JPH05164801A - Diagnostic method and its device - Google Patents

Abstract

PURPOSE:To effectively respond to an unexpected trouble by extracting universal signs of troubles from plural information of behavior conditions, accumulating troubles possible to take place as virtual troubles, and devising counter-measures thereto in accordance with priority order. CONSTITUTION:A virtual trouble formation mechanism 75-2 reads the change of signs causing troubles from the contents of varieties of troubles example-base, and consolidates signs just one step before happening of troubles, creates new virtual troubles and adds these signs to a virtual trouble example-base 74. Besides, when a trouble verification mechanism 76 decides a new trouble, the virtual trouble formation mechanism 75-2 adds it to the virtual trouble example- base 74 through a virtual trouble holding mechism 77. The holding mechanism 77 decides the priority order of a trouble according to importance and frequency of the trouble, effectually renews the virtual trouble example-base 74, and a trouble analyzing mechanism 78 specifies the cause of a trouble, decides the countermeasure and prepare an alerting signal Q. This method is effectual to not only cope with an unexpected trouble, but also previously prevent the occurrence of a trouble by foreseeing it.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a diagnostic method and apparatus for diagnosing an electronic device such as a copying machine.

[0002]

BACKGROUND OF THE INVENTION The use of expert systems as a means of problem solving is known in the art. For example, as described in Japanese Patent Application Laid-Open No. 2-100141, in a production rule type expert system, inference is performed by executing a production rule in which a condition and a conclusion are paired.

Further, there is a technique in which a supervisory controller makes an inference based on information from an object to be monitored.
It is disclosed in the publication.

In a diagnostic system having such an inference mechanism, it is necessary to prepare a large number of rules in advance as a knowledge base. The accuracy of the diagnosis result by the expert system depends on whether the prepared rules are appropriate.

[0005]

However, in the conventional diagnostic method, the rule is such that the basic principle is originally clear, or the rule that can be easily predicted from the experience so far is adopted, There was a problem that we could not deal with troubles and obstacles that could not be predicted at all.

For example, before launching a new model copying machine to the market, sufficient consideration is made to prevent the occurrence of troubles, and countermeasures are taken in advance for factors that may cause troubles. Factors that are likely to cause this trouble are incorporated into the knowledge base as rules and used when diagnosing by the expert system. However, in reality, there are not a few troubles that become apparent after a large amount of products are put on the market. Moreover, there are many types of trouble that cannot be predicted based on past experience.
Particularly in the paper transport system of a copying machine, not only the mechanical conditions of the transport mechanism but also the nature of the paper and changes in the atmosphere such as temperature and humidity have a delicate influence on the transport characteristics. It cannot be reproduced in the laboratory environment alone.

In addition, even if a new trouble occurs, the service person simply takes measures to eliminate the trouble, and the information that a new trouble which has not been known until now has occurred is sent to the service center. And other servicemen do not provide effective feedback.

Further, in order to effectively perform the diagnosis, it is necessary to incorporate the newly revealed trouble and its cause into the knowledge base of the expert system as a new rule, but the device for realizing the expert system is For example, there is a problem that it takes time and effort to update the knowledge base because the service stations are located all over the country. In addition, the expert system itself may not be able to handle the update of the knowledge base, and in this case, the knowledge becomes obsolete immediately.

Further, when performing diagnosis, since all possible troubles are sequentially diagnosed, there is a problem that the diagnosis takes a very long time and the efficiency of the diagnosis is low when there are many diagnosis items. there were.

Therefore, it is an object of the present invention to provide a diagnostic method and apparatus capable of coping with an unexpected trouble.

Another object of the present invention is to make a diagnosis efficiently by prioritizing the troubles to be diagnosed.

[0012]

SUMMARY OF THE INVENTION The present invention is a diagnostic method for diagnosing a device to be diagnosed on the basis of operating state information indicating the operating state of the device to be diagnosed. The operating state information indicating the operating state is collected, a universal symptom showing common characteristics with respect to the plurality of diagnosed devices is extracted from the plurality of operating state information, and the universal symptom is generated in the diagnosed device based on the universal symptom Determine a trouble that may occur, accumulate the trouble as a virtual trouble in the virtual trouble case base, assign a priority according to the urgency of the trouble countermeasure to a plurality of virtual troubles in the virtual trouble case base, and It is characterized in that the plurality of virtual troubles are sequentially verified based on the priority order.

The present invention is also a diagnostic device for diagnosing a device to be diagnosed based on operating state information consisting of a plurality of factors indicating the operating state of the device to be diagnosed.
From a plurality of communication control devices, which are provided corresponding to each of the plurality of devices to be diagnosed, which obtain operation state information indicating an operation state of each device to be diagnosed and which are transmitted to the outside, A host computer that collects the plurality of operation state information, and a diagnosis unit that diagnoses the state of the device to be diagnosed based on the plurality of operation state information obtained from the host computer. Analyzing means for analyzing the plurality of operating state information to obtain a universal symptom exhibiting common characteristics for the plurality of devices to be diagnosed, and universal for accumulating the universal symptom obtained by the analyzing means. And a virtual trouble generating means for generating a virtual trouble that may occur in the device to be diagnosed based on data from the universal symptom case base. A virtual trouble case base for accumulating the virtual troubles, a means for giving priority to a plurality of virtual troubles in the virtual trouble case bases according to a trouble countermeasure urgency level, and a virtual trouble case in the virtual trouble case bases. And a trouble verification means for sequentially verifying the validity in accordance with the priority order.

[0014]

In the present invention, the operating condition information indicating the operating condition of a plurality of devices to be diagnosed, such as a copying machine, is collected in the diagnostic device. When a trouble occurs in a certain device to be diagnosed, all operating state information is sent to the diagnosis device together with data indicating the type of trouble that has occurred. In diagnostic equipment,
A universal symptom showing common characteristics with respect to a plurality of devices to be diagnosed is extracted from a plurality of operating state information. This universal symptom indicates a symptom of trouble common to all the diagnosed devices. Therefore, when the frequency of occurrence of a particular trouble is high, it can be predicted that the same trouble will occur in the future even if the trouble is not actually occurring at present. If trouble countermeasures are taken in advance based on this prediction, trouble occurrence can be prevented in advance.

For the troubles extracted as described above, the priority order is determined according to the urgency of the trouble countermeasure, and the troubles having higher importance are sequentially verified.

[0016]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The features of the present invention will be specifically described below with reference to the accompanying drawings.

FIG. 1 is a schematic block diagram of a diagnostic device to which the diagnostic method of the present invention is applied. In the present embodiment, the case of diagnosing a copying machine will be described as an example.

