JP2793422B2 - Self-diagnosable image forming device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image forming apparatus having a self-diagnosis system. More specifically, the present invention relates to an image forming apparatus capable of performing self-diagnosis of an operation state and the like of the apparatus using artificial intelligence and knowledge engineering, which have been actively studied in recent years.

[0002]

2. Description of the Related Art Recently, in the field of development of precision machines and industrial machines, in order to realize labor saving of maintenance work and prolonged automatic operation, experts using artificial intelligence (AI) technology have recently been used. Research on the system is being actively conducted. Some expert systems perform self-diagnosis as to whether or not a failure has occurred in the device and self-repair the resulting failure.

However, conventional expert systems (automatic adjustment systems and failure diagnosis systems) basically only actuate corresponding actuators based on the output of a certain sensor. It was not perfect. Therefore, the applicant of the present application has found a machine control method using diagnosis / repair inference on a target model based on qualitative physics, and a new self-diagnosis and self- Invented a restoration system and filed a patent application.
Flat No. 2-252191 (JP-A-4-130459) reference).

The self-diagnosis and self-repair system for the image forming apparatus according to the prior application has the following features. That is, (1) A detection value of a sensor provided in a target machine (image forming apparatus) is converted into a qualitative value and used for control. (2) The structure and characteristics of the image forming apparatus are qualitatively expressed using a causal relation network (parameter model) of parameters representing the properties of the image forming apparatus.

(3) A qualitative simulation for fault diagnosis and fault repair inference is performed by applying the sensor values converted into qualitative values to a parameter model. In other words, qualitative model-based systems (Qualitative Mo
Del Based System (QMS)) to perform fault diagnosis and repair.

According to the self-diagnosis and self-healing system of the applicant of the present application having such features, even if a failure such as a structural change of the image forming apparatus occurs, the image forming apparatus can flexibly deal with the failure. It is possible to correspond to. This is because the control points and control loops of the target machine can be dynamically changed by using the qualitative simulation.

[0007]

However, in the self-diagnosis and self-repair system according to the prior application,
When converting the detection value of the sensor into a qualitative value, there is a possibility that an error occurs in the conversion. This is because, when converting a sensor detection value into a qualitative value, a landmark is defined on the qualitative quantity space, and is converted into a different qualitative value depending on whether the detected value is larger or smaller than the boundary mark. As such, the landmarks must be correctly defined.

However, this landmark may change depending on the use environment of the image forming apparatus and the like, and the sensor detection value itself may not always be an accurate value due to the limit of the measurement accuracy of the sensor. . Therefore, in the conventional system, the qualitative value of the sensor detection value as the basis of the control varies, and as a result, an accurate qualitative simulation cannot be performed, and an error may occur in the failure diagnosis and the failure repair. .

When operating the self-diagnosis system according to the prior application as a practical machine embedded system,
It was not optimal in terms of system size and execution speed. therefore,
It is desired to reduce the size and speed of the self-diagnosis system. The present invention has been made in view of such a background, and an object of the present invention is to provide an image forming apparatus capable of achieving more miniaturization and higher speed and capable of performing accurate failure diagnosis.

[0010]

According to the first aspect of the present invention,
An image forming apparatus capable of performing self-diagnosis of a failure occurring in the apparatus, wherein a virtual case storage unit storing a plurality of virtual cases in which at least a failure and a qualitative state of a predetermined parameter are set; A plurality of sensors for detecting the state of a plurality of parts determined in advance, and a boundary mark of a qualitative quantity space which is provided corresponding to each sensor and which is necessary when converting the detection value of each sensor into a qualitative value is a fuzzy theory. At a predetermined timing, forcibly causing a failure in the device,
Using the detection values of the plurality of sensors read before causing the failure and the detection values of the plurality of sensors read when causing the failure, the membership function stored in the landmark storage unit is corrected. Landmark correction means,
When a failure symptom appears in the device, a conversion unit that reads the detection values of the plurality of sensors and converts the detection value of each sensor into a fuzzy qualitative value using a membership function stored in the landmark storage unit. Comparing the fuzzy qualitative value converted by the conversion means with a plurality of virtual cases stored in the virtual case storage means, and specifying a fuzzy qualitative value And selecting means for selecting a virtual case having a predetermined relationship.

According to a second aspect of the present invention, the image forming apparatus further comprises a fault repairing means operable to repair a fault set in the virtual case selected by the selecting means. Is what you do. According to a third aspect of the present invention, in the image forming apparatus, the timing at which the fault is forcibly caused to cause a fault in the apparatus in order for the landmark correcting means to correct the membership function is such that the fault repair is completed by the fault repairing means. It is characterized by the fact that

According to a fourth aspect of the present invention, in the image forming apparatus, the timing at which the failure is forcibly caused in the apparatus in order to correct the membership function by the landmark correcting means is performed by a manual operation or the like. Is input. According to a fifth aspect of the present invention, in the image forming apparatus, the membership functions stored in the landmark storage unit are stored for each failure symptom.

According to a sixth aspect of the present invention, in the image forming apparatus, the plurality of virtual cases stored in the virtual case storage means are stored by being classified according to failure symptoms. is there. The invention according to claim 7 is
In the image forming apparatus, the selection unit defines a qualitative quantity space, and compares a qualitative state of a predetermined parameter set in each virtual case with the fuzzy qualitative value based on a positional relationship in the qualitative quantity space. Then, a virtual case in which the parameter of the qualitative state closest to the position of the fuzzy qualitative value is set is selected.

