JP2021093664A - Image processing system, image processing apparatus, program, control management system, and apparatus - Google Patents

Abstract

To make it possible to change a control parameter in conformity with a standard.SOLUTION: An image processing system 100 comprises an image processing apparatus 200 and a server 300. The image processing apparatus 200 includes: an image forming unit; a machine learning execution unit that executes machine learning related to a control parameter of the image forming unit; a temporary determination unit that temporarily determines, based on a result of the machine learning, a new control parameter as a control parameter after change (post-change parameter); and a communication unit that transmits the post-change parameter to the server before updating the post-change parameter. The server 300 determines whether or not the post-change parameter has passed a standard test and transmits a result to the image processing apparatus. The image processing apparatus 200 includes an update unit that updates the post-change parameter in accordance with the test result.SELECTED DRAWING: Figure 1

Description

The present invention relates to image processing systems, image processing devices, programs, control management systems and devices.

There are regulations (standards) related to ensuring safety and environmental protection for electrical products and electronic devices used in homes and offices, manufacturing equipment installed in factories, and medical equipment. One example is the regulation of electromagnetic interference. A voltage including a high-voltage DC component and a high-frequency square wave AC component is applied to a developing device provided in an electrophotographic image processing device such as a multifunction device. Since the AC component is a square wave, it contains many high-frequency components, and because the voltage is high, electromagnetic interference waves are likely to be generated. For this reason, the manufacturer of the image processing device ships after testing and confirming that the regulation of the electromagnetic interference wave is cleared (the radio wave intensity of the electromagnetic interference wave is less than the regulation limit value). ..

On the other hand, image processing devices are required to have high image quality, high durability, and support for many paper types, and adjustment of voltage and frequency is required even after shipment. Further, adjustments are required to cope with the environment (temperature, humidity, etc.) in which the image processing device is installed. In addition, high durability means that even if the image processing device changes over time, the usage period is extended by adjusting control parameters such as voltage and frequency. Such adjustment of control parameters after shipment is required not only for image processing devices but also for various devices and devices including manufacturing devices and medical devices.

In general, manufacturers or distributors of equipment and devices adjust control parameters as part of maintenance services. In addition, with the progress of machine learning technology, a technology has emerged in which devices and devices adjust control parameters according to their states. For example, in the invention described in Patent Document 1, in a machine tool, the tool correction interval is optimized by using reinforcement learning from the viewpoint of the amount of machining error and the machine operating rate. Further, in the invention described in Patent Document 2, in the control device of the robot, the operation parameters of the robot are optimized by using reinforcement learning from the viewpoints of operation time, deviation from the target position, vibration and the like.

Japanese Patent No. 5996676 Japanese Unexamined Patent Publication No. 2018-126796

In the inventions described in Patent Documents 1 and 2, control parameters are optimized from the viewpoints of machine operation accuracy, operation time, efficiency, and the like. On the other hand, devices and equipment are required to comply with national and industry regulations and standards. For example, regarding regulations established by the Voluntary Control Council for Interference by Information Technology Equipment (VCCI) (abbreviated as VCCI (Voluntary Control Council for Interference by Information Technology Equipment)), EU's EMC (ElectroMagnetic Compatibility) directive, and safety of electrical products. It is necessary to comply with EU low voltage directives.

When the control parameter is changed, the value subject to regulation (for example, the radio field intensity of the electromagnetic interference wave emitted by the device in the EMC directive) changes and may exceed the limit value. However, the above invention has no description regarding regulation and is not considered.

The present invention has been made in view of such a background, and an image processing system, an image processing device, a program, and a control capable of changing within a range conforming to a standard of control parameters calculated by machine learning. The subject is to provide a management system and equipment.

(1) An image processing system including an image processing device and a server, wherein the image processing device includes an image forming unit, a machine learning executing unit that executes machine learning related to a predetermined control parameter of the image forming unit, and an image processing unit. A tentative determination unit that tentatively determines a new control parameter as a changed control parameter based on the result of the machine learning, and the change before updating the changed control parameter as a control parameter to be executed by the image processing apparatus. The server includes a communication unit that transmits the later control parameters to the server, and the server includes a pass / fail result transmission unit that determines the pass / fail result of the standard test in the changed control parameters and transmits the pass / fail result to the image processing device. The image processing system further includes an update unit that updates the changed control parameter as a control parameter executed by the image processing device according to the pass / fail result.

(2) An image processing device of the same type as the image processing device is connected to the server, and the pass / fail result transmission unit changes the control parameter of the image processing device of the same type to the changed control parameter. The image processing system according to (1), wherein the standard test is performed using the same type of image processing apparatus.

(3) The control parameter that has passed the standard test is stored in the storage unit of the server, and the pass / fail result transmission unit is described when the changed control parameter is included in the passed control parameter. The image processing system according to (1), characterized in that it passes a standard test and transmits it.

(4) A control parameter group that affects the pass / fail of the standard test and a control parameter group that does not affect the standard test are stored in the storage unit of the image processing device, and the update unit stores the pass / fail of the standard test. The control parameters included in the control parameter group that affect the standard test are changed after receiving the pass from the server by the communication unit, and the control parameters included in the control parameter group that does not affect the pass / fail of the standard test are described above. The image processing system according to (1), wherein the image processing system is changed regardless of the pass / fail result.

