FIG. 4 is a table illustrating some of control parameters and operation settings thereof.

The control parameters are defined for the following sheet processing, respectively. The sheet processing includes sheet conveyance processing such as sheet feed, conveyance, and ejection of the sheet on which the image has been formed. The sheet processing includes image forming processing for forming the image on the sheet, such as conditions for transferring the color material to the sheet and conditions for fixing the color material on the sheet. The sheet processing includes post-processing to be performed on the ejected sheet on which the image has been formed.

As a control parameter of the air sheet feeding, there are a PFU tray tip fan air volume and a PFU tray side fan air volume in a paper feed unit (PFU) for sending out a sheet. These represent the air blowing amounts at the leading end and the side surface of the tray, respectively. By adjusting an air volume according to the mass (density) of the sheets such as thick paper and thin paper, it is possible to appropriately separate and deliver one sheet from a bundle of sheets. Furthermore, the sheet feed control parameters include a PFU registration loop amount. The registration loop amount is an amount by which the sheet is fed for a predetermined time in a state where the leading end of the sheet is in contact with a nip line of the registration rollers. Thus, a loop (warp) is formed on the sheet surface, and the leading edge position of the sheet is aligned. Therefore, the sheet is sent out from the tray in accordance with the setting of the control parameter of the conveyance timing.

Thereafter, at least a part of the plurality of rollers is rotated by a motor, and thus the sheet is moved at the set conveyance speed while being nipped by the plurality of rollers. A nip pressure of the roller can be controlled according to a sheet type such as a thickness and a material of the sheet. These are included in the control parameters related to conveyance.

As a control parameter of the transfer condition, there is a transfer current value for transferring the toner from the intermediate transfer belt 32 to the sheet. As for the transfer current, a transfer trailing edge correction current value and a transfer leading edge correction current value are determined in accordance with different transfer states between the leading edge and the trailing edge of the sheet.

Furthermore, at the time of transfer, a voltage is applied in order to receive the charged toner on the sheet. A separation voltage is applied to the sheet after the transfer so that the sheet after the toner transfer is appropriately separated from the roller and the intermediate transfer belt 32. The AC component and the DC component are the separated AC applied voltage and the separated DC applied voltage, respectively.

Control parameters of the fixing conditions include a heating temperature related to heating and pressing, that is, a set temperature of the fixing rollers (upper and lower). Furthermore, a rotation speed of the fixing roller, an air volume of a fixing separation fan for separating the sheet from the fixing roller, and the like are also set as control parameters of the fixing conditions.

The static electricity is removed from the sheet before being ejected from the image forming apparatus 3 or in the post-processing apparatus 4. Thus, adhesion of the ejected sheet to the tray or another sheet due to static electricity is suppressed. The current value related to the static electricity removal is also a control parameter to be set. In addition, the control parameters related to the post-processing may include setting conditions for the sorting, the cutting, the stapling, the folding, and the like described above.

These control parameters are optimally set from a combination (characteristic information) of a plurality of sheet physical properties. The sheet physical properties are detection results by the various sensors of the sheet characteristic detection device 2 and properties uniquely specified from the detection results. In the image forming system U of the present embodiment, the controller 301 of the image forming apparatus 3 acquires the detection result from the sheet characteristic detection device 2, and sets the control parameter based on the detection result.

FIG. 5 is a diagram schematically illustrating a flow of setting of control parameters.

The image forming apparatus 3 acquires the detection result from the sheet characteristic detection device 2 (U1).

Among the plurality of physical property measurement values included in the detection result, some or all of the values necessary for determination of the control parameters are passed to the control parameter calculation processing for each category (U2). The control parameter calculation processing includes, for example, conveyance and sheet feed control parameter determination processing, transfer control parameter determination processing, fixing control parameter determination processing, and post-processing control parameter determination processing.

When each of the control parameters is determined, the determined control parameters are updated and stored as the setting of each drive process for each category (U3).

In the image forming system U, these control parameters can be set in two kinds of setting patterns which are different methods from each other.

In the first setting pattern, the value of the control parameter is determined in real time based on the physical property measurement value of the sheet measured by the sheet characteristic detection device 2 immediately before the image formation on the sheet. In this case, even when the physical properties of the sheet change during continuous image formation on a plurality of sheets, the control parameters are set following this change in real time. Thus, the image is appropriately formed.