According to an eighth aspect of the present invention, in the image forming apparatus, a repair method according to a failure is further set for each of the plurality of virtual cases. According to a ninth aspect of the present invention, in the image forming apparatus, the set restoration method includes a plurality of registration methods with priorities.

According to a tenth aspect of the present invention, in the image forming apparatus, the failure repair means operates to execute a repair method set in the virtual case selected by the selection means. Things. Claim 11
The described invention is a method for generating a plurality of virtual cases stored in the virtual case storage means, wherein the image forming apparatus is qualitatively represented by a causal relation network of parameters representing properties of the apparatus. Then, when it is assumed that a failure symptom has occurred in the device, all possible states of the device are qualitatively simulated using the causal network of the parameters, and the result is used to determine the failure and the qualitative state of the predetermined parameter. And a plurality of virtual cases including the virtual cases.

According to a twelfth aspect of the present invention, in the method of generating a virtual case, when it is assumed that a failure symptom has occurred in the device, the cause of the occurrence of the failure symptom is caused by a single parameter change. It is characterized by performing under the conditions. According to a thirteenth aspect of the present invention, in the method of generating a virtual case, when it is assumed that a failure symptom has occurred in the apparatus, a change in a certain parameter always affects another parameter. It is characterized by performing.

[0017]

According to the present invention, when the state of the image forming apparatus is self-diagnosed, the qualitative simulation does not have to be performed based on the sensor value every time the diagnosis is performed. In advance, all possible states as a result of such a qualitative simulation are stored as virtual cases. Therefore, basically, the failure of the machine can be specified in a short time only by comparing the sensor value converted into the fuzzy qualitative value with the virtual case.

In the present invention, the trouble symptoms that can be handled are limited to those caused by changes in the parameters of the target machine (image forming apparatus). Many failures of the image forming apparatus can be dealt with by changing parameters. If a failure other than that caused by a parameter change, such as a failure involving a structural change of a device or a failure caused by a limitation of an operation range of an actuator, can be dealt with, a qualitative analysis based on the already proposed QMS described above can be performed. What is necessary is just to add a failure diagnosis utilizing simulation. In addition, in order to appropriately deal with failure repair after structural changes of the target machine, etc., the already proposed Case Based Planning System (Case Based Planning System)
(CBS)). In addition, QM
Both the qualitative simulation by S and the CBS are described in, for example, Japanese Patent Application No. 2-252191 (Japanese Patent Application No.
Kaihei 4-130459) and / or Japanese Patent Application No. 3-2.
No. 51073 (Japanese Unexamined Patent Publication No. 5-88468) will be referred to here, and the description thereof will be omitted.

Specifically, according to the first aspect of the present invention,
When a failure symptom occurs in the device, the detection values of a plurality of sensors are read, and each detection value is converted into a fuzzy qualitative value. The membership function as a landmark required for converting the sensor detection value into a fuzzy qualitative value is modified by a simulated failure method that forcibly causes a failure in the device. Therefore, the membership function required for the conversion to the fuzzy qualitative value can be constantly corrected to the optimum value, and the sensor detection value can be accurately converted to the fuzzy qualitative value. The converted fuzzy qualitative value is compared with a plurality of virtual cases, and a virtual case having a predetermined relationship with the fuzzy qualitative value, for example, a closest relationship with the fuzzy qualitative value is selected.
Then, the failure set in the selected virtual case is presumed to be the cause of the current failure symptom.

According to the second aspect of the present invention, the estimated failure is automatically repaired. According to the third aspect of the present invention, the timing of correcting the membership function is every time the failure repair is completed. Therefore, even if the parameter of the device changes due to the repair of the failure and the boundary mark changes, the change does not occur. And the membership function can be modified.

According to the present invention, the membership function is corrected by the pseudo failure method every time a correction request signal is input manually by a service person or the like. Function can be modified. According to the fifth aspect of the present invention, since the membership function is stored for each failure symptom, the detection value of the sensor can be more accurately converted to a fuzzy qualitative value for each failure symptom.

According to the sixth aspect of the present invention, since the virtual cases are classified according to the failure symptoms, when a certain failure symptom appears, the virtual cases are classified from the virtual cases classified into the corresponding failure symptoms. The virtual case can be selected quickly. In the invention according to claim 7, when comparing the fuzzy qualitative value with the virtual case, a qualitative space is defined, and a virtual case is selected based on a positional relationship in the qualitative space. When the virtual case is selected based on the positional relationship in the qualitative quantity space, the selection of the virtual case does not vary, and the selection accuracy is improved.

According to the eighth, ninth or tenth aspect of the present invention, since a repair method corresponding to a failure is further set in the virtual case, when a virtual case is selected, the repair work of the failure is performed. Easy to do. According to the method of the present invention, the virtual cases can be grouped into a plurality of cases including a fault and a qualitative state of a predetermined parameter. can do.

In particular, according to the twelfth or thirteenth aspect, when a virtual case is generated, filtering is performed under a certain condition, so that the generated virtual case is only an appropriate case that can exist in the apparatus.