(5) The storage unit of the image processing device stores a range of control parameters that do not affect the standard test, and the update unit stores control parameters that are included outside the range of control parameters that do not affect the standard test. The control parameters are changed after receiving the pass from the server by the communication unit, and the control parameters included in the range of the control parameters that do not affect the standard test are changed regardless of the pass / fail result. The image processing system according to (1).

(6) In the storage unit of the server, a pass range, which is a range of control parameters that have passed the standard test, and a fail range, which is a range of control parameters that have failed the standard test, are stored. The pass / fail result transmission unit transmits that the changed control parameter has passed the standard test when it is included in the pass range, and the changed control parameter is included in the fail range. If the changed control parameter is not included in the pass range or the fail range, the changed control parameter corresponds to the changed control parameter in the pass range. The control parameter value closest to the control parameter to be used is transmitted, or the control parameter of the image processing device of the same type is changed so as to correspond to the changed control parameter, and the pass / fail result of the standard test is determined and transmitted. The image processing system according to (2), wherein the updating unit changes the control parameter value when the communication unit receives the control parameter value.

(7) The communication unit of the image processing device transmits either one or both of the environment information and the usage history information to the server together with the changed control parameters, and the environment information is sent to the storage unit of the server. The control parameters when the standard test is passed are stored in association with the usage history information, and the pass / fail result transmission unit uses the changed control parameters as the environmental information and usage history information transmitted by the image processing apparatus. The image processing according to (1), wherein when the control parameter when the standard test is passed, which is associated with any one or both of the above, is transmitted, the result is transmitted as passing the standard test. system.

(8) The storage unit of the server stores the control parameters and firmware that have passed the standard test for each type of the image processing device, and the pass / fail result transmission unit has passed the standard test for the type. The image processing system according to (1), wherein the firmware of the type is changed so that the control parameter value becomes a specified value of the firmware of the type.

(9) The storage unit of the server stores the control parameters and the firmware when the standard test is passed for each model of the image processing device, each usage history, and each environment, and the pass / fail result transmission unit stores the firmware. , The firmware is changed so that the control parameter values that have passed the standard test in the type, the usage history, and the environment are the specified values of the firmware of the type, the usage history, and the environment. The image processing system according to (1), which is a feature.

(10) The image processing system according to (1), wherein the control parameter is a physical parameter that controls the image processing apparatus.

(11) The image processing system according to (1), wherein the standard of the standard test is a standard corresponding to at least one of a legal regulation and a safety regulation.

(12) The image processing system according to (1), wherein the machine learning is reinforcement learning accompanied by a predetermined reward determination.

(13) After changing the image processing unit, the machine learning execution unit that executes machine learning related to a predetermined control parameter of the image forming unit, and a new control parameter based on the result of the machine learning. A tentative determination unit that tentatively determines as a control parameter of the above, a communication unit that transmits the changed control parameter to a server before updating the changed control parameter as a control parameter to be executed by the image processing device, and the above. An image characterized by comprising an update unit that updates the changed control parameter as a control parameter executed by the image processing device according to a pass / fail result of a standard test in the changed control parameter received from a server. Processing equipment.

(14) The image processing apparatus according to (13), wherein the control parameter is a physical parameter that controls the image processing apparatus.

(15) The image processing apparatus according to (13), wherein the standard of the standard test is a standard corresponding to at least one of a legal regulation and a safety regulation.

(16) A program for being executed by an image processing device that is a computer, based on a step of forming an image, a step of executing machine learning related to control parameters related to the formation of the image, and a result of the machine learning. A step of tentatively determining a new control parameter as a changed control parameter and transmitting the changed control parameter to a server before updating the changed control parameter as a control parameter to be executed by the image processing apparatus. To execute the step and the step of updating the changed control parameter as a control parameter to be executed by the image processing device according to the pass / fail result of the standard test in the changed control parameter received from the server. program.

(17) The program according to (16), wherein the control parameter is a physical parameter that controls the image processing apparatus.

(18) The program according to (16), wherein the standard of the standard test is a standard corresponding to at least one of a legal regulation and a safety regulation.

(19) A control management system that manages changes in the control parameters of a device having control parameters, a machine learning execution means for executing machine learning related to the control parameters, and new control parameters based on the result of the machine learning. As a tentative determination means for tentatively determining the changed control parameter, a pass / fail determination means for determining the pass / fail result of the standard test in the changed control parameter, and the changed control parameter according to the change result. A control management system equipped with an update means for updating as a control parameter to be executed in.

(20) A machine learning execution unit that executes machine learning related to control parameters, a tentative determination unit that tentatively determines a new control parameter as a changed control parameter based on the result of the machine learning, and a changed control parameter. Before updating as a control parameter to be executed by the own device, the change is made according to the pass / fail result of the standard test in the communication unit that transmits the changed control parameter to the server and the changed control parameter received from the server. A device including an update unit that updates later control parameters as control parameters to be executed by the own device.

According to the present invention, it is possible to provide an image processing system, an image processing device, a program, a control management system and a device capable of changing a control parameter calculated by machine learning within a range conforming to a standard.