Each of the plurality of copying machines 1 has a copying machine main body 2
And the communication control device 3 is built in. Each communication control device 3 obtains from each copying machine main body 2 the status of each copying machine main body 2, for example, operation status information such as presence / absence of paper jam or jam, quality of image quality, etc., through a communication line 4 such as a telephone line. Send to the host computer 5. An expert system server 6 is connected to the host computer 5, and the expert system server 6 makes a diagnosis based on the information sent from the host computer 4, and sends back the diagnosis result, coping method, etc. to the host computer 5. The expert system server 6 is constructed on an ordinary workstation or computer, and includes a knowledge base system 7 including a knowledge base that is expert knowledge and an inference engine that makes an inference using this knowledge base. It comprises a knowledge base system 7 and a shell 8 for controlling communication between the host computer 5.

Each of the copying machines 1 is installed in a customer's office or business office, and the host computer 5 and the expert system server 6 include a factory, a business office, a service station, etc. of the manufacturer who manufactures and sells the copying machine 1. Is installed in. It should be noted that the models of the plurality of copying machines 1 are not necessarily the same, and a plurality of models may coexist.

FIG. 2 schematically shows the basic structure of the sheet conveying mechanism and the image forming mechanism of the copying machine 1.

Paper feed trays 11a to 11a in the upper, interrupted, and lower stages
The paper 12 stored in any one of the paper feed trays 1c is taken out by any of the paper feed devices 13a to 13c. The sheet taken out from the sheet feeding tray 11 is conveyed along the conveying path A by the conveying device 14 including a plurality of conveying rollers, and is further abutted on the alignment gate 16 by the slip roller 15.

On the other hand, a toner image is formed on the belt-shaped photosensitive member 17 rotating in the direction of the arrow by the electrophotographic image forming means described later. The alignment gate 16 is opened in synchronization with the movement of the toner image on the photoconductor 17,
The sheet is conveyed to the transfer position 19 by the alignment roll 18 and is brought into close contact with the photoconductor 17, and the toner image on the photoconductor 17 is transferred onto the paper by the transfer means described later. The transferred sheet is conveyed to the fixing device 21 by the conveying belt 20 and undergoes a fixing process. The paper after fixing is the exit roll 22 of the fixing device.
The fixing device outlet gate 23 switches the conveyance path. In the case of single-sided printing, it is conveyed to the route B, and in the case of double-sided printing or single-sided multiple printing, it is conveyed to the route C.

The path of the sheet passing through the path B is further switched to the path D and the path E by the reversing gate 24. In the case of the path D, the paper is conveyed by the conveying rolls 25 and the exit rolls 26 with the image surface facing upward, passes through the path F, and is discharged onto a paper discharge tray (not shown) with the image surface facing upward. In the case of the path E, the front and back of the sheet are reversed by the reversing device 27, and then the paper passes through the paths G and F and is discharged with the image surface facing down.

After the sheet passing through the path C has its warp corrected by the decurler 28, the double-sided printing gate 29 switches between the path H and the path I. In the case of the path H, the paper is discharged to the intermediate tray 31 by the double-sided printing roll 30 with the image side facing down. Further, in the case of the path I, the sheet conveying direction is reversed by the reversing device 32, passes through the path J, and is discharged onto the intermediate tray 31 by the double-sided printing roll 30 with the image side facing upward.

The paper in the intermediate tray 31 is sent out by the paper feeding device 33 and is composed of a plurality of carrying rolls.
4, the toner image on the photoconductor 17 is transferred onto the paper as is the paper from the paper feed trays 11a to 11c. Here, when the image surface of the paper in the intermediate tray 31 faces downward, that is, when the paper is discharged through the path H, a new image is formed on the side opposite to the surface on which the image is already formed. An image is formed on the sheet and double-sided printing is achieved. Further, when the image surface of the paper in the intermediate tray 31 faces upward, that is, when the paper is discharged via the paths I and J, it is newly added to the same side as the surface on which the image is already formed. Images are formed and single-sided multiplex printing is achieved.

In the case of double-sided printing or single-sided multiplex printing,
After the second image is formed, it is subjected to the fixing process by the fixing device 21 as in the case of single-sided printing, and then is ejected to a paper ejection tray (not shown) in the same posture or with the front and back reversed. It

A plurality of paper sensors 35a to 35j are arranged along the paper conveyance path from the paper feed tray 11 or the intermediate tray 31 of the copying machine to a paper discharge tray (not shown). The presence or absence of paper at the position is detected. For example, the sheet sensor 35a arranged in the sheet feeding device 13a detects that the sheet is fed out from the sheet feeding tray 11a, and the sheet sensor 35g arranged immediately after the fixing device 21 causes the sheet to pass through the fixing device 21. Detect what you have done.

The outputs of these sheet sensors 35a to 35j are supplied to a conveyance control CPU (central processing unit) 41 as shown in FIG. The conveyance control CPU 41 controls the operation and stop of various drive mechanisms such as the conveyance rollers via the paper conveyance control device 42 according to the outputs of the paper sensors 35a to 35j, and conveys the paper at a predetermined timing and speed. .. Further, the transport control CPU 41 is configured to control each paper sensor 35a.
The occurrence of a paper jam, that is, a jam in the transport path is detected based on the output state of .about.35j. In the following description, the locations where the sheet sensors 35a to 35j are arranged are referred to as jam detection locations J1 to J10. It should be noted that in the present embodiment, in order to simplify the explanation,
Only one sheet sensor is disclosed, and therefore, there are only 10 jam detection points, but in reality, a larger number of sheet sensors are provided so that a jam can be detected at many points. To do.

The occurrence of a jam is due to whether the paper sensor detects the paper within a fixed time after the paper feeding is started, or whether the paper is not detected within a fixed time after the paper sensor detects the paper. Alternatively, it can be detected by determining whether the paper has passed between a plurality of paper sensors within a predetermined time. The presence / absence of a jam is detected for each sheet sensor and accumulated in a nonvolatile memory or the like built in the conveyance control CPU 41 as the number of jams at each location. These pieces of information are used as a copy machine history (fault log) indicating jam detection locations J1, J2, J3, ..., J10, the number of jam occurrences (N ₁ , N ₂ , N ₃ , ..., N ₁₀ ). For example, it is stored in the form as shown in Table 1. Note that the jam detection location here means a location where a jam occurrence can be detected regardless of whether or not an actual jam occurs.

[0030]

[Table 1] The copy machine history stores the number of paper feeds, the average paper passage time at each location, the total number of copies, and the like, in addition to the above-described jam detection locations and jam occurrence counts.