The machine learning model 392 may be used to set the value of the control parameter in the first setting pattern. The machine learning model 392 outputs the value (setting) of the control parameter in response to an input of the necessary physical property measurement value (characteristic information). The machine learning model 392 has been trained in advance using learning data with teacher data. The learning may be performed outside the image forming system U. Data of the learned machine learning model is acquired from the outside and stored in the image forming apparatus 3 in advance. The machine learning model 392 may be different for each control parameter of the output target. The machine learning model 392 may be updated at appropriate timing. In this case, since the physical property measurement values are directly converted into the values of the control parameters, the type of sheet (sheet type) is not specified from the physical property measurement values.

For the machine learning model 392, for example, any of various algorithms for ensemble learning may be used. Values of the control parameters are selected such that an optimal result is obtained for a combination of a plurality of physical property measurement values. The machine learning model 392 may be trained and used independently for each control parameter. Alternatively, there may be a case where a plurality of control parameters change with correlation therebetween. For such a plurality of control parameters, a common machine learning model may be trained and used.

Alternatively, the setting of the control parameters in the first setting pattern may be performed, for example, by a statistical method without using a machine learning model. Alternatively, the optimal solution of the control parameters may be obtained more analytically using a polynomial regression equation, but in this case, the arithmetic processing tends to be complicated. Therefore, the polynomial regression equation is likely to be used in a case where the number of control parameters obtained by the first setting pattern is small.

In the second setting pattern, the value of the control parameter is initially determined according to the sheet type or a combination of physical property reference values corresponding to the sheet type. In the storage section 309, a list of type names (first type information) for classifying sheet types, combinations of physical property reference values (characteristic information and second type information) corresponding to the sheet types, and control parameters associated with these is held in advance as a sheet type table 393 (sheet type information). For example, the type of sheet having a combination of physical property reference values that is closest to the obtained combination of physical property measurement values is identified. Whether or not the physical property values are the closest to each other may be determined by a distance between vectors when the physical property values are appropriately normalized and converted into multidimensional vectors, for example, a Euclidean distance or the like. Before calculating a specific distance, sheet types to be candidates may be narrowed down by similarity of some control parameters. On the other hand, in a case where there is a significantly different physical property between the physical property reference value and the physical property measurement value, it may be determined to be a different sheet type without considering the distance. The setting content of the control parameter corresponding to this type is determined as the control parameter related to the image formation. In this case, in a standard state of the sheet, the sheet is appropriately conveyed and an image is formed thereon. In these processes, the sheet type itself may not be considered. The sheet type information including the control parameters is registered in advance in a sheet type table 393 for each sheet type. In addition, a new sheet type may be sequentially added and registered in the sheet type information. The additional registration may be manually performed by the user. Alternatively, the additional registration may be performed by automatically acquiring additional content periodically or each time there is an additional contact from a database server or the like that aggregates various sheet type information.

The setting operation by the first setting pattern by the controller 301 corresponds to a first control parameter determination means of the present embodiment. The setting operation according to the second setting pattern by the controller 301 corresponds to a second control parameter determination means of the present embodiment.

Some of the physical property measurement values may change during formation of a plurality of images. For example, the moisture percentage may greatly decrease or increase due to the effect of absorbing heat generated during an image forming operation or surrounding water vapor. Therefore, images may not be appropriately formed on all the sheets with the second setting pattern.

On the other hand, it is difficult to perform all the physical property measurements and all the control parameter settings based on the physical property measurements in real time every time before the image forming operation on each sheet. It is not so preferable to reduce the feeding speed (time interval) of each sheet for the above-described measurement and setting because productivity decreases. In the image forming system U, some of the control parameters are adjusted according to the first setting pattern for each sheet in real time based on the necessary measurement results and are used for each sheet processing. The control parameters other than the some of the control parameters are set according to the second setting pattern before the start of the image forming operation and are continuously used thereafter. That is, during the operation related to image formation, the plurality of control parameters include both of the parameters set by the first setting pattern and the parameters set by the second setting pattern. The controller 301 can set which of the first setting pattern and the second setting pattern is used to determine the value of each control parameter as a selection means of the present embodiment.