It is an overall block diagram of the image processing system which concerns on this embodiment. It is an internal block diagram of the image processing apparatus which concerns on this embodiment. It is a functional block diagram of the image processing apparatus which concerns on this embodiment. It is a figure which illustrates the voltage applied to the developer of the image processing apparatus which concerns on this embodiment. It is a figure for demonstrating an example of the standard limit value of an electromagnetic interference wave. It is a functional block diagram of the server which concerns on this embodiment. It is a data structure diagram of the test image processing apparatus database stored in the server which concerns on this embodiment. It is a sequence diagram (1) of the control parameter change processing of the image processing system which concerns on this embodiment. It is a sequence diagram (2) of the control parameter change processing of the image processing system which concerns on this embodiment. It is a flowchart (1) of the printing process of the image processing apparatus (user machine) which concerns on this embodiment. It is a flowchart (2) of the printing process of the image processing apparatus (user machine) which concerns on this embodiment. It is a functional block diagram of the server which concerns on the modification 1 of this embodiment. It is a data structure diagram of the pass value parameter database stored in the server which concerns on the modification 1 of this embodiment. It is a sequence diagram (1) of the control parameter change processing of the image processing system which concerns on the modification 1 of this embodiment. It is a sequence diagram (2) of the control parameter change processing of the image processing system which concerns on the modification 1 of this embodiment. It is a data structure diagram of the parameter type database stored in the image processing apparatus which concerns on modification 2 of this embodiment. It is a data block diagram of the test exemption parameter database stored in the image processing apparatus which concerns on modification 3 of this embodiment. It is a data structure diagram of the pass / fail history parameter database stored in the server which concerns on the modification 4 of this embodiment. It is a data structure diagram of the pass history parameter database stored in the server which concerns on the modification 5 of this embodiment. It is a functional block diagram of the server which concerns on the modification 6 of this embodiment. It is a data structure diagram of the average pass value parameter database stored in the server which concerns on modification 6 of this embodiment. It is a data structure diagram of the average pass value parameter database stored in the server which concerns on the modification 7 of this embodiment.

Hereinafter, the image processing system according to the embodiment (embodiment) for carrying out the present invention will be described. The image processing system includes an image processing device (user machine) used by the user, a server, and an image processing device (test target machine) installed in the test room. The user machine prints a test chart, reads the print result, evaluates the image under the current control parameters, and determines new control parameters. The user machine sends new control parameters to the server. The server sets new control parameters on the test target machine, tests whether the test target machine meets the electromagnetic interference regulation, and notifies the user machine of the pass / fail. The user machine sets a new control parameter if it passes, and keeps the current setting if it fails.

Reinforcement learning is one of the methods for the user machine to determine a new control parameter value. By setting the set value of the control parameter as a state, changing the set value as an action, and determining the reward based on the print result of the test chart, the change (optimization) of the control parameter to improve the image quality by reinforcement learning can be performed. It will be possible. Machine learning techniques other than reinforcement learning may be used.
By executing a test (standard test) to see if the control parameter electromagnetic interference wave regulation is met on the test target machine, it is possible to guarantee that the user machine also meets the regulation, and the user machine within the range that meets the regulation. It is possible to improve the image quality for each. It is possible to optimize control parameters in consideration of variations in the parts of the image processing device (user machine), the environment in which the user machine is installed, and changes over time.
In the following embodiment, the electromagnetic interference wave will be described as an example of the regulation (standard), but other regulations and standards may be used.

≪Overall configuration of image processing system≫
FIG. 1 is an overall configuration diagram of the image processing system 100 according to the present embodiment. The image processing system 100 includes an image processing device 200 (user machine), a server 300, and an image processing device 600 (test target machine). The image processing device 200 is an image processing device used by the user, and can communicate with the server 300 via the network 800.
The image processing device 600 (test target machine) is installed in the test room 500 (anechoic chamber), and the intensity of the electromagnetic interference wave emitted by the image processing device 600 is measured by the standard tester 550. The measurement result of the standard tester 550 is transmitted to the server 300.

In FIG. 1, one image processing device 600 and one standard tester 550 are installed in one test room 500, but the present invention is not limited to this, and a plurality of image processing devices 600 and a plurality of standard testers 550 are installed. May be done. Further, although a plurality of test rooms 500 are connected to one server 300, one test room 500 may be used.

<< Configuration of image processing device (user machine) >>
FIG. 2 is an internal configuration diagram of the image processing device 200 according to the present embodiment. The image processing device 200 includes a paper feed tray 281,282,283, a transport path 270, an image forming unit 260, a scanner 291 and a paper ejection tray 292. Paper (recording medium) is set in the paper feed trays 281, 228, 283. The paper is conveyed on the transport path 270, printed (image is formed) by the image forming unit 260, and discharged to the paper ejection tray 292. A scanner 291 is provided on the transport path 270 between the image forming unit 260 and the output tray 292, and reads an image on the printed paper.

The image forming unit 260 includes a laser 264, a photoconductor 261 and a developing device 263, a band electrode 265, a primary transfer roller 262, a primary transfer belt 268, a secondary transfer roller 267, and a fixing device 266. The laser 264, the photoconductor 261 and the developer 263, the band electrode 265, and the primary transfer roller 262 are provided with four each corresponding to four colors of YMCK.

The image forming unit 260 forms an image on the paper conveyed along the conveying path 270 by the electrophotographic method. Specifically, the photoconductor 261 is charged by the band electrode 265, a latent image is formed on the photoconductor 261 by the laser 264, and the toner is applied to the photoconductor 261 by the developer 263. The toner on the photoconductor 261 is transferred to the primary transfer belt 268 by the primary transfer roller 262, and the toners of four colors are overlapped on the primary transfer belt 268 to form a toner image. The toner image on the primary transfer belt 268 is transferred to the paper on the transport path 270 by the secondary transfer roller 267.