[0025]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a self-diagnosis and self-repair system applied to a small-sized electrophotographic copying machine will be described below with reference to the drawings. FIG. 1 is a mechanical configuration diagram of a small-sized electrophotographic copying machine to which the present invention is applied, in which only parts related to the present invention are schematically illustrated. FIG.
, 1 is a photosensitive drum, 2 is a main charger,
Reference numeral 3 denotes a halogen lamp for illuminating a document, 4 denotes a developing device, and 5 denotes a transfer / separation charger.

The main charger 2 is connected to a main charger controller 2C for changing the discharge voltage of the main charger. The halogen lamp 3 is connected to a halogen light amount controller 3C for controlling the light amount of the halogen lamp 3. further,
The transfer / separation charger 5 includes a transfer charger controller 5 for controlling a discharge voltage of the charger 5, that is, a transfer voltage between the photosensitive drum 1 and a copy sheet.
C is connected.

In an electrophotographic copying machine, the most important thing is whether or not the obtained copy image is beautifully finished (normal). Therefore, this embodiment automatically detects whether the obtained copy image is normal, has an image fog, or whether the image is faint. The cause of the symptoms,
That is, a description will be given of an example of a device that locates a failure and self-repairs the failure.

In this embodiment, for example, four sensors are provided. That is, a light amount sensor X for measuring the amount of light for exposing the photosensitive drum 1 (in other words, the light amount of the halogen lamp 3), a surface potential sensor Vs for measuring the surface potential of the photosensitive drum 1 after exposure, Photoconductor drum 1
Density sensor D for detecting the above toner density
s and the copy density sensor Os. The copy density sensor Os detects the density of a copy image formed by the electrophotographic copying machine. Based on the detection output Os of the copy density sensor Os, it is determined whether the electrophotographic copying machine is normal, whether image fogging has occurred as a failure symptom, or whether the image is faint.

FIG. 2 is a functional block diagram of the small-sized electrophotographic copying machine shown in FIG. 1, and shows only parts related to the present invention. In FIG. 2, blocks with rounded corners represent so-called hardware functions, and blocks with sharp corners represent so-called software functions (program processing executed in a computer). Note that the division of functions by hardware and functions by software is an example, and the functions of software may be realized by hardware.

The correspondence between the functional blocks in FIG. 2 and the machine configuration in FIG. 1 is as follows. That is, the sensor of FIG. 2 includes the light amount sensor X, the surface potential sensor Vs,
A toner density sensor Ds and a copy density sensor Os are included. The actuator controller of FIG. 2 includes the main charger controller 2C, the halogen light amount controller 3C, and the transfer charger controller 5C of FIG. The actuator of FIG.
The main charger 2, the halogen lamp 3, and the transfer / separation charger 5 of FIG. 1 are included.

In FIG. 2, a functional block by software is divided into, for example, four functional blocks.
That is, the diagnosis / repair inference unit 11 and the virtual case holding unit 1
2. The membership function generator 13 and the simulated fault generator 14. The virtual case holding unit 12 stores virtual cases illustrated in Tables 1 and 2 generated in advance. Here, the virtual case is a qualitative simulation of the state of the apparatus at the time of the failure under the condition that the failure occurring in the electrophotographic copying machine at a time is a single failure,
The state of the simulation result that can be present in the device.

[0032]

[Table 1]

[0033]

[Table 2] Generation of virtual case may be performed using a qualitative simulation described in applicant's earlier application described above (Japanese Patent Application No. flat No. 2-252191 (JP-A-4-130459)), and the like. The method of the qualitative simulation will be briefly described as follows. The electrophotographic copying machine, represented as binding of a plurality of elements captured from a physical view point, the qualitatively expressed using parameters binding relationship between behavior and attributes as well as the elements of each element, FIG. The parameter model shown in FIG. Note that the parameter model shown in FIG. 3 is a simplified model in which only parameters related to the copy density parameter Os are extracted.

In the parameter model shown in FIG.
l is the light amount parameter of the halogen lamp 3, D is the optical density parameter of the document, X is the parameter of the light amount for exposing the photosensitive drum 1, β is the sensitivity parameter of the photosensitive drum 1,
Vn is the surface potential parameter of the photosensitive drum 1 after the main charge, Vs is the surface potential parameter of the photosensitive drum 1 after exposure, Vb is the parameter of the developing bias, γ ₀ is the parameter of the toner sensitivity, and Ds is the parameter on the drum. An image density (toner density) parameter, Vt indicates a transfer voltage parameter, and ζ indicates a paper sensitivity parameter. Among these parameters, D, β, γ ₀ and ζ can be regarded as fixed values because the possibility of variation is small. Therefore, the cause of the change in the copy density parameter Os is H
It can be inferred that any one of 1, Vn, Vb, and Vt has changed. Then, when these four parameters Hl, Vn, Vb or Vt change and Os changes, the change always changes three sensing target parameters X, Vs or Ds (circled in FIG. 3) ( However, only when it is caused by a change in Vt,
X, Vs, and Ds do not change).

The qualitative simulation for generating a virtual case is based on the assumption that a single failure occurs in the electrophotographic copying machine at a time, as described above. Therefore, in the case of Hl (halogen lamp) failure, Vn (main charger) failure, Vb (development bias) failure, and Vt (transfer charger) failure, sensing target parameters X, Vs , Ds
Are different. Therefore, this state is inferred and stored as a virtual case.