In the setting example 1 and the setting example 2 of FIG. 4, the control parameter described as “1” is set in real time by the first setting pattern. The control parameter marked as “2” is initially set once according to the second setting pattern. In this way, for each of the plurality of control parameters, which of the first control parameter determination means and the second control parameter determination means is used to determine the value is set.

In setting example 1, the transfer current value in transfer and the upper fixing control temperature and the lower fixing control temperature in fixing are set according to the first setting pattern. These are important control parameters in image forming processing, in particular, for image quality and stable conveyance. These are set in real time, and thus stable image formation in which image quality is emphasized is likely to be continued while occurrence of sheet jam or the like is reduced. Conversely, the control parameters related to sheet feed (sheet conveyance processing) and post-processing are set in the second setting pattern.

In the setting example 2, the control parameters set in the first setting pattern with respect to the fixing and sheet static elimination are included. All of the control parameters in the sheet feed and the transfer are set with the second setting pattern. In the image forming system U, since the fixing and the sheet static elimination are performed after the transfer, a grace time until the control parameter is set after the measurement is longer than that of the transfer. Therefore, since the settings in the first setting pattern are limited to the control parameters for the fixing and thereafter, the possibility that the image forming speed is adversely affected is reduced. In addition, the fixing process is more likely to affect a difference in final appearance, for example, gloss than the transfer. Therefore, the control parameters related to fixing are preferentially set in the first setting pattern, which tends to reduce the deviation from the originally expected image quality.

The image forming system U may be switchable between the setting example 1 and the setting example 2. The switching may be performed in response to a user's input operation or the like.

FIG. 6 is a flowchart illustrating a control procedure of setting control processing executed by the controller 301 in the image forming apparatus 3 of the present embodiment. The setting control processing is executed at the start of the image forming operation.

The controller 301 causes the sheet feed device 1 to feed the first sheet from the tray that supplies sheets on which images are to be formed (S1). The controller 301 causes the sheet characteristic detection device 2 to measure the first sheet. The sheet is sent to the second conveyance path 25S, and measurement by the second sensing section 24 is also performed. The controller 301 acquires the measurement results of the sheet physical properties from the sheet characteristic detection device 2 (S2).

The controller 301 determines the sheet type based on the measurement result (S3). The controller 301 refers to the sheet type table 393 to identify a sheet type associated with a combination of physical property reference values closest to the measurement result. Further, the controller 301 calculates the degree of coincidence between the measurement result and the physical property reference value of the specified sheet type. The degree of coincidence may be calculated by any calculation method as long as the degree of coincidence can be quantitatively evaluated.

The controller 301 sets, according to the identified sheet type and matching degree, the value of a control parameter for which the setting of the second control parameter is defined (S4). The processing S4 is included in the second control parameter determination means of the present embodiment. The processing S4 also corresponds to a function as a second control parameter determination means of the controller 301. The controller 301 ejects the first sheet from the second conveyance path 25S to the purge tray 251. Note that the processing S7 may be performed at appropriate timing in parallel with the processing S3 and the S4.

The controller 301 determines whether the set number of sheets for image formation has been fed from the sheet feed device 1 (S11). Note that the number of fed sheets in this case does not include the first sheet described above. If it is determined that the set number of sheets have been fed from the sheet feed device 1 (S11; YES), the controller 301 ends the setting control processing.

If it is determined that the set number of sheets has not been fed from the sheet feed device 1 (S11; NO), the controller 301 causes the sheet feed device 1 to feed the next sheet (S12). The controller 301 causes the sheet to be conveyed from the first conveyance path 25M to the image forming apparatus 3 in the sheet characteristic detection device 2. The controller 301 determines whether or not there is a setting of the first control parameter for real-time control (S13). If it is determined that there is no setting of the first control parameter (S13; NO), the processing by the controller 301 returns to processing S11.

If it is determined that the real-time control is to be performed (S13; YES), the controller 301 acquires the measurement result of the sheet physical property by the first sensing section 23 (S14). The controller 301 inputs the content selected from the measurement result for each first control parameter to be acquired to the corresponding machine learning model 392. The controller 301 acquires the value of the first control parameter output from each machine learning model 392 (S15). The processing S15 constitutes a first control parameter determination step of the present embodiment. The processing S15 also corresponds to a function as a first control parameter determination means of the controller 301. The controller 301 updates the setting of the control parameters with the acquired value of the first control parameter (S18). For the control parameter for which the value of the first control parameter has not been acquired, the value of the second control parameter acquired in the processing S4 is maintained as it is. Thereafter, the processing of the controller 301 returns to the processing S11.