Further, the latest N times (for example, several tens of times) of trouble data, for example, codes K1, K2, K3, ..., KN indicating the kind and place of trouble and a copy at the time when the trouble occurs. Number of sheets L ₁ , L ₂ , L
₃ , ..., L _N are stored in the form as shown in Table 2 as a short-term trouble history (shutdown history). The value of N is, for example, 20-40.

[0032]

[Table 2] As for trouble data N times or more, the oldest data is discarded every time new data is written. The above-mentioned short-term trouble history is reset by the service person when the service person performs maintenance on the copying machine.

The jam occurrence location is information on the paper jam location detected by each paper sensor. For example, if the sheet is still detected after a predetermined time has elapsed after the presence of the sheet was detected by a certain sheet sensor, it is considered that a jam has occurred at that position. The average sheet passage time indicates an average value of sheet passage times from one sensor to another sensor.

Various kinds of information related to the above-mentioned jam are supplied from the conveyance control CPU 41 to the communication control device 3. The details of the communication control device 3 will be described later.

Next, an electrophotographic image forming means for forming an image on the photoconductor 17 will be described with reference to FIG.

The belt-shaped photoreceptor 17 is wound around a plurality of support rollers 51. This photoconductor 1
7, a charging device 52, a laser output scanner (RO
S) or other exposure device 53, patch generation device 54, surface potential sensor 55, first and second developing devices 56 and 57, pre-transfer corotron 58, density sensor 59, transfer device 60, peeling device 61, cleaning device 62. Etc. are sequentially arranged.

The surface of the photoconductor 17 is uniformly charged by the charging device 52 and then exposed by the exposure device 53 in a pattern corresponding to the image to be output, and the surface of the photoconductor 17 corresponding to the image to be output. An electrostatic latent image is formed. This electrostatic latent image is developed by the first or second developing device 56, 57 to form a toner image on the photoconductor 17. This toner image is transferred by the transfer device 60 onto the paper conveyed from the paper feed tray 11. The sheet after transfer is peeled from the photoconductor 17 by the peeling device 61, and is conveyed by the conveying belt 20 toward the fixing device 21. Photoconductor 17 after transfer
The toner remaining on the top is removed by the cleaning device 62,
Prepare for the next image forming cycle.

Although not shown, an image input device for reading an image of a document to be copied and converting it into an image signal is provided above the photoconductor 17, and the document is automatically fed to the image input device. An automatic document feeder is provided. The image signal obtained from the image input device is supplied to the exposure device 53 shown in FIG.

At the time of image formation, a predetermined reference voltage is applied from the high voltage controller 43 shown in FIG. 3 to the patch generator 54 before copying the image of the document, and the patch generator 54 causes the photoreceptor 17 to be imaged. A patch having a predetermined potential on it is generated. This patch is the developing device 56 or 57.
The density of the patch after development is detected by the density sensor 59. The output of the density sensor 59 is supplied to the image forming control CPU 44 shown in FIG. 3, and the density control device 45 is controlled so that the detected density matches a predetermined reference density. Specifically, the developing bias, the toner supply amount, and the like in the developing devices 56 and 57 are controlled.

Further, the surface potential of the photoconductor 17 other than the patch portion is detected by the surface potential sensor 55. The output of the surface potential sensor 55 is supplied to the image forming control CPU 44, and the charging device 52 and the like are controlled via the high voltage control device 43 so that the detected surface potential matches a predetermined reference potential.

When the image of the original is copied, the copying operation is performed based on the image forming conditions initialized as described above.

Further, the voltage applied to the patch generator 54 is detected by the voltage detection circuit 46, and the data indicating the voltage applied to the patch generator 54 is supplied to the image forming control CPU 44.

The above-mentioned development density, the surface potential of the photosensitive member 17,
Each data indicating the voltage applied to the patch generation device 54 is supplied from the image formation control CPU 44 to the communication control device 3.

The above-mentioned conveyance control CPU 41 and image formation control CPU 44 are connected to the main CPU 47. Based on an instruction from the main CPU 47, the paper conveyance operation by the conveyance control CPU 41 and the image formation operation by the image formation control CPU 44 are performed. The copying operation is performed synchronously.

The communication control device 3 controls the image forming control CP and the information related to the sheet conveying operation from the conveyance control CPU 41.
The information related to the image forming operation is received from U44 and the information is stored. And the communication control device 3
Detects whether or not a trouble has occurred in the copying machine, and if a trouble has occurred, an identification code indicating the location and type of the trouble (determined trouble I
Along with D), all the information stored in the communication control device 3 is transmitted to the host computer 5 shown in FIG. 1 via the communication line 4.

The transmission information sent from each copying machine 1 to the host computer 5 includes the model name indicating the model of the copying machine, the serial number, the customer name, the output state of each sensor, and the like. Is represented by, for example, binary data and is transmitted in a packet format.

The host computer 5 sends various information in binary representation sent from each copying machine 1 to, for example, A
It is converted to a data file of SCII representation and sent to the shell 8 of the expert system server 6. The shell 8 converts this data file into a hash table. The knowledge base system 7 takes out necessary information from the hash table and makes a diagnosis.

Next, the structure of the knowledge base system 7 for making a diagnosis will be described with reference to FIG.

Knowledge-based system 7 in this embodiment
Has a basic principle fact base 71, a universal symptom case base 72, an individual case base 73, and a virtual trouble case base 74 as knowledge bases. Each base will be described below.

The basic principle fact base 71 is a fact base that memorizes facts that are clear as a basic principle, and the investigation of the cause and the selection of coping means for a confirmed trouble are all based on the knowledge from the basic principle fact base 71. It is decided based on. The meaning of the confirmed trouble will be described later. The universal symptom case base 72 organizes and stores the symptom peculiar to a model in a statistically processed form.
Holds statistically universal manifestations of observed phenomena for a model and the relationships between phenomena. The individual case base 73 is for organizing and storing the signs of individual specific models in a statistically processed form, and peculiar information such as defective products is stored in the individual case base 73. Virtual trouble case base 74
Is for accumulating and saving cases of troubles that may occur in a certain model, and trouble prediction diagnosis described later is performed by applying and verifying the cases in these cases to the machine to be diagnosed.

Furthermore, the knowledge base system 7 shown in FIG.
Includes a learning mechanism 75, a trouble verification mechanism 76, a virtual trouble maintenance mechanism 77, and a trouble analysis mechanism 78.