Next, a specific example of how to generate a virtual case will be described using the parameter model of FIG. It is assumed that the copy density becomes abnormal and Os becomes high (+). If the cause of Os: high (+) is Hl: low (-), X becomes low (-). If the cause of Os: high (+) is caused by a change in Vn, Vb or Vt,
X is normal (N). This is because the generation of the virtual case is based on the premise that a failure occurring in the electrophotographic copying machine at a time is a single failure. Thus, for Os: high (+), X must be normal (N) or high (+), and low (-) is not possible.

On the other hand, the root cause of Os: high (+) is Hl
Then, Hl must be low (-) and changes in Hl should affect X, Vs and Ds on the parametric model. This is because if the change of Hl is such that it has no effect, Os does not change as a result. Thus, if Hl is the root cause of the failure symptom, ie, failure, then X, Vs, and Ds
Cannot be normal (N).

As described above, in the virtual case, when the copy density parameter Os indicates an abnormality, the cause is always limited to a change in a single parameter, and the parameter change is a parameter to be sensed (in FIG. 3, a circle). (Enclosed parameters). When a virtual case is generated under such a condition, only what can actually occur in the device can be obtained as a virtual case.

Tables 1 and 2 show specific examples of the virtual case. Table 1 shows four virtual cases when the failure symptom “image fogging” occurs in the electrophotographic copying machine. Table 1
Is obtained as follows. When an image fogged occurs in an image copied by an electrophotographic copying machine, the cause of the image is H based on the parameter model shown in FIG.
l (halogen lamp) failure, Vn (main charger)
A defect, a Vb (development bias) defect, or a Vt (transfer charger) defect can be estimated.

In this case, if the above-mentioned faults are limited to a single fault, and if the change in Hl always affects other parameters, the cause of the image fogging is H1 fault. In this case, the parameter X should be low (-), the parameter Vs should be high (+), the parameter Ds should be high (+), and no other state is taken.

If the cause of image fogging is Vn failure, the parameter X should be normal (N), the parameter Vs should be high (+), the parameter Ds should be high (+). I will not take it. When the cause of image fogging is a Vb defect, the parameter X is normal (N),
The parameter Vs should be normal (N), the parameter Ds should be high (+), and no other state is taken.

If the cause of the image fog is Vt failure, the parameter X should be normal (N), the parameter Vs should be normal (N), and the parameter Ds should be normal (N). I will not take it. The numerical value of “1.0” added to each parameter state in Table 1 indicates a degree in a membership function of fuzzy logic described later. The advantages of introducing the membership function of the fuzzy theory will be described later.

Similarly, when the density of the copy obtained by the electrophotographic copying machine is low, the causes are as follows: Hl (halogen lamp) failure, Vn (main charger) failure, Vb (development bias) according to the parameter model of FIG. ) A defect or a Vt (transfer charger) defect is estimated, and the parameters X, Vs and D to be sensed when each defect occurs.
The state of s is as shown in Table 2.

The virtual cases exemplified in Tables 1 and 2 are obtained in advance by a qualitative simulation for each failure symptom and stored in the virtual case holding unit 12. The virtual case holding unit 12 also stores, for each failure in each failure symptom,
The restoration method is inferred and stored in advance. For example,
For the failure "Hl failure" of the failure symptom "image fogging", a repair method as shown in Table 3 below is stored.

[0045]

[Table 3] The repair method is stored for each of the other faults. The feature of this embodiment is that such a virtual case is calculated in advance using a qualitative simulation, and is held. This eliminates the need to perform a qualitative simulation of the state of the electrophotographic copying machine every time a failure is diagnosed, thereby increasing the processing speed.

Next, the membership function generator 1 shown in FIG.
As shown in FIG. 4 and FIG. 5, a membership function used to qualify the detection values of the light quantity sensor X, the surface potential sensor Vs, and the toner density sensor Ds for each failure symptom is stored in 3. ing. As is well known, a membership function is a function that defines the degree (grade) of a certain element belonging to a certain set in fuzzy logic.

For example, FIG. 4 shows a failure symptom “image fog”.
The X, Vs, and Ds membership functions used at times are shown. If the diagnosis / repair inference unit 11 (see FIG. 2) determines that an image fogging has occurred in a copy output from the electrophotographic copying machine based on the output of the copy density sensor Os (see FIG. 1), Time light intensity sensor X,
The detection values of the surface potential sensor Vs and the toner density sensor Ds are stored in the membership function generation unit 13 in FIG.
Qualitative value based on the membership function shown in
For example, the detection value of the light amount sensor X is 2.2 as a quantitative value.
Below (V), the parameter X (-: 1.0, N: 0.
0). 1. The detected quantitative value of the light amount sensor X is 2. 29 (V), the parameter X (-: 0.7, N:
0.3). When the detected quantitative value of the light quantity sensor X is 2.5 (V) or more, the parameter X (−:
0.0, N: 1.0).

Similarly, the detected quantitative value of the surface potential sensor Vs and the detected quantitative value of the toner density sensor Ds are also qualitatively determined using the membership function of Vs and the membership function of Ds shown in FIG. . When it is determined that the image density is low, the quantitative detection values of the light amount sensor X, the surface potential sensor Vs, and the toner density sensor Ds are calculated using the membership functions of X, Vs, and Ds shown in FIG. Each is qualitatively converted.