The processes S13 to S14 constitute a selection step of the present embodiment. The processes S13 to S14 also correspond to a function as a selection means of the controller 301.

Modification Example 1

Modification example 1 of the setting control processing will be described.

In the above description, the first control parameters and the second control parameters are fixedly determined at the beginning, but may be changed during the image forming operation.

Examples of control parameters that can be changed between the first setting pattern and the second setting pattern are illustrated in the rightmost column of FIG. 4. That is, the transfer current value can be changed between the setting by the first setting pattern and the setting by the second setting pattern depending on the situation. In this way, whether or not the setting pattern can be changed may be determined depending on whether or not it depends on the measured value of the above-described physical property that is likely to greatly change, or the like. Alternatively, without the settings as described above, parameters set in the first setting pattern may be uniformly changeable to the settings in the second setting pattern.

In addition, in a case where it takes more time for arithmetic processing than originally assumed time, the setting of the first control parameters might not be done in time for actual processing contrary to assumption. The image forming system U determines a setting upper limit time (reference time) of the first control parameter in advance. For a control parameter for which the value of the first control parameter has not been determined within the set upper limit time, the value of the second control parameter may be set instead. Once the setting is changed to the value of the second control parameter, this value may be continuously used.

FIG. 7 is a flowchart illustrating a control procedure of a setting control process according to a first modification example.

In the setting control processing, processing S5, S6, S16, S17, and S21 are added to the setting control processing of the above-described embodiment. In addition, the processes S4 and S15 of the setting control process of the above-described embodiment are changed to a process S4a and S15a, respectively. In addition, the position of the processing S11 is changed. Other processes are the same between the two setting control processes. The same processing contents are denoted by the same reference numerals, and detailed description thereof will be omitted.

After the processing S3, the controller 301 acquires all the second control parameters according to the sheet type and the matching degree (S4a). That is, the controller 301 acquires the second control parameters in advance for the control parameters for which the first control parameters are to be acquired.

The controller 301 determines whether or not there is a control parameter for which the first control parameter is selected (S5). If it is determined that there is no control parameter for which the first control parameter has been selected (S5; NO), the controller 301 updates the setting of the control parameter with the acquired second control parameter (S21). The processing of the controller 301 proceeds to processing S11.

If it is determined that there is a control parameter for which the first control parameter has been selected (S5; YES), the controller 301 inputs necessary measurement results to the machine learning model 392. The controller 301 acquires the value of the selected first control parameter from the machine learning model 392 (S6). Next, the processing of the controller 301 proceeds to processing S7.

If it is determined not to acquire the first control parameter and perform the real-time control in the processing S13 (S13; NO), the processing by the controller 301 proceeds to processing S11. When the real-time control is performed (S13; YES), the process of the controller 301 proceeds to a process S14. After the processing S14, the controller 301 starts the processing of calculating the value of the selected first control parameter (S15a).

The controller 301 determines whether a specified time has elapsed before the value of the selected first control parameter is calculated (S16). If it is determined that the specified time has elapsed before the value of the first control parameter is calculated (S16; YES), the controller 301 replaces the uncalculated control parameter with the calculated second control parameter (S17). At this time, the controller 301 changes the selection of the uncalculated first control parameter to the selection of the second control parameter. When all the selections are changed to the second control parameter, the real-time control is stopped. The processing of the controller 301 proceeds to processing S18. If it is determined that the value of the first control parameter has been calculated before the specified time elapsed (S16; NO), the processing by the controller 301 proceeds to processing S18.

In the processing S18, the controller 301 updates the control parameters with the newly obtained first control parameters or the replaced second control parameters (S18). Next, the processing of the controller 301 proceeds to processing S11.

Modification Example 2

Modification example 2 of the setting control processing will be described.

In the above description, the selection of the first control parameter may be changed to the selection of the second control parameter in accordance with the accuracy of the machine learning model 392.