The learning mechanism 75 is a mechanism for analyzing observation data from the machine to be diagnosed and for learning and holding necessary information. It should be noted that whether or not learning is necessary is determined in advance for most items, but may be newly determined for items having a high correlation described later. The learning mechanism 75 is composed of an input information analysis mechanism 75-1 and a virtual trouble generation mechanism 75-2. The input information analysis mechanism 75-1 analyzes the input information received from the input / output mechanism (not shown), that is, the machine information P to be diagnosed,
The accumulated universal symptom is statistically processed, and then the universal symptom case base 72 is updated. In addition, the individual information of the machine to be diagnosed is secured in the internal work area until the end of the diagnostic session, that is, until the alert output described below is generated. Base 7
Update 3 and store this. The expert system itself automatically determines whether or not to continue observation based on a predetermined rule. That is, when the diagnosed copying machine shows a sign different from the general sign, the copying machine is unique and is designated as a device requiring continuous observation.

When the input information includes the confirmed trouble ID, the confirmed trouble T1 is passed to the trouble analysis mechanism 78 described later. In addition, fixed trouble I
D is predetermined according to the type of trouble, and is generated on the copying machine side when the trouble actually occurs on the copying machine side.

Further, the virtual trouble generating mechanism 75-2 is
When the input information includes the fixed trouble ID, the change of the symptom that causes the trouble is calculated backward and read from the contents accumulated in the various case bases. Then, a set X is created by organizing various signs just before the occurrence of the trouble. Here, a new virtual trouble H is generated. The above set X becomes a logical consequential element when proving this virtual trouble H. Both of them are added to the virtual trouble case base 74 and updated. Details of this processing will be described later.

Next, the trouble verification mechanism 76 is a mechanism for verifying whether or not the virtual trouble in the virtual trouble case base 74 is a true trouble, and the trouble verification mechanism 76 has rules necessary for judgment. And data (details will be described later).

The true trouble verified by the trouble verification mechanism 76 is output as a confirmed trouble T2 and passed to the trouble analysis mechanism 78. In addition, fixed trouble T
2 is passed to the correlation analysis mechanism 75-2a, and another trouble having a high correlation is passed to the trouble verification mechanism 76 as a new trouble candidate. When it is determined that this candidate is a new trouble, the virtual trouble case base 7 is set as a new virtual trouble via the virtual trouble maintenance mechanism 77.
Added to 4.

The virtual trouble maintenance mechanism 77 is for efficiently and constantly updating the virtual trouble case base 74, and includes a virtual trouble ranking determination mechanism 77-1. The virtual trouble ranking determination mechanism 77-1 determines the priority of the trouble to be verified based on the importance of the trouble and the occurrence frequency of the trouble, and is composed of necessary rules and data.

The trouble analysis mechanism 78 is a mechanism for generating the alert output Q for determining the cause and determining the coping means based on the basic principle fact base 71 when the confirmed troubles T1, T2 are received. .. The trouble analysis mechanism 78 includes a cause investigation mechanism 78-1 configured by rules and data necessary for cause determination, and a handling determination mechanism 78-2 configured by rules and data required for handling determination. .. Note that the rules and data used in the above-mentioned cause investigation mechanism 78-1 and handling decision mechanism 78-2 are general ones that can be estimated from conventional experience.

Next, the diagnostic operation based on the knowledge base shown in FIG. 4 will be described. The diagnostic operation is started, for example, when information indicating the operating state of the copying machine, that is, data indicating a trouble (confirmed trouble ID) appears in the diagnosed machine information P, or periodically. Further, it may be activated by a manual operation of a customer or a service person.

Here, an outline of the diagnosis process when the copier is diagnosed will be described with reference to FIG. The details of each process will be described later.

First, the long-term observation information is analyzed (step 101). The long-term observation information is the accumulation of various information obtained from all the copiers to be diagnosed over a long period of time. In other words, after a sufficient number of diagnoses have been performed, It is the diagnostic result obtained in. With this long-term observation information, it is possible to obtain statistics on what kind of trouble occurs at which place in a certain model.

Next, a virtual trouble is determined based on the analysis result of the long-term observation information (step 102). A virtual trouble is a trouble that may occur in a certain model when a certain type of copier is targeted for diagnosis.
This trouble is determined based on the analysis result of the long-term observation information. For example, if an analysis result has been obtained that a jam is likely to occur at a specific position on a certain model, even if the copying machine under diagnosis is not actually jammed, the above-mentioned identification may be performed in the future. Jam is likely to occur at the position. Therefore, the occurrence of a jam at this specific position is determined as a virtual trouble with the highest degree of precedence. For other troubles, prioritize according to the overall occurrence frequency.

Next, the short-term observation information is analyzed (step 103). Short-term observation information means the latest information obtained for an individual copying machine within a limited period of time. This short-term observation information corresponds to the shutdown history described above and has the latest information for each copying machine. Therefore, in the copying machine to be diagnosed, the recent operation history of the copying machine is used. Can know.

Next, the virtual trouble determined in step 102 is weighted based on the analysis result of the short-term observation information (step 104). This weighting is performed in order to enable reasoning for each individual copying machine in the shortest time in hypothesis reasoning described below.

Next, based on this weighted virtual trouble, hypothesis inference is carried out sequentially from the one having the highest importance and the highest occurrence frequency (step 105). It should be noted that the importance and the frequency of occurrence are weighted respectively, and the inference is performed from the one that has the greatest influence when a trouble occurs. This hypothesis is verified by analyzing the input information of the copier under diagnosis (step 106). That is, regarding the copying machine under diagnosis, the number of jams, the number of fed sheets,
By sequentially examining the average paper passing time, etc., based on a predetermined diagnostic algorithm, whether or not a virtual trouble is appropriate, that is, whether this trouble actually occurs or may occur. Verify whether it is high. This verification is performed by a predetermined number of troubles in the order of virtual troubles that are likely to occur in the copying machine under diagnosis. For verified virtual troubles,
A confidence value of 0 to 100 indicating the certainty is attached and output. Then, while pursuing the cause of the verified virtual trouble, the coping method is determined (step 10).
7). Further, a trouble having a high correlation with the hypothesis whose hypothesis has been verified is generated as a new virtual trouble (step 108), and the process returns to step 102. The details of the diagnostic algorithm will be described later.

The above-mentioned diagnosis is carried out for all copying machines when a trouble occurs in a certain copying machine.
The above-mentioned long-term observation information and short-term observation information are sequentially updated. Therefore, the long-term observation information includes universal symptoms common to a certain model, and the short-term observation information includes individual symptoms of each copying machine. From the universal signs, troubles that are likely to occur in the model can be known in advance, and at the time of periodic inspection etc., by replacing parts and adjusting the device in advance and eliminating the cause of the trouble It is possible to prevent troubles that cause trouble.

Details of the diagnosis will be described below with reference to FIG.