FIG. 3 is a functional block diagram of the image processing device 200 according to the present embodiment. The image processing device 200 includes a general control CPU (Central Processing Unit) 211, a storage unit 212, a printer control unit 220, a network interface 219 (described as "net I / F" in FIG. 3, also referred to as a communication unit), a laser 264, and the like. And a scanner 291.

The overall control CPU 211 functions as a control unit that controls the entire image processing device 200 by executing the program 214 stored in the storage unit 212. Further, the overall control CPU 211 controls ON / OFF of the laser 264 to form a latent image on the photoconductor 261. Another control is changing (adjusting) control parameters related to image quality. The control parameter change processing of the image processing system 100 including the image processing device 200 will be described later with reference to FIGS. 8A and 8B.

The storage unit 212 stores the machine learning model 213, which will be described later, in addition to the program 214. The network interface 219 (communication unit) transmits and receives communication data to and from the server 300, which is a component of the image processing system 100. In addition, the network interface 219 transmits and receives communication data with other devices such as a personal computer, and receives print job data. The scanner 291 reads the image formed on the paper on the transport path 270 and outputs it to the overall control CPU 211.

The printer control unit 220 includes a printer control CPU 221, a ROM (Read Only Memory) 222, a RAM (Random Access Memory) 224, an I / O unit 225, and a D / A unit 226. The printer control CPU 221 operates according to the firmware 223 stored in the ROM 222 to control the image forming unit 260, and temporarily stores the temporary data required for the control in the RAM 224. The printer control CPU 221 controls the transfer roller 271 and the fuser 266 provided in the transfer path 270 via the I / O unit 225. Further, the printer control CPU 221 controls the laser 264, the band electrode 265, the developing device 263, the primary transfer belt 268, and the secondary transfer roller 267 via the D / A unit 226.

≪Relationship between development voltage and image≫
FIG. 4 is a diagram illustrating a voltage applied to the developing device 263 of the image processing device 200 according to the present embodiment. By applying a voltage to the toner of the developing device 263, the toner moves to the latent image on the photoconductor 261. The applied voltage includes a DC component and an AC component.

The voltage (development voltage) V1 of the DC component in FIG. 4 is 800V. The AC component is a square wave, the amplitude V2 is 200 V, and the frequency f is 3 KHz. Before time t1 and after time t2, the engine is idling and no voltage is applied. The printing state is between time t1 and time t2, and the voltage shown in the figure is applied to the developing device 263.
By adjusting the voltage V1 of the DC component, the maximum density of the image changes. Further, adjusting the amplitude V2 of the AC component changes the intermediate density, and adjusting the frequency f changes the image noise and unevenness. By adjusting these voltages and frequencies, the image quality can be adjusted and improved. However, since the rectangular wave contains a high frequency component and has a high voltage, an electromagnetic interference wave is likely to be generated, and it is necessary to adjust the rectangular wave within the range of regulations such as VCSI.

FIG. 5 is a diagram for explaining an example of a standard limit value of an electromagnetic interference wave. In FIG. 5, PK (peak), QP (quasi-peak), and AV (average) indicate the peak value, the quasi-peak value, and the average value of the electromagnetic interference wave, respectively, and the unit is dBμV / m. .. The bottom row in FIG. 5 shows that the limit value of the electromagnetic interference wave of 3 GHz to 6 GHz at the measurement distance of 3 m is 54 dBμV / m on average and 74 dBμV / m on the peak value. ..

<< Configuration of image processing device (test target machine) >>
The configuration of the image processing device 600 (test target machine) is the same as that of the image processing device 200 (user machine) except for the storage unit 212 (see FIG. 3). The image processing device 600 does not include the machine learning model 213. Further, the image processing device 600 receives a control parameter from the server 300 and executes printing. The standard tester 550 measures the electromagnetic interference wave generated by the image processing device 600 at the time of printing, and transmits the measurement result to the server 300. The control parameter change processing of the image processing system 100 including the image processing device 600 will be described later with reference to FIGS. 8A and 8B.

≪Server configuration≫
FIG. 6 is a functional block diagram of the server 300 according to the present embodiment. The server 300 is a computer and includes a control unit 310, a storage unit 320, and a communication unit 340. The server 300 receives the change value of the control parameter from the image processing device 200 (user machine), and uses the image processing device 600 (test target machine) to determine whether the change value of the control parameter complies with the regulation (standard). To test. Specifically, the control unit 310 (pass / fail result transmission unit) of the server 300 sets the control parameter of the image processing device 600 to a changed value and instructs the execution of the standard test. The control unit 310 receives the measurement result of the intensity of the electromagnetic interference wave from the standard tester 550, determines the pass / fail of the standard test, and notifies the image processing device 200 of the determination result. The control parameter change processing of the image processing system 100 including the server 300 will be described later with reference to FIGS. 8A and 8B.
The communication unit 340 transmits and receives communication data with the image processing devices 200 and 600 and the standard tester 550. The storage unit 320 stores the test image processing device database 330 (see FIG. 7 described later).