Next, how to set the membership function shown in FIG. 4 or 5 will be described. In general, it is necessary to define landmarks (landmarks) on a quantity space in order to convert a detected quantitative value of a sensor into a qualitative value. However, in consideration of a change in the normal state of the electrophotographic copying machine after restoration and a limit of the measurement accuracy of the sensor, it is not easy to determine the landmark as a static one. If the landmarks are determined as static and there is an error in the determination, the qualitative conversion of the sensor values, which is the premise of this control, will not be performed accurately, and it will be used in subsequent fault diagnosis and fault repair. ,
The possibility of misdiagnosis or misrepair is increased.

Therefore, in this embodiment, as described above, a landmark is defined for each failure symptom, and the landmark is defined using a membership function of fuzzy logic. If the detected quantitative value of the sensor is converted into a qualitative value using a membership function corresponding to the failure symptom, a reading error of the sensor,
It is possible to flexibly and suitably cope with a change in the sensor output due to a change in the use environment.

In addition, when a quantitative detection value of a sensor is converted into a qualitative value, a membership function of fuzzy logic is introduced to allow a correspondence between an actual measured value and a qualitative value depending on a measurement accuracy of a sensor, a change in a use environment, and the like. It is possible to flexibly deal with the problem relating to the attachment, and it is possible to prevent errors from occurring when converting the sensor value into a qualitative value. At this stage, the qualitative parameter is not immediately used for selecting any fault by applying it to the virtual case. As described later, in order to select one of a plurality of faults included in the virtual case, a fault of the virtual case having the closest state to the parameter is obtained based on a predetermined calculation formula.

Further, in this embodiment, the pseudo failure method (Im
itation Fault method: IF method) was introduced. The IF method forcibly causes a failure in an electrophotographic copying machine by operating an actuator at an initial time before shipping the electrophotographic copying machine, after repairing a failure, or at an arbitrary timing based on manual input, This is a method of dynamically determining a landmark using sensor information at the time of normal operation and at the time of failure before causing a failure. The membership functions shown in FIGS. 4 and 5 are determined using this IF method. I
By using the F method, it is possible to dynamically determine the landmarks in the quantity space necessary to qualify the quantitative value detected by the sensor for each electrophotographic copying machine that is the actual control target. Can be accurately defined for each device.

Further, if the IF method is used, as will be described later, the landmarks defined when the device is in the initial state can be corrected each time the repair of the failure is completed. It is possible to always update the landmarks in the quantity space to optimal values in accordance with the above. Returning to FIG. 2, the functional block of this electrophotographic copying machine is provided with a simulated failure generator 14 for executing the above-mentioned IF method.

In this embodiment, as shown in FIGS. 4 and 5, the landmarks in the quantity space are defined using a membership function of fuzzy logic. That is, the landmarks are fuzzy. By fuzzifying the landmarks,
As described above, there is an advantage that it is possible to flexibly cope with a reading error of the sensor or a change in the landmark due to a disturbance such as an environmental change.

FIG. 6 shows the diagnosis / repair inference unit 11 shown in FIG.
Qualitative Reasoning (Fuzzy Qualitativ)
5 is a flowchart illustrating an algorithm of e Reasoning (FQR). Next, a failure diagnosis and a failure repair process in the electrophotographic copying machine will be described with reference to the flow of FIG. When the control operation starts, the diagnosis / repair inference unit 1
The detection value of the copy density sensor Os is read by 1 (step S1). Then, the read copy density Os is compared with a predetermined reference value, and it is determined whether or not the electrophotographic copying machine has failed (step S2).

For example, it is assumed that a condition shown in FIG. 7 is stored as a reference value. That is, the detection voltage is 2.5
If the voltage is less than (V), the image is thin. If the detection voltage is 2.5 (V) or more and less than 2.9 (V), the image is normal.
(V) Above, it is assumed that a failure determination reference value of image fog is set. At this time, the copy density sensor O
If the detected value of s is 3.1 (V), it is determined that the failure symptom "image fogging" has occurred (step S3).

The processes in steps S1 to S3 are performed because the electrophotographic copying machine according to the present embodiment is a machine for automatically determining the presence or absence of a failure. This process is performed manually. Is also good. Step S manually
When performing the processes of 1 to S3, the copy density sensor Os need not be provided. In the manual process, the serviceman or the like may determine that the copy has an image fog, for example, by looking at the copy output from the electrophotographic copying machine.
In this case, an image fog is input to the apparatus as a failure symptom. The input of the failure symptom may be made by a ten key or the like usually provided in the electrophotographic copying machine.

When the failure symptom "image fog" is determined in step S3, the detection values of the light quantity sensor X, the surface potential sensor Vs, and the toner density sensor Ds are read (step S4). Now, the detected values of the respective sensors read are X: 2. 26 (V), Vs: 2. 6 (V), D
s: Suppose that it was 1.9 5 (V). Each read sensor value is applied to the membership function at the time of image fogging (FIG. 4) stored in the membership function generation unit 13, and a temporary qualitative value is determined (step S5). In this specific example, X: 2. 26 , Vs: 2. 6 , Ds: 1.9
5 are respectively applied to the membership function of FIG. 4, and X: -0.8, Vs: +0.9, Ds: +0.
7 is obtained.