The diagnosis target machine information, that is, the diagnosis target machine information P from each copying machine 1 shown in FIG.
Of the input information analysis mechanism 75-1. The diagnostic machine information P is input, for example, to each copying machine 1
The value of various information indicating the operating state of is above a predetermined threshold value. This input information includes, for example,
Jam occurrence location, number of jam occurrences, number of paper feed at the time of occurrence, average paper passage time at each location, copier serial number, total copy number, photoconductor surface potential, patch generator potential, copy density value, etc. Contains information. The input information analysis mechanism 75-1 analyzes these input information. Hereinafter, a specific example of analysis of this input information will be described.

In analyzing the input information, the input information analyzing mechanism 75-1 sequentially accumulates and holds the number of jams, the number of fed sheets, etc. at each location for each diagnosis. Then, from these pieces of information, the variation in the occurrence of each jam is obtained by a statistical method. Specifically, the statistics of the following items are collected every time.

Total Number of Occurrences at Each Jam Occurrence Point Total Number of Paper Feeds at Each Jam Occurrence Point Standard Deviation in Jam Occurrence Deviation Value Based on Standard Deviation Each Average Occurrence Rate Each Average Paper Passage Time The universal symptom case base 72 is updated with the symptom universally found in the model, that is, the data indicating the property.

Further, the individual information of the machine to be diagnosed, that is, the copying machine, is input to the input information analysis mechanism 75-until one diagnosis is completed.
For a copying machine which is secured in the work area provided inside 1 and requires continuous observation, the individual case base 73 is updated with individual information. In addition, the copier requiring continuous observation means a copier exhibiting a peculiar symptom deviating from the general symptom.

When the input information includes a confirmed trouble, the confirmed trouble T1 is passed to the trouble analysis mechanism 78. Since the raw data obtained from the copying machine includes data that is not necessary for trouble analysis, only the necessary data is selected and the fixed trouble ID is added to make it the fixed trouble T1.

Next, how the input information is specifically analyzed will be described.

For example, the total number of occurrences at each jam occurrence point will be described. As shown in Table 2, at which position the jam has occurred is determined for each diagnosis. The circles in the table indicate the locations where the jam occurred. Also,
At the location where the jam occurs, the cumulative number of jams at the time of the diagnosis is stored.

[0075]

[Table 3] In the example of Table 3, in the first diagnosis, jams occur at three jam detection points J1, J2, and J10. In the second diagnosis, a jam occurs at the jam detection point J2 and the other two points, for a total of three points, and the total number of confirmed jam points is five points. Similarly, the location of the jam is confirmed for each diagnosis. In the example of Table 3, the number of times of diagnosis is M times counted from the time when the content of the short-term history is reset by the service person, and it is shown that jams are confirmed at all places, that is, 10 places in the example of Table 3. There is.

The information on the number of jam occurrences shown in Table 3 is stored in the buffer for the latest M times of diagnosis together with the data indicating the number of jam occurrences at each location in chronological order. Old data is discarded.
The capacity of the buffer (buffer length) is set in correspondence with the average number M of times of diagnosis that jams are confirmed at all locations (10 locations in this embodiment). The value of M is 100
The above is desirable. This buffer exists in the universal symptom case base 72 shown in FIG. 4, and the information in the buffer corresponds to the short-term observation information.

Further, in the storage area provided separately from the buffer for storing the information shown in Table 3, as long-term observation information,
The number of jams generated at the jam detection points J1 to J10 is accumulated for each diagnosis and stored as the total number of jams.

If these pieces of information are accumulated in a large number of copying machines of the same model over a long period of time, in other words,
If the number of times of diagnosis is sufficiently large, for example, about 1000 times in total, it is possible to know the tendency of trouble occurrence commonly occurring in a certain model based on these information. That is, it is possible to learn not only the characteristics peculiar to a specific device but also the characteristics universally found in a certain model.
Therefore, based on this learning result, it is possible to predict in which part a jam easily occurs in a certain model, and a virtual trouble can be generated as described later.

However, if learning is simply performed from the information on the total number of jams at each location, inconvenience may occur in some cases. This inconvenience will be described below.

For example, it is assumed that jams frequently occur at the jam detection location J1 immediately after the introduction of the copying machine. In this case, the learning mechanism 75 shown in FIG. 4 learns that a jam is likely to occur at the jam detection location J1 and makes a diagnosis based on the learning result. As a result of the diagnosis, the cause of the jam is found to be an initial defect of the component, and if the component is replaced, the jam at this position is reduced. However, it is judged from the past history that a jam is still likely to occur at the jam detection location J1, and virtual troubles are generated in the priority order based on this, and hypothesis inference is performed. In other words, for a while after the parts are replaced, hypothetical reasoning is performed based on old knowledge, even though the current state of the device is changing, so it takes longer to determine the true cause. The problem arises.

Therefore, in this embodiment, the input information is divided into long-term observation information and short-term observation information, for example, M diagnostic results in the buffer are used as short-term observation information, and the virtual information is calculated based on this short-term observation information. Determine the priority of the trouble. As a result, after the countermeasure is taken against the trouble, this trouble has a lower priority as a virtual trouble in the device, and the hypothesis is not unnecessarily verified.

Next, the virtual trouble generation mechanism 75-2 of the learning mechanism 75, when the input information includes the confirmed trouble ID, the contents accumulated in various case bases such as the universal symptom case base 72. From the above, read back the change in the symptoms that caused the trouble. Then, a set X is created in which only the necessary signs are sorted out from the various signs just before the occurrence of the trouble.

For example, the history of signs of jam occurrence at each jam detection location is stored in the universal sign case base 72 as short-term observation information, and the history is shown in FIG. Now, when the confirmed trouble ID indicating the occurrence of a jam is detected in the nth diagnosis,
(N-1) Read changes in the contents of the universal symptom case base 72 at the time of the diagnosis. In the example shown in FIG. 6, when it is detected that a jam has occurred at the jam detection location J1, other jam detection locations J2, J3, ...
・ Read the paper passage time at. The set of sheet passage times at each of these jam detection locations is X. Here, a new virtual trouble corresponding to the set X is defined as a virtual trouble H. At this point, the virtual trouble H is only a name, and the substance does not yet exist. This set X and virtual trouble H are virtual trouble case base 7
4 and the case base 74 is updated.

The trouble verification mechanism 76 refers to the virtual trouble case base 74 to verify whether the virtual trouble is a true trouble. That is, it verifies whether the hypothesis is correct. This determination process will be described with reference to the trouble verification tree shown in FIG.

FIG. 7 is an explanatory diagram schematically showing the verification logic in the trouble verification mechanism 76 based on the virtual trouble case base 74. In this embodiment, all knowledge bases such as the virtual trouble case base 74 are designed based on the object-oriented representation. Therefore, all knowledge and rules are defined in the object. The object is composed of a plurality of slots, and the data and rules required for verification are held in the slots.