FIG. 7 is a data structure diagram of the test image processing apparatus database 330 stored in the server 300 according to the present embodiment. The test image processing device database 330 is tabular data, and one row (record) indicates one image processing device 600 (test target machine). The record of the test image processing apparatus database 330 is configured to include columns (attributes) of identification information 331, model 332, address 333, test room 334, and tester 335.
The identification information 331, the model 332, and the address 333 are the identification information, the model name, and the network address of the image processing apparatus 600. The test room 334 is identification information of the test room 500 (see FIG. 1) in which the image processing device 600 is installed. The test device 335 is identification information of the standard test device 550 (see FIG. 1) for measuring the intensity of the electromagnetic interference wave generated by the image processing device 600.

The model name included in the model 332 is not limited to one, and a plurality of model names may be included as long as they can be regarded as equivalent for the electromagnetic interference test. Further, the identification information of the standard tester 550 included in the tester 335 is not limited to one, and when the electromagnetic interference wave is measured by a plurality of standard testers 550, a plurality of identification information is included.
The record 339 is the information of the image processing apparatus 600 whose identification information 331 is "T1234", the model name is "C1234", and the network address is "111.22.23.4". Further, the image processing device 600 is installed in the test room 500 of "R34" and is measured by the standard tester 550 whose identification information is "A1234".

≪Control parameter change processing≫
Before explaining the control parameter change process, the reinforcement learning of the control parameters executed by the overall control CPU 211 (control unit) of the image processing device 200 (user machine) will be described. The control parameters that are the targets of reinforcement learning in this embodiment are control parameters that set the voltage V1 of the DC component, the amplitude V2 of the AC component, and the frequency f (see FIG. 4), and are optimally optimized by different reinforcement learning. It is adjusted to be.

The optimum condition is that the print result of the test chart is the best. Specifically, it is optimal that the test chart is printed, the image of the print result is read by the scanner 291 and there is no maximum density error, halftone density error, and image noise in the read image. The image processing device 200 executes reinforcement learning to obtain optimum voltage V1, amplitude V2, and frequency f set values (control parameter values).

The state in the reinforcement learning of the voltage V1 (also referred to as reinforcement learning A) is the set value (control parameter value) of the voltage V1, the action is the change of the set value, and the reward is the amount of decrease in the error of the maximum concentration before and after the change. (Amount of improvement). The state in the reinforcement learning of the amplitude V2 (also referred to as the reinforcement learning B) is the set value of the amplitude V2, the action is the change of the set value, and the reward is the amount of decrease in the error of the halftone density before and after the change. The state in the reinforcement learning (also referred to as reinforcement learning C) of the frequency f is the set value of the frequency f, the action is the change of the set value, and the reward is the amount of reduction of the image noise before and after the change.

The image processing device 200 includes reinforcement learning A for improving the error of the maximum density (corresponding to the developing voltage V1), reinforcement learning B for improving the error of the halftone density (corresponding to the amplitude V2), and reinforcement learning C for improving the image noise (frequency f). (Corresponding to), three reinforcement learnings are executed in parallel at the same time. Specifically, the image processing device 200 (overall control CPU 211) has a plurality of test patches with a plurality of development voltages near the currently set development voltage V1 (DC component of the development bias), and an amplitude V2 (AC component of the development bias). A test chart containing a plurality of test patches having a plurality of amplitudes in the vicinity and a plurality of test patches having a plurality of frequencies in the vicinity of the frequency f is printed (see step S35 in FIG. 9A described later), and the print result is read by the scanner 291. (See step S36).

Subsequently, the image processing apparatus 200 calculates a reward from the reduction amount of the error of the maximum density in the read image, the reduction amount of the error of the halftone density, and the reduction amount of the image noise, and the above-mentioned reinforcement learnings A and B are performed. , C are executed (see step S37). The larger the amount of decrease, the larger the reward. Q-learning may be adopted as reinforcement learning, and the image processing device 200 may update the value (action value) of the action (change of the set value) in the state (set value) based on the reward. Based on this reward, the machine learning model 213 for improving image quality is updated. The machine learning model 213 (see FIG. 3) is data showing this behavioral value. Then, based on the machine learning model 213 obtained as described above, the change values (voltage V1', amplitude V2', and frequency f') of the control parameters are tentatively determined in order to further improve the image quality (see step S38). ).

In the control parameter change process, the image processing device 200 tentatively determines the change values of the voltage V1, the amplitude V2, and the frequency f (see step S38), and then transmits the change values to the server 300 (see step S39). The image processing apparatus 200 changes after confirming that the changed value passes the electromagnetic interference test (see steps S41 to S47 in FIG. 9B described later).

≪Control parameter change processing: Operation of the entire image processing system≫
FIG. 8A is a sequence diagram (1) of the control parameter change processing of the image processing system 100 according to the present embodiment. FIG. 8B is a sequence diagram (2) of the control parameter change processing of the image processing system 100 according to the present embodiment. The control parameter change process executed by the image processing system 100 will be described with reference to FIGS. 8A and 8B. The reinforcement learning executed by the image processing device 200 (user machine) will be described with reference to FIGS. 9A and 9B described later.

In step S11, the image processing device 200 (the overall control CPU 211 that functions as the control unit or the machine learning execution unit) executes reinforcement learning. At this time, the image processing device 200 tentatively updates the machine learning model 213.
In step S12, the image processing device 200 (the overall control CPU 211 that functions as a tentative determination unit) tentatively determines (tentatively determines) the changed values (control parameter values) of the voltage V1, the amplitude V2, and the frequency f. The details of the processes of steps S11 to S12 will be described later with reference to FIG. 9A.