[0059] That is, (X, Vs, Ds) = (2. 26, 2. 6, 1.9 5) = p (-0.8, + 0.9, + 0.7) are obtained. If the qualitative conversion of the sensor detection value is performed based on a specific landmark instead of fuzzy qualitative conversion using a membership function, (X, Vs, Ds) = (−, +, +) Is obtained.

Next, the failure symptom "image fog" shown in Table 1 is shown.
The degree of coincidence C between the failures listed in the virtual case and the temporary qualitative value calculated in step S5 is calculated (step S5).
6). The calculation of the degree of coincidence C is performed as follows. First, the failures listed in the virtual case of the failure symptom “image fogging” are expressed in a three-dimensional quantity space of X, Vs, and Ds. This expression can be expressed by the following equation.

Hl defect: (X, Vs, Ds) = f1 (−1.0, +1.0, +1.0) Vn defect: (X, Vs, Ds) = (N1.0, +1.0, +1) .0) = f2 (−0.0, +1.0, +1.0) Vb defect: (X, Vs, Ds) = (N1.0, N1.0, +1.0) = f3 (−0.0 , +0.0, +1.0) Vt defect: (X, Vs, Ds) = (N1.0, N1.0, N1.0) = f4 (−0.0, +0.0, +0.0) When the equation is plotted, the three-dimensional quantity space shown in FIG. 8 is obtained. In FIG. 8, f1, f2, f3, and f4 are the positions of the Hl defect, the Vn defect, the Vb defect, and the Vt defect, respectively.

The temporary qualitative value p (−0.8, +0.9, +0.7) obtained in step S5 is located at p in the three-dimensional quantity space of FIG. Therefore, next, from the point p, the positions f1, f2, f of the faults listed in the virtual case
Calculating the distance D to 3 and f4 is as follows. D (f1) = √ {( 0.8-1.0) 2 + (0.9 -1.0) 2 + (0.7 -1.0) 2} = 0.374 D (f2) = √ {(0.8-0.0) 2 + (0.9 -1.0) 2 ^{+ (0.7 -1.0) 2} =} 0.86 D (f3) = √ {(0.8-0.0) 2 + (0.9 -0.0) 2 + (0.7 -1.0) 2} = 1.241 D (f4) = √ {(0.8- 0.0) ² + (0.9−0.0) ² + (0.7−0.0) ² ｝ = 1.393 Then, the distance D calculated by the above equation is normalized, and the coincidence C is calculated. The normalization of the distance D is performed based on the following equation.

C = 1−D / √n (where n is the number of sense parameters: in this case, n =
3) Accordingly, each degree of coincidence C is expressed as follows: C (f1) = 1−0.374 / √3 = 0.784 C (f2) = 1−0.86 / √3 = 0.503 C (f3) = 1 −1.241 / √3 = 0.284 C (f4) = 1−1.393 / √3 = 0.196 As a result, f1, which is the closest distance D from the point p,
That is, f1 (Hl failure) having the highest coincidence C is determined as a failure candidate (step S7).

The equation for calculating the degree of coincidence C can be expressed by the following general equation. C = 1− {C (p1) ² + C (p2) ² +... + C (pn) ² } / Δn C (pn) = Gm (qn) −Gs (qn) (where C: coincidence of the entire model) Degree, pn: measurable variable, C (pn): degree of coincidence with variable pn, qn: qualitative value that variable pn can take, Gm (qn): grade of qualitative value qn in failure model, Gs (qn): measurement (Quality of qualitative value qn in value) When performing normal qualitative qualification based on a specific landmark instead of fuzzy qualitative qualification, the degree of coincidence C in step S6
Is omitted, and from (X, Vs, Ds) = (−, +, +), the failure is immediately determined to be an H1 failure.

In step S7, since the failure is determined to be the H1 failure, the repair method corresponding to the failure “H1 failure” of the failure symptom “image fog” stored in the virtual case holding unit 12 (the method shown in Table 3) ) Is executed according to the priority. In order to perform the restoration method in the priority order, the counter x is cleared in step S8, and the counter x is set to x = 1 in step S9. Next, when it is confirmed that the value of the counter x does not exceed the registered number of the stored repair methods (step S10), the priority of the value of the counter x among the stored repair methods (for example, When the first restoration is performed, Hl: UP of priority No. 1
(Increase the amount of halogen lamp light)) is performed (step S11).

Then, it is determined whether or not the restoration has been successful (step S12). Whether or not the restoration was successful is determined by copying after the restoration and reading the density of the copy output as a result by the copy density sensor Os. If the repair was not successful, step S
9, the count value of the counter x is incremented by one, and the next priority is restored. For example, priority N
o. Vn: DOWN (reducing the main charger voltage), which is the repair of No. 2, is performed. If the restoration of the next priority is not registered, the process ends at that point.

If it is determined in step S12 that the restoration has been successful, the process proceeds to step S13, where the IF method is executed, and the process ends. FIG. 9 shows the processing content of the IF method performed in step S13 described above. Next, the IF method will be described in detail with reference to FIG. When the failure repair is successful, the diagnostic / repair inference unit 11 reads the detection values of the light quantity sensor X, the surface potential sensor Vs, and the toner density sensor Ds (Step S21). The detection values of each sensor read at this time are, for example, X: 2.9 (V), Vs:
It is assumed that 1.6 (V) and Ds: 1.4 (V).