The operation of the trouble verification mechanism 76 will be described below, focusing on jam troubles. The object here means the same thing as a frame in a frame type language.

In FIG. 7, there are "jam trouble" and "image quality trouble" as virtual trouble cases. Further, in the "jam trouble" case, a plurality of cases "jam trouble 1", "jam trouble 2", ... .. is indicated. Then, there are two logical consequential elements (“Certification element 1” and “Certification element 2”) necessary for “Jam trouble 1” to be established, and a consequent element logically obtained from the hypothesis of the element. Are two (“certification element 1-1” and “certification element 1-2”). "Jam trouble 1", "Jam trouble 2", ... Correspond to each jam occurrence point, and "certification element 1", "certification element 2"
.. corresponds to the severity of jam occurrence, severity of increased jam, severity of jam frequency, and importance of jam deviation value, which will be described later.

In FIG. 7, each of the “virtual trouble cases”, ..., “Proof element 1-2” surrounded by a frame indicates an object, and each object stores necessary data in its own slot. And keep the necessary data in my method slot.

The method included in each object is activated by a message sent from another object from another object. The trouble verification mechanism 76 controls activation of this method, that is, control of verification such as message transmission.

Here, the case of verifying "jam trouble 1" will be described. In this case, for each of the "proof element 1" and the "proof element 2", it is tested whether or not each holds. For this verification, a message "Verify" for certifying the two objects "certification element 1" and "certification element 2" is sent. Each of the objects "certification element 1" and "certification element 2" that have received the message "Verify" activates the method corresponding to the message and makes the determination. However, at this stage, the pros and cons thereof are not determined, and an object of a lower hierarchy to prove itself, for example, "certification element 1-
Similar messages are sent to "1" and "certification element 1-2".

When the object at the lowest level is reached in this way, each object returns the result to the upper level object. The upper object sums up the results from the lower objects, decides the propriety of itself, and returns the propriety to the upper object. In this way, the pros and cons of "jam trouble 1" are finally judged.

In the above proof, each object returns a belief value having a value between 0 and 100, instead of returning whether the result is binary or not. The upper-level object also determines its own certainty value in consideration of the plurality of results and the certainty values associated with them. Finally, "Jam Trouble 1" is 0 to 1
With the certainty value of 00, the verification result of the trouble on the machine to be diagnosed is estimated. That is, the confidence value of the verification result is 10
The closer to 0, the higher the possibility of trouble, and the closer to 0, the lower the possibility of trouble.

Next, specific examples of "certification element 1", "certification element 2", etc., which are subordinate to "jam trouble 1", will be described. In the present embodiment, “jam occurrence severity”, “jam increase severity”, “jam frequency severity”, and each proof element are subordinate to “jam trouble 1”.
Four severity levels named "Jam Deviation Severity" are set, and the degree of jam trouble necessary for determining a jam trouble is determined from these severity levels. Although the trouble verification tree is generally shown in FIG. 7, in the following description, there is only one proof element under “certification element 1”.

Each severity will be described below.

Jam occurrence severity (X1): Jam occurrence rate j / f obtained by dividing the total number j of jams generated up to the present at a position by the current number f of sheets fed at that position.
With the argument as an argument, the one obtained by the following function is the jam occurrence severity.

When f = 0 X1 = 0 When f ≠ 0 X1 = j / f × 10000 When X1> 100 X1 = 100 This value was found in the short-term observation information (shutdown history) described above. It is calculated not only by the number of jams, but also by the total number of jams obtained by examining all of the long-term observation information (fault log).

[0097] Jam increasing severity (X2): Focusing on the most recent three jam generation timing T _1, T _2, the number of sheet feeding in the _{T 3 N P1, N P2,} N P3, jam occurs at time T ₁ The number of fed sheets ΔA (= N _P2 −N _P1 ) from the beginning to the time point T ₂ ,
Number of fed sheets ΔB (= N _P3 − from time T ₂ to time T ₃
N _P2 ) and shows that the jam occurrence tends to increase or decrease depending on the length of the occurrence interval. The jam increase severity X2 is calculated from the following equation, and the larger the value, the higher the jam occurrence frequency. Note that this value is calculated only for jams found in the shutdown history.

When ΔB = 0, X2 = 0 When ΔA ≠ 0, X2 = (ΔA / ΔB−1.0)
X100 X2> 100 X2 = 100 X2 <0 X2 = 0 Jam frequency severity (X3): When the number of jams at a certain position is N _J , it is expressed by the following formula, and this value N _J is It means that the larger the size, the higher the possibility of determining a trouble. Note that this value is calculated only for jams found in the shutdown history.

When N _J ≧ 5 X3 = 100 When N _J <5 X3 = a × N _J ³ + N _J Jam Deviation Value Severity (X4): The standard deviation regarding occurrence is calculated from the number of jams at each position, Further, each deviation value S is calculated. The deviation value S is applied to the following formula to obtain the jam deviation value severity X4. Note that this value is calculated not only by the jam found in the shutdown history but also by the total number of jams obtained by inspecting all the fault logs. Also, this jam deviation value severity X
4 is calculated based on the information obtained for all the copiers diagnosed so far. That is, while the other three severity levels X1, X2, and X3 indicate the diagnostic evaluation depending on the individual information of each copying machine, the jam deviation value severity level X
4 indicates a universal diagnostic evaluation common to a certain model.

When S> 75 X4 = 100 When S <25 X4 = 0 When 25 ≦ S ≦ 75 X4 = 2 × (S-25) This jam deviation value severity X4 is a jam that occurs at each location. Among them, it is required to be able to deal with the most troublesome ones.

The above three severity levels X1, X2, X3, X
4 is weighted as in the following formula to obtain the jam trouble degree Y required for jam trouble determination.

Y = (X1 × n1 + X2 × n2 + X3 × n3 + X4 × n4) However, 0 ≦ Y ≦ 100, 0 ≦ n1 ≦ 1,0 ≦ n2 ≦ 1,0 ≦ n3 ≦ 1,0 ≦ n4 ≦ 1 n1 + n2 + n3 + n4 = 1 0 ≦ X1 ≦ 100, 0 ≦ X2 ≦ 100, 0 ≦ X3 ≦ 100, 0 ≦ X4 ≦ 100.

The adequacy of the virtual trouble is judged by the jam trouble degree Y obtained as described above. That is, the hypothesis is verified. Jam trouble level Y
When is equal to or greater than a certain value, the virtual trouble is regarded as a true trouble and becomes a confirmed trouble.