In step S13, the image processing device 200 transmits its own model name and the three changed values to the server 300. The changed values of the voltage V1, the amplitude V2, and the frequency f (also referred to as the changed control parameters) are referred to as the voltage V1', the amplitude V2', and the frequency f', respectively.
In step S14, the server 300 executes steps S15 to S19 for each combination of voltage V1', amplitude V2', and frequency f'. As the combination, there are a total of seven combinations, one is a combination of three change values, three is a combination of two change values, and three is a combination of one change value. Although one changed value cannot be said to be a combination of changed values, it is described as a combination for convenience.

In step S15, the server 300 transmits the changed value included in the combination to the image processing device 600 (test target machine), and instructs the image processing device 600 to change the set value of the control parameter to the changed value. Specifically, the server 300 searches the test image processing device database 330 (see FIG. 7) for a record including the model name received by the model 332, and obtains an image of the same model as the image processing device 200 (user machine). The processing device 600 (test target machine) is specified. Next, the server 300 transmits a change value to the specified image processing device 600, and instructs the server 300 to change the set value of the control parameter of the image processing device 600 to the transmitted change value.

In step S16, the server 300 (control unit 310 (pass / fail result transmission unit)) instructs the image processing device 600 and the standard tester 550 to perform the test. The standard tester 550 is the standard tester 550 (see FIG. 1) shown in the tester 335 (see FIG. 7) corresponding to the image processing device 600 specified in step S15.
In step S17, the image processing device 600 (test target machine) sets the change value received in step S15 as a control parameter, and prints a test chart. At this time, printing is executed by combining the changed value (any of the voltage V1', the amplitude V2', and the frequency f') received in step S15 with the default value for each parameter. After printing, the image processing device 600 returns the settings.
In step S18, the standard tester 550 measures the electromagnetic interference wave for each of the above printing processes.
In step S19, the standard tester 550 transmits the measurement result to the server 300.

In step S20, the server 300 proceeds to step S21 if steps S15 to S19 are executed for all combinations of voltage V1', amplitude V2', and frequency f', and returns to step S15 if there are unprocessed combinations.
In step S21, the server 300 compares the measurement result with the regulation standard limit value (see FIG. 5) for each of the above combinations (7 patterns) of the voltage V1', the amplitude V2', and the frequency f', and determines pass / fail. Then, the pass / fail result and the measurement result of each are transmitted to the image processing device 200.

Moving on to FIG. 8B, the image processing device 200 (overall control CPU 211 functioning as an update unit) has one of steps S23, S25, S27, and S28 described later according to the received pass / fail result and measurement result for each of the above combinations. Perform one process.
In step S22, the image processing apparatus 200 proceeds to step S23 if the three combinations of voltage V1', amplitude V2', and frequency f'pass (step S22 → YES), and if it fails (step S22 → NO). The process proceeds to step S24.
In step S23, the image processing apparatus 200 updates all the control parameters of the voltage V1, the amplitude V2, and the frequency f to the tentatively determined change values (voltage V1', amplitude V2', frequency f'). Further, the image processing device 200 updates the machine learning model 213 by giving a positive reward to all the reinforcement learnings (reinforcement learning A, B, C) of the voltage V1, the amplitude V2, and the frequency f.

In step S24, the image processing apparatus 200 proceeds to step S25 if there is a passing combination of the two combinations of the voltage V1', the amplitude V2', and the frequency f'(step S24 → YES), and the two changed values If all the combinations are unacceptable (step S24 → NO), the process proceeds to step S26.
In step S25, the image processing apparatus 200 extracts a combination of two changed values that have passed. When there is only one extracted combination, the image processing apparatus 200 updates the control parameter to the changed value of this combination. When there are a plurality of extracted combinations, the image processing apparatus 200 selects the combination having the best measurement result (for example, the measurement result in step S18 is the lowest), and updates the control parameter to the changed value of this combination. Next, the image processing device 200 gives a positive reward for the reinforcement learning corresponding to the two control parameters updated to the changed value, and a negative reward for the reinforcement learning corresponding to the control parameter not updated. Give and update the machine learning model 213.

In step S26, the image processing apparatus 200 proceeds to step S27 if there is a pass combination in one of the voltage V1', the amplitude V2', and the frequency f'(step S26 → YES), and the image processing apparatus 200 proceeds to step S27. If all the combinations are unsuccessful (step S26 → NO), the process proceeds to step S28.
In step S27, the image processing apparatus 200 extracts one combination of changed values that has passed. When there is only one extracted combination, the image processing apparatus 200 updates the control parameter to the changed value of this combination. When there are a plurality of extracted combinations, the image processing apparatus 200 selects the combination having the best measurement result (for example, the measurement result in step S18 is the lowest), and updates the control parameter to the changed value of this combination. Next, the image processing device 200 gives a positive reward for the reinforcement learning corresponding to one control parameter updated to the changed value, and a negative reward for the reinforcement learning corresponding to the control parameter not updated. Give and update the machine learning model 213.
In step S28, the image processing device 200 does not update any of the control parameters, and gives a negative reward to all the reinforcement learning (reinforcement learning A, B, C) of the voltage V1, the amplitude V2, and the frequency f, and the machine. The learning model 213 is updated.