Next, the simulated fault generator 14 (see FIG. 2)
, The halogen light amount controller 3C is operated, and the light amount of the halogen lamp 3 is reduced (step S2).
2). Each time the amount of light of the halogen lamp 3 is reduced by a small amount, the copying machine is caused to perform a copying operation, and the density of the copy obtained at that time is detected by the copy density sensor Os, and the detected value is read (step S23). ).
The detection value of the copy density sensor Os is illuminated by the above-described failure determination reference value shown in FIG. 7, and when the value of Os reaches the reference value at which image fogging occurs, the process of lowering the light amount of the halogen lamp 3 is stopped. Is performed (step S24).

Then, the detection values of the light quantity sensor X, the surface potential sensor Vs, and the toner density sensor Ds when the light quantity of the halogen lamp 3 is reduced until image fog occurs are read (step S25). The read detection values are, for example, X: 2.6 (V), Vs: 2.5 (V), and Ds:
Let it be 1.8 (V). The detected values of X, Vs, and Ds read at step S21 after the repair and the detected values of X, Vs, and Ds at the time when the image fog is read at step S25 are sent to the membership function generating unit 13. Given, a membership function for image fogging is generated. That is, the value detected in step S21 is a landmark when normal, and the value read in step S25 is a landmark when image fogging starts, and the membership function at the time of image fogging shown in FIG. Is restored to the membership function shown in

Next, the pseudo failure generator 14 operates the halogen light quantity controller 3C to increase the light quantity of the halogen lamp 3 (step S26). And every time the light amount of the halogen lamp 3 is increased by a minute amount,
The copying is performed by the electrophotographic copying machine, and the copy density obtained at that time is detected by the copy density sensor Os, and the value is read (step S27).

Then, the read value of the copy density sensor Os is illuminated with the reference value for judging the presence or absence of a failure shown in FIG. 7 and when the image reaches a light copy (step S28), the light quantity sensor X and the surface potential sensor Vs at that time. Then, the detection value of the toner density sensor Ds is read (step S29). This read value is, for example, X: 3.5 (V), Vs: 0.6
(V) and Ds: 0.5 (V).

The read value is sent to the membership function generator 13. In the membership function generating unit 13, the sensor value in the normal state read in step S21,
Using the sensor value read at the time of generation of the low-density image read in step S29 as a landmark, a membership function at the time of density reduction is generated. As a result, the membership function at the time of density reduction shown in FIG. 5 is restored to that shown in FIG.

The above-described IF method may be performed, for example, in response to a manual input of an IF method execution signal by a service person or the like, instead of being performed after the failure has been successfully repaired. The present invention is not limited to the contents of the above-described embodiments, and various modifications can be made based on the scope of the claims. For example, in the above embodiment, a small electrophotographic copying machine has been described as an example. However, the self-diagnosis and self-repair system according to the present invention is also applied to other image forming apparatuses such as a laser beam printer and a facsimile. be able to.

Further, the present embodiment is directed to an electrophotographic copying machine in which a failure symptom appears when an obtained copy image is not beautifully finished, and a failure which causes the occurrence of the failure symptom is self-repaired. It is all about explaining. However, the present invention can also be applied to self-diagnosis and self-repair for other failures of the image forming apparatus different from whether or not the copy image is beautifully finished.

Various other changes are possible.

[0076]

According to the present invention, when a detection value of a sensor provided in an image forming apparatus is converted into a qualitative value, the conversion is affected by the use environment of the apparatus, detection accuracy of the sensor, and the like. It is possible to accurately estimate a failure causing a failure symptom in the image forming apparatus using the converted fuzzy qualitative value.

Further, according to the present invention, it is possible to provide an image forming apparatus having a self-diagnosis and self-repair system which has been reduced in size and speed, thereby providing a highly practical apparatus. it can. Furthermore, according to the present invention, a membership function as a landmark that forms the basis for converting a sensor detection value into a qualitative value can be accurately defined for each device. In addition, since the membership function can be updated at a predetermined timing, the qualitative conversion of the sensor detection value can always be performed accurately, and a device free from erroneous diagnosis and erroneous repair can be provided.

[Brief description of the drawings]

FIG. 1 is a mechanical configuration diagram of a small electrophotographic copying machine to which the present invention is applied.

FIG. 2 is a functional block diagram of the small electrophotographic copying machine shown in FIG.

FIG. 3 is a simplified parameter model of the small electrophotographic copying machine shown in FIG.

FIG. 4 is a diagram showing membership functions of X, Vs, and Ds used at the time of image fogging.

FIG. 5 is a diagram illustrating a membership function of X, Vs, and Ds used when the image density decreases.

FIG. 6 is a flowchart illustrating an algorithm of fuzzy qualitative inference.

FIG. 7 is a diagram illustrating an example of a failure presence / absence determination reference value.

FIG. 8 is a diagram expressing the failures listed in the virtual case of the failure symptom “image fogging” in a three-dimensional quantity space of X, Vs, and Ds.

FIG. 9 is a flowchart showing processing contents of a pseudo failure method (IF method).

FIG. 10 is a diagram showing a membership function at the time of image fogging corrected by the IF method.