In the above description, the hypothesis verification regarding the jam has been described in detail, but the hypothesis verification is similarly performed for the image quality and other items. At the time of verifying this hypothesis, a plurality of trouble cases accumulated in the virtual trouble case base 74 are sequentially verified. This verification is performed by the virtual trouble order determining mechanism 77 in the virtual trouble maintaining mechanism 77. Based on the priority order determined in advance by -1, only a predetermined number of virtual troubles are executed. The confirmed trouble T2 determined as described above is passed to the trouble analysis mechanism 78.

Further, the trouble T2 determined by the trouble verification mechanism 76 is passed to the virtual trouble generation mechanism 75-2 in order to seek a true trouble by further obtaining a correlation with another trouble (route R). The correlation analysis mechanism 75-2a provided in the virtual trouble generation mechanism 75-2 is
Analyze the correlation between each trouble and each symptom. That is, the correlation between the occurrence of a certain trouble and another trouble or symptom is obtained by regression analysis for each diagnosis. And
For troubles with high correlation, the trouble verification mechanism 76
(Route R) to verify whether or not there is a real trouble.

The reason why a correlation between a certain trouble and another trouble or symptom is obtained in this way is to investigate the cause of a true trouble which cannot be predicted from the trouble appearing on the surface.

For example, it is assumed that jams frequently occur at the fixing device 21 portion (jam detection location J7) of the copying machine shown in FIG. In this case, the operation of the fixing device 21 seems to be malfunctioning on the surface, and therefore, in general, the parts in the vicinity of the fixing device 21 are exchanged or adjusted, but if these measures alone cannot solve the trouble. There is. This is because the true cause of trouble may be hidden and invisible. An example of the true cause of the hidden trouble will be described below.

Although the paper is fed from the paper feed tray 11 and reaches the fixing device 21 via a predetermined conveying path, there is a deviation from the standard paper passage timing in each part before the fixing device 21. For example, the paper feed tray 11a shown in FIG.
The paper passage timing delay is 20 m, the delays at the passage points J5 and J6 in the conveyance path are 1 ms and 2 ms, respectively, and the fixing device 21 has a delay of 10 ms. It is assumed that the delay time is 25 ms and the allowable delay time in the fixing device 21 is 30 ms. In this case, the jam is not detected in the portion before the fixing device 21 because it is within the allowable delay time, but the delay is accumulated as the sheet is conveyed. Then, when reaching the fixing device 21, the cumulative delay amount becomes 33 ms, which exceeds the allowable delay time 30 ms. Therefore, the paper sensor 35g detects a jam. However, considering the details of the delay, the paper feed tray 11a has the largest delay amount, and if the delay amount here is small, the fixing device 21 should not detect a jam.
That is, the true cause of the jam detected in the fixing device 21 portion is the paper feed tray 11a, and the true cause hidden in the conventional diagnostic method although the countermeasure for this portion should be taken originally. I couldn't know.

In such a case, since the delay amount in the paper feed tray 11a is large, not only the fixing device 21 portion (jam detection portion J7) but also the paper feed tray 11a is less frequent.
It is expected that a jam will also be detected at a (jam detection location J1). That is, as shown in FIG. 8, it can be expected that there is a positive correlation between the jam occurrence location J7 and the jam occurrence frequency at the jam detection location J1. Note that Jn indicates a jam detection point having no correlation.

Therefore, in this embodiment, for example,
When a trouble occurs in a certain part, the correlation between the trouble and other troubles occurring at the same time or the symptom at that time is obtained by regression analysis, and the one having a high correlation is set as a new virtual trouble.

In the above example, when a jam has occurred in the fixing device 21, the jam of the paper feed tray 11a has a high correlation, so the jam of the paper feed tray 11a is added as a new virtual trouble. To do.

In order to detect the correlation, for example, the simple regression analysis as shown below is used.

Now, n pairs of data (X ₁ , Y ₁ ) ,.
, (X _n , Y _n ) is obtained, the regression equation of Y with respect to X is expressed as follows.

[0114]

[Equation 1] Here, the correlation coefficient r is obtained by the following formula.

[0115]

[Equation 2] Then, if the correlation coefficient is, for example, 0.5 or more, it is considered as a candidate for a trouble having a correlation, and further, the candidate for this trouble determines a sign in the long-term observation information and obtains the correlation. It is calculated by adjusting it so that it falls within a certain range equal to or less than the certainty value of the original trouble. For example, if the trouble obtained from this correlation is actually occurring in the relevant device, the certainty value is set to the maximum, and if there is a symptom of trouble occurrence but no trouble occurrence, an intermediate value is assigned. , If neither, assign the lowest value.

As described above, if the trouble obtained from the correlation is confirmed as a new trouble by the trouble verification mechanism 76, the learning mechanism 75 determines this as a confirmed trouble T2.
It is passed to the trouble analysis mechanism 78 and the virtual trouble maintenance mechanism 77 in the same manner as the normal confirmed trouble T1 from (1) (route R).

In the virtual trouble maintaining mechanism 77, the built-in virtual trouble order determining mechanism 77-1 determines the priority order of a new virtual trouble, and updates the virtual trouble case base 74. The priority is determined, for example, according to the importance and frequency of the trouble. Basically, the higher the importance and the higher the frequency, the higher the priority. However,
The importance and the frequency are weighted. For example, even when the frequency is high, the priority is lowered when the importance is low. It should be noted that the term "importance" as used herein means that the customer is greatly affected when the trouble occurs. The virtual trouble maintenance mechanism 7
Step 7 is carried out every time the diagnosis is made in accordance with the frequency of occurrence troubles, regardless of the presence or absence of highly correlated troubles.

As described above, the parameters for determining the priority are the importance and the frequency. The importance is predetermined as an alert category according to the type of trouble.

The above-mentioned priority order is specifically determined as follows.

The importance alert category is roughly divided into a plurality of stages, for example, 6 stages. The frequency is statistically processed as a deviation value of occurrence of trouble. When deciding the order, priority is given to the six levels of hierarchy (trouble importance levels 1 to 6) where the importance levels are determined, and then the deviation values of items within the same importance level range are compared to make a final decision. The ranking is decided. It should be noted that each trouble is always displayed in the specific area of the case base in the order of priority in the form of a list.

The trouble analysis mechanism 78 causes the confirmed trouble T1 from the learning mechanism 75 or the confirmed trouble T2 from the trouble verification mechanism 76 by the cause investigation mechanism 78-1 based on the basic principle fact base 71. The coping method and the like are determined by the coping determination mechanism 78-2. Then, the alert output Q is generated according to the determined content. The cause investigating mechanism 78-1 and the countermeasure determining mechanism 78-2 are general ones, and are used to investigate the cause and determine the coping method based on the fact that is well known in the past. For example, if the image is not recorded on the paper, the causes are troubles such as failure of the charging device, failure of the exposure device, failure of the developing device, lack of toner, failure of the transfer device, etc. Instruct to inspect each device and replenish toner.