<< Control parameter change processing: Operation of image processing device >>
FIG. 9A is a flowchart (1) of the printing process of the image processing device 200 (user machine) according to the present embodiment. FIG. 9B is a flowchart (2) of the printing process of the image processing device 200 (user machine) according to the present embodiment. The printing process executed by the image processing apparatus 200 (overall control CPU 211 (control unit)) corresponding to steps S11 to S13 and S21 to S28 of FIGS. 8A and 8B will be described with reference to FIGS. 9A and 9B.
In step S31, the image processing apparatus 200 proceeds to step S32 if there is a print instruction (step S31 → YES), and returns to step S31 if there is no print instruction (step S31 → NO).

In step S32, the image processing apparatus 200 applies a set voltage (development voltage) based on the set value of the current control parameter to the developer 263.
In step S33, the image processing apparatus 200 proceeds to step S35 if a predetermined number of sheets have been printed since the previous test chart printing (step S33 → YES), and proceeds to step S34 if not printing (step S33 → NO).
In step S34, the image processing apparatus 200 prints one sheet and returns to step S31.

In step S35, the image processing apparatus 200 prints a test chart.
In step S36, the image processing apparatus 200 acquires the print result of the test chart print read by the scanner 291 and calculates the error of the maximum density, the error of the halftone density, and the amount of image noise.
In step S37, the image processing device 200 first executes reinforcement learning. Specifically, the image processing device 200 calculates from the value calculated in step S36 the amount of reduction (improvement amount) of the error and noise amount in the previous change (reinforcement learning action) as a reward, and the action value (FIG. 3). The machine learning model 213) is tentatively updated to execute reinforcement learning. This update is performed in steps S41 to S47 (see FIG. 9B).

In step S38, the image processing apparatus 200 provisionally determines the changed values (voltage V1', amplitude V2', and frequency f') of the voltage V1, the amplitude V2, and the frequency f based on the machine learning model temporarily updated in the step S37. To do. After the tentative determination, the image processing apparatus 200 undoes the tentative update of the machine learning model 213 in step S37.
In step S39, the image processing device 200 transmits the change value tentatively determined in step S38 to the server 300.
The steps after step S40 are the same as those after step S21 in FIGS. 8A and 8B.

≪Characteristics of control parameter change processing≫
In the control parameter changing process, the image processing device 200 tentatively determines the changed values of the voltage V1, the amplitude V2, and the frequency f (see step S12 in FIG. 8A and step S38 in FIG. 9A). Subsequently, the image processing apparatus 200 changes the control parameters and updates the machine learning model 213 (step S23 in FIG. 8B) with respect to the changed value that the electromagnetic interference wave after the change does not exceed the limit value (passes the standard test). S25, S27, see steps S42, S44, S46 of FIG. 9B). Therefore, the image processing apparatus 200 can improve the image quality (reduce the error of the maximum density, the error of the halftone density, and the image noise) while observing the standard.

In the above-described embodiment, the process of changing the control parameters related to the voltage V1 of the DC component, the amplitude V2 of the AC component, and the frequency f has been described. The control parameters of the image processing device 200 are not limited to these three, and other control parameters also exist, and may be the target of optimization by reinforcement learning. The reward is not limited to the image quality, and may be calculated from, for example, short processing time, low noise, low power consumption, low toner consumption, and the like.

≪Modification example 1: Server is pass value parameter database≫
In the above-described embodiment, the server 300 sets the change value of the control parameter transmitted from the image processing device 200 (user machine) in the image processing device 600 (test target machine), tests the image processing device 600 (test target machine), and notifies the pass / fail result. (See steps S14 to S21 of FIG. 8A). On the other hand, the server 300 may store the changed value (control parameter value) that has passed, and notify the changed value that has passed the test of passing without a test.

FIG. 10 is a functional block diagram of the server 300A according to the first modification of the present embodiment. Compared to the server 300 (see FIG. 6), the server 300A further stores the pass value parameter database 410 (see FIG. 11 below).

FIG. 11 is a data configuration diagram of the pass value parameter database 410 stored in the server 300A according to the first modification of the present embodiment. The pass value parameter database 410 is tabular data, and stores pass values for each model and parameter. The columns (attributes) of the pass value parameter database 410 include the model 411, the parameter 412, and the pass value 413. The model 411 is the model name of the image processing apparatus 200, the parameter 412 indicates the name of the control parameter of the model, and the pass value 413 indicates the pass control parameter value.

Record 419 indicates that the image processing apparatus 200 having the model name "C1234" passed the test with the control parameter "AC frequency" being "2.9 KHz".
FIG. 12A is a sequence diagram (1) of control parameter change processing of the image processing system according to the first modification of the present embodiment. FIG. 12B is a sequence diagram (2) of control parameter change processing of the image processing system according to the first modification of the present embodiment.
Steps S51 to S53 are the same as steps S11 to S13 (see FIG. 8A).

In step S54, the server 300A determines whether or not the received change value is a pass value. Specifically, the server 300A searches for a record of the pass value parameter database 410, which matches the model name received by the model 411 and whose changed values correspond to the parameter 412 and the pass value 413. If there is a record (step S54 → YES), the server 300A proceeds to step S55, and if there is no record (step S54 → NO), the server 300A proceeds to step S56.