In a case where the physical property measurement value is largely deviated from a normal range, it may be difficult to set an appropriate control parameter by the change in real time. In the case where the machine learning model 392 is used for the first setting pattern as described above, the amount of learning data outside of the normal range is generally small in many cases. That is, the machine learning model 392 is not learned with high accuracy outside the normal range. The image forming apparatus 3 defines in advance an allowable range of a physical property measurement value in which the machine learning model 392 can be used with necessary accuracy. It is sufficient that this range be defined for each of a plurality of control parameters. When input data outside the defined range is acquired, the controller 301 stops the setting of the control parameters in the first setting pattern. In this case, the controller 301 uses the control parameters obtained in the second setting pattern. The control parameter obtained in the second setting pattern may be acquired in advance.

FIG. 8 is a flowchart illustrating a control procedure of the setting control processing of modification example 2.

The setting control processing includes processing S16 instead of processing S20 in the setting control processing of modification example 1. Furthermore, the setting control processing includes the processing S15a of the present embodiment described above instead of the processing S15 in modification example 1. The process S20 is executed before the processes S15 and S16. The other processing is the same between modification examples 1 and 2. The same processing contents will be denoted by the same reference numerals and the description thereof will be omitted.

After the processing S14, the controller 301 determines whether or not the acquired measurement result is within a specified range (S20). When the measurement result is not within the defined range (S20; NO), the processing by the controller 301 proceeds to processing S17. When the measurement result is within the defined range (S20; YES), the processing by the controller 301 proceeds to processing S15. Note that in a case where whether or not a plurality of control parameters are within the defined ranges is divided, one of processing S15 and processing S17 may be individually performed for each control parameter. After the processing S15 or S17, the processing of the controller 301 proceeds to the processing S18.

As described above, the image forming system U of the present embodiment includes the image forming apparatus 3, and the image forming apparatus 3 includes the controller 301. The controller 301 as the first control parameter determination means determines the value of the control parameter for sheet processing. The controller 301, as the second control parameter determination means, determines the value of the control parameter for sheet processing by a method different from that of the first control parameter determination means. The controller 301, as the selection means, can set which of the first control parameter determination means and the second control parameter determination means is used to determine the value of at least one of the plurality of control parameters.

There are many control parameters related to image formation. When an attempt is made to uniformly optimize a combination of these on the basis of data on a large number of sheet physical properties, the amount of calculation may become large. As a result, it is difficult for the image forming system to reflect the optimized control parameters in real time during the high-speed image forming operation. Therefore, there is a desire to enable more appropriate operation settings related to image formation while suppressing a decrease in productivity. In contrast, the image forming system U can use two different methods of determining control parameters in combination and select and set which of the methods is to be used. Therefore, the control parameters are more preferably determined in accordance with the importance of each control parameter, the degree of variation of the physical property measurement values for determining the control parameters, and the like. Thus, the image forming system U can perform operation setting related to image formation more appropriately without reducing productivity. Therefore, the image forming system U can output a stable image with higher accuracy.

Furthermore, as a first control parameter determination means, the controller 301 can determine, based on the characteristic information corresponding to the sheet physical properties, the values of the control parameters for the sheet processing without identifying the sheet type. Thus, the controller 301 can determine the control parameters by appropriately reflecting the variation within the sheet type, in particular, the sheet physical property which tends to change with time, such as the moisture percentage. Therefore, the image forming system U can suppress the occurrence of sheet conveyance trouble during image formation. Further, the image forming system U can appropriately maintain the image quality of the formed image.

Furthermore, the controller 301 as the second control parameter determination means may determine the value of the control parameter for sheet processing according to sheet type information corresponding to the sheet type. Conversely to the above, the control parameters that are generally fixed according to the sheet type may be determined to be values according to the sheet type, a combination of the physical property measurement values associated with the sheet type, or the like. Accordingly, the controller 301 can reduce the load and the required time related to the setting of the control parameter. On the other hand, troubles related to image formation are easily suppressed.

In addition, the sheet type information may include a sheet type specified based on characteristic information obtained by combining physical property measurement values and the like. As long as a combination of physical property measurement values by which a sheet type is uniquely determined is defined, name information on the sheet type itself is not necessarily required for setting of control parameters. However, since the sheet type itself can be specified, the user can easily determine the suitability when the user confirms the setting contents.