FIG. 11 is a diagram showing a membership function at the time of image density reduction corrected by the IF method.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 photoconductor drum 2 main charger 3 halogen lamp 4 developing device 5 transfer / separation charger 2C main charger controller 3C halogen light amount controller 5C transfer charger controller 11 diagnostic / repair inference unit 12 virtual case holding unit 13 membership function generation unit 14 pseudo failure Generator X light intensity sensor Vs surface potential sensor Ds toner density sensor Os copy density sensor

──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor ▲ Yoshi ▼ Hiroyuki Kawa 804 Yonbancho, Yobancho, Chiyoda-ku, Tokyo 804 (56) References JP-A-58-66967 (JP, A) JP-A-58 -94012 (JP, A) JP-A-62-235916 (JP, A) JP-A-2-82271 (JP, A) JP-A-3-27058 (JP, A) JP-A-2-311860 (JP, A) JP-A-3-7963 (JP, A) JP-A-3-10269 (JP, A) JP-A-4-74224 (JP, A) JP-A-4-258080 (JP, A) JP-A-5-Japanese 173821 (JP, A) JP-A-58-221856 (JP, A) JP-A-63-233655 (JP, A) JP-A-62-52601 (JP, A) JP-A-1-219697 (JP, A) JP-A-1-169611 (JP, A) JP-A-62-2328 (JP, A) JP-A-1-278865 (JP, A) JP-A-1-291918 (JP, A) JP-A-2-235074 (JP, A) JP-A-2-302828 (JP, A) JP-A-2-113262 (JP, A) JP-A-63-70268 (JP, A) Kaihei 4-130330 (JP, A) (58) Field surveyed (Int. Cl. ⁶ , DB name) G03G 21/00 370-540

Claims (13)

(57) [Claims] An image forming apparatus capable of self-diagnosing a failure occurring in an apparatus, wherein a virtual case storage storing a plurality of virtual cases in which at least a failure and a qualitative state of a predetermined parameter are set is stored. Means, a plurality of sensors for detecting the state of a plurality of predetermined parts of the apparatus, and a qualitative quantity space which is provided corresponding to each sensor and which is necessary when converting the detection value of each sensor into a qualitative value. A landmark storage means in which the landmark is stored as a membership function of fuzzy theory, at a predetermined timing, forcibly causing a failure in the device, and detecting values of the plurality of sensors read before causing the failure; Using the detection values of the plurality of sensors read when causing the landmark, a landmark correction means for correcting the membership function stored in the landmark storage means, Conversion means for reading the detection values of the plurality of sensors when a failure symptom appears in the device, and converting the detection values of each sensor into fuzzy qualitative values using a membership function stored in the landmark storage means; Comparing the fuzzy qualitative value converted by the converting means with a plurality of virtual cases stored in the virtual case storing means, and identifying a fault causing an onset failure symptom; And selecting means for selecting a virtual case having a predetermined relationship. 2. The image forming apparatus according to claim 1, further comprising: a failure repair unit that operates to repair a failure set in the virtual case selected by the selection unit. It is. 3. The image forming apparatus according to claim 2, wherein the timing at which the landmark correcting means forcibly causes the device to malfunction in order to correct the membership function includes:
It is characterized in that each time the failure repair is completed by the failure repair means. 4. The image forming apparatus according to claim 2, wherein the timing at which the boundary marker correcting means forcibly causes the device to malfunction in order to correct the membership function is:
This is characterized in that the correction request signal is inputted by manual operation or the like. 5. The image forming apparatus according to claim 1, wherein the membership functions stored in the landmark storage unit are stored for each failure symptom. 6. The image forming apparatus according to claim 1, wherein the plurality of virtual cases stored in the virtual case storage means are stored by being classified according to failure symptoms. 7. The image forming apparatus according to claim 1, wherein said selecting means defines a qualitative quantity space, and a qualitative state of a predetermined parameter set in each virtual case, and A fuzzy qualitative value is compared with a positional relationship in a qualitative quantity space, and a virtual case in which a parameter of a qualitative state closest to the position of the fuzzy qualitative value is set is selected. 8. The image forming apparatus according to claim 2, wherein a repair method corresponding to a failure is further set for each of the plurality of virtual cases. 9. The image forming apparatus according to claim 8, wherein the set restoration method includes a plurality of registration methods with priorities. 10. The image forming apparatus according to claim 8, wherein said failure repair means operates to execute a repair method set in the virtual case selected by the selection means. is there. 11. A method for generating a plurality of virtual cases stored in virtual case storage means according to claim 1, wherein said image forming apparatus is controlled by a causal network of parameters representing properties of the apparatus. Expressed qualitatively, when it is assumed that a failure symptom has occurred in the device, all possible states of the device are qualitatively simulated by using a causal network of the parameters, and the result is represented by a failure and a predetermined parameter. A method of generating a virtual case, wherein the virtual case includes a plurality of virtual cases including a qualitative state. 12. The virtual case generation method according to claim 11, wherein when it is assumed that a failure symptom has occurred in the apparatus, the cause of the occurrence of the failure symptom is limited to a single parameter change. And is carried out in 13. The method of generating a virtual case according to claim 12, wherein when it is assumed that a failure symptom has occurred in the device, a change in a certain parameter always affects another parameter. It is characterized by performing in.