The alert output Q includes information such as a serial number of the copying machine to be diagnosed, a code indicating a model, and the like, as well as "signs" and "signs" indicating signs regarding the current state of the machine to be diagnosed. It includes diagnostic information such as "factor" indicating the inferred factor or part information that led to the cause, "treatment" which is the appropriate handling instruction information. The alert output Q is sent back from the knowledge base 7 shown in FIG. 1 to the host computer 4 via the shell 6.

This diagnostic information is sent to the host computer 4, accumulated as a file, and output by the printer as required. Alternatively, this diagnostic information is sent from the host computer 4 to the communication control device 2 via the communication line and is retrieved from the communication control device 2. From this information, the service engineer can check the "symptom", "factor", and
It is possible to know the "handling", and before the actual failure of the copier, the parts can be replaced and adjusted to prevent the trouble of the device which causes trouble to the customer.

[0124]

As described above, in the present invention, a universal symptom showing a common characteristic for a plurality of diagnosed devices is extracted from a plurality of operation state information indicating the operation states of a plurality of diagnosed devices. is doing. As a result, a trouble that may occur in the future can be predicted, and if trouble countermeasures are taken in advance based on this prediction, the occurrence of trouble can be prevented in advance and the customer will not be bothered.

In the present invention, the trouble to be diagnosed can be efficiently diagnosed by prioritizing the troubles according to the urgency of the trouble countermeasure.

Further, in the present invention, the correlation between the extracted trouble and other symptoms is detected, and the one having a high correlation is regarded as a candidate for a new trouble, and the trouble is also verified for this new trouble. .. As a result, even if a real trouble is hidden behind the apparent trouble, the hidden trouble can be found and dealt with.

[Brief description of drawings]

FIG. 1 is a schematic block diagram of a diagnostic device to which a diagnostic method including an inference mechanism of the present invention is applied.

FIG. 2 is an explanatory diagram schematically showing the basic structure of a copying machine to which the present invention is applied.

FIG. 3 is a block diagram showing a configuration for obtaining input information about a copying machine necessary for diagnosis.

FIG. 4 is an explanatory diagram schematically showing the configuration of a knowledge base for making a diagnosis.

FIG. 5 is a flow chart showing an outline of a diagnostic process when a copier is diagnosed.

FIG. 6 is a graph for explaining generation of virtual trouble.

FIG. 7 is a graph for explaining a correlation analysis.

FIG. 8 is a graph for explaining the trouble verification.

[Explanation of symbols]

1 copy machine, 2 copy machine body, 3 communication control device, 4
Communication line, 5 host computer, 6 expert system server, 7 knowledge base system, 8 shell,
11, 11a to 11 paper feed tray, 12 papers, 13a
˜13c Paper feeder, 14 Conveyor, 15 Slip roller, 16 Alignment gate, 17 Photoconductor, 18
Alignment roll, 19 transfer position, 20 conveyor belt, 21 fixing device, 22 fixing device outlet roll, 23 fixing device outlet gate, 24 reversing gate, 25 conveying roll, 26 exit roll, 27 reversing device, 28 decurler, 29 double-sided printing Gate, 30 double-sided printing roll,
31 intermediate tray, 32 reversing device, 33 paper feeding device, 3
4 transport devices, 35a to 35j paper sensor, 41 transport control CPU, 42 paper transport control device, 43 high voltage control device, 44 image forming control CPU, 45 density control device, 46 voltage detection circuit, 47 main CPU, 51 support roller, 52 charging device, 53 exposure device, 54 patch generation device, 55 surface potential sensor, 56, 57 developing device, 58 pre-transfer corotron, 59 density sensor, 6
0 transfer device, 61 peeling device, 62 cleaning device, 71 basic principle fact base, 72 universal symptom case base, 73 individual case base, 74 virtual trouble case base, 75 learning mechanism, 75-1 input information analysis mechanism, 75- 2 virtual trouble generation mechanism, 76 trouble verification mechanism, 77 virtual trouble maintenance mechanism, 77-1 virtual trouble ranking determination mechanism, 78 trouble analysis mechanism, 7
8-1 Cause investigation mechanism, 78-2 Coping decision mechanism

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁵ Identification code Internal reference number FI Technical display location G06F 9/44 330 C 9193-5B 15/20 N 7218-5L // B41J 11/42 J 9011- 2C G06F 11/22 360 E 9290-5B

Claims (2)

[Claims] 1. A diagnostic method for diagnosing a device to be diagnosed based on operating condition information indicating an operating condition of the device to be diagnosed, the operating condition information indicating an operating condition in each device to be diagnosed from a plurality of devices to be diagnosed. Collecting universal signs from the plurality of operating state information, and extracting a universal symptom exhibiting common characteristics with respect to the plurality of diagnosed devices, and determining a trouble that may occur in the diagnosed device based on the universal symptom. Then, the troubles are accumulated in the virtual trouble case base as virtual troubles, a plurality of virtual troubles in the virtual trouble case base are given a priority according to a trouble countermeasure urgency level, and the plurality of virtual troubles are assigned based on the priority order. A diagnostic method characterized by sequentially verifying virtual troubles. 2. A diagnostic device for diagnosing a device to be diagnosed based on operating state information consisting of a plurality of factors indicating an operating state of the device to be diagnosed, wherein the diagnostic device is provided for each of the plurality of devices to be diagnosed. Correspondingly provided, a plurality of communication control devices that obtain operating state information indicating the operating state of each of the devices to be diagnosed and transmit to the outside, and a plurality of operating state information from the plurality of communication control devices. A host computer that collects data, and a diagnostic unit that diagnoses the state of the device to be diagnosed based on the plurality of operating state information obtained from the host computer, the diagnostic unit analyzing the plurality of operating state information. And analysis means for obtaining a universal symptom showing common characteristics for the plurality of devices to be diagnosed, and a universal symptom case for accumulating the universal symptom obtained by the analysis means. A virtual trouble generating means for generating a virtual trouble that may occur in the device to be diagnosed based on data from the universal symptom case base, and a virtual trouble case base for accumulating the virtual trouble. A means for assigning a priority order to a plurality of virtual troubles in the virtual trouble case base according to a urgency level of trouble countermeasures, and a validity of the virtual troubles in the virtual trouble case base for sequentially verifying in accordance with the priority order. A diagnostic device comprising a trouble verification means.