The sheet type information may also include information associating the identified sheet type with the value of the control parameter. The value of an appropriate control parameter is directly associated with a sheet type, so that the image forming system U can easily and promptly obtain a control parameter corresponding to the sheet type. In principle, the physical properties of sheet are not continuous, but change discontinuously among different sheet types. Therefore, the image forming system U can appropriately set standard control parameters according to the sheet type by being associated with the control parameters in units of the sheet type.

The sheet type information may also include information associating the characteristic information with the value of the control parameter. As described above, as long as the combination of physical property measurement values by which the sheet type is uniquely determined is defined, name information on the sheet type itself is not necessarily required for setting of control parameters. The image forming system U can directly and easily obtain appropriate control parameters from the combination of physical property measurement values.

Furthermore, the sheet processing may include at least one of image formation processing for forming the image on the sheet, sheet conveyance processing for feeding or conveying the sheet, and post-processing to be performed on the sheet after image formation. Appropriate control parameters are set for these processes, so that the image forming system U can effectively reduce the occurrence of troubles related to sheets. Further, the image forming system U can form the image of appropriate image quality on the sheet.

In particular, the sheet processing may include image forming processing for forming the image on the sheet. The controller 301 determines, as the first control parameter determination means, the control parameter for the image formation processing. Control parameters related to image formation processing greatly affect image quality. At the same time, these control parameters are important for appropriately separating the sheet from the roller or the like while causing the color material to adhere to the sheet. Therefore, the image forming system U can more reliably form the image with optimum control parameters. In addition, the image forming system U can more reliably discharge the sheet on which the image is formed.

Furthermore, the control parameters for the image formation processing may include at least one of control parameters related to fixing conditions of the color material to the sheet and control parameters related to transfer conditions of the color material to the sheet. Transfer and fixing of the color material to the sheet is particularly important in image formation. When these are appropriately defined, the image forming system U can stably form the image with higher image quality.

Furthermore, the sheet processing may include at least one of sheet conveyance processing for feeding or conveying the sheet and post-processing to be performed on the sheet after image formation. The controller 301 may, as second control parameter determination means, determine the control parameter for the at least one process. Thus, the number of control parameters set according to the first setting pattern can be reduced. Therefore, the image forming system U can reduce the time and load required for setting the control parameters while maintaining appropriate image formation. Thus, the image forming system U can assign more effective control parameter settings to the first setting pattern. Further, the image forming system U can form a high-quality image without reducing the image forming speed, that is, the productivity.

Furthermore, the control parameters determined by the controller 301 as the second control parameter determination means may include at least one of the conveyance speed of the sheet, the conveyance timing, the loop amount, and the nip pressure. For these control parameters, accuracy of values generally corresponding to the sheet type information is sufficient. Therefore, since these are defined by the second setting pattern, the burden of the setting is reduced. Furthermore, since these control parameters are used from the start of the conveyance of the sheet, they need to be already determined in a situation where the physical property measurement values are acquired by the sheet characteristic detection device 2. That is, the control parameter that cannot be set in time by the first setting pattern may be preferentially allocated to the setting by the second setting pattern.

Furthermore, the control parameters determined by the controller 301 as the second control parameter determination means may include at least one of conditions related to stapling, folding, and cutting. The post-processing is performed after the image is formed on the sheet. Therefore, these parameters are less likely to require detailed setting than various control parameters related to the image forming operation. These are preferentially allocated to the second setting pattern, so that the image forming system U can set more effective control parameters by the first setting pattern.

The characteristic information may include information corresponding to at least one of basis weight, thickness, moisture percentage, surface properties, electrical resistance, and size. These are important physical properties of the sheet, in particular, for conveyance of the sheet and application and fixing of the color material. Therefore, the image forming system U can determine optimal control parameters more accurately by determining control parameters based on characteristic information that includes these as much as possible.

The controller 301 may include, as the first control parameter determination means, the machine learning model 392 that outputs the control parameter in response to input of the characteristic information. In order to identify an empirical optimum solution from a combination of many physical property measurement values, it is useful to use the machine learning model. As a result, the image forming system U can set the control parameter based on a large number of physical property measurement values so that an appropriate image can be formed without deteriorating the image quality.

Alternatively, the controller 301 may output, as the first control parameter determination means, the control parameter corresponding to the characteristic information without using the machine learning model. If it is possible to determine appropriate control parameters from a combination of many physical property measurement values by a statistical method or the like, the controller