In step S55, the server 300A notifies the pass. The image processing device 200 that has received the pass proceeds to step S66 (see FIG. 12B).
Steps S56 to S61 are the same as steps S14 to S19 (see FIG. 8A).
In step S62, the server 300A compares the measurement result with the standard limit value of the regulation (see FIG. 5), determines pass / fail, and if it passes (step S62 → YES), proceeds to step S63, and if it fails (see FIG. 5). Step S62 → NO) Proceed to step S64.

In step S63, the server 300A adds the passed change value to the pass value parameter database 410. Specifically, the server 300A adds a record to the pass value parameter database 410. Next, the server 300A updates the model 411 of the added record to the model name received in step S53, and updates the parameter 412 and the pass value 413 to the parameter of the changed value and its value, respectively.
Steps S64 to S72 are the same as steps S20 to S28 (see FIGS. 8A and 8B), respectively.

Since the server 300A includes the pass value parameter database 410, the response of the server 300A becomes faster. Specifically, if the change value transmitted by the image processing apparatus 200 is a change value that has passed in the past, the server 300A notifies the pass without testing (see step S55). Therefore, in the printing process of the image processing apparatus 200 (see FIG. 9A), the waiting time from transmitting the changed value to receiving the pass / fail result in steps S39 to S40 is shortened. In addition, useless tests that have passed the test can be reduced, and the power required for the test is reduced. Further, even when the changed value is continuously received from a plurality of image processing devices 200 (user machines) of the same model, the server 300A can omit the test of the changed value having a passing record, and can quickly perform the test without a passing record. It can be started and the performance of the entire image processing system 100 is improved.

<< Modification 2: Image processing device parameter type database >>
In the first modification of the present embodiment, the server 300A stores the pass value parameter database 410, and notifies the pass without performing a test on the stored changed value. On the other hand, the image processing device 200 (user machine) may store the control parameters that do not need to be tested, and reduce the inquiry of the changed value (see step S53 in FIG. 12A).

FIG. 13 is a data configuration diagram of the parameter type database 420 stored in the image processing apparatus 200 according to the second modification of the present embodiment. The parameter type database 420 stores the control parameters in which the change in the control parameter value affects the intensity of the electromagnetic interference wave and the control parameters in which the change does not affect the intensity of the electromagnetic disturbance. In FIG. 13, “AC frequency”, which is the frequency f of the AC component, and “DC voltage”, which is the voltage V1 of the DC component, are stored as parameters that affect the intensity of the electromagnetic interference wave. Further, the rotation speed of the fixing roller motor provided in the fixing device 266 (see FIG. 2) is stored as a parameter that does not affect the intensity of the electromagnetic interference wave.

When the image processing apparatus 200 tentatively determines the change value (see step S12 in FIG. 8A), if the change value is a change value of a parameter that affects the intensity of the electromagnetic interference wave, the image processing apparatus 200 inquires of the server 300 (step S13). (See), if there is no effect, change without contacting the server. As described above, when there is no influence, the process of changing the control parameter of the image processing device 200 can be speeded up by not making an inquiry to the server 300. Moreover, the load on the server 300 can be reduced.

<< Modification 3: Test exemption parameter database for image processing equipment >>
In the first modification, the server 300A stores the pass value parameter database 410 to prevent unnecessary tests. The image processing device 200 (user machine) may store control parameter values that do not require testing so as not to make unnecessary inquiries to the server 300.

FIG. 14 is a data configuration diagram of the test exemption parameter database 430 stored in the image processing apparatus 200 according to the third modification of the present embodiment. The test exemption parameter database 430 is tabular data and stores control parameter values that are known to pass the test and are exempt from the test (does not affect the test). The test exemption parameter database 430 is configured to include an attribute of parameter 431 indicating the name of the parameter exempted from the test and an attribute of value 432 indicating the parameter value. Record 439 shows that the "AC frequency", which is the frequency f of the AC component, is exempt from the test if it is between 2.8 KHz and 3.2 KHz.

If the record corresponding to the tentatively determined change value (see step S12 in FIG. 8A) is in the test exemption parameter database 430, the image processing apparatus 200 changes it without inquiring the server, and inquires the server if it does not. By referring to the test exemption parameter database 430, the number of inquiries to the server 300 can be reduced, and the process of changing the control parameters of the image processing device 200 can be speeded up. Moreover, the load on the server 300 can be reduced.

<< Modification 4: Server pass / fail history parameter database >>
In the first modification, if the changed value in the inquiry is included in the pass value parameter database 410 (see FIG. 11) (see step S54 → YES in FIG. 12A), the server 300A notifies the pass without testing (step S55). See). On the other hand, the server 300A may store the pass value and the fail value.

FIG. 15 is a data configuration diagram of the pass / fail history parameter database 440 stored in the server 300A according to the fourth modification of the present embodiment. The pass / fail history parameter database 440 is tabular data, and stores the pass value and the fail value of the control parameter value for each model and control parameter. The pass / fail history parameter database 440 includes attributes of model 441, parameter 442, pass history 443, and fail history 444.

The model 441 includes the model name of the image processing apparatus 200, and the parameter 442 includes the name of the control parameter. The pass history 443 (pass range) includes control parameter values that have a pass record and do not require a test. The failure history 444 (failure range) includes control parameter values that have a failure record and do not require a test. Record 449 is a pass for the image processing device 200 whose model is "C1234" if the control parameter value of "AC frequency" is 2.9 KHz to 3.1 KHz, and no test is required. If it is 5.5 KHz or higher, it means that the test is not passed and the test is unnecessary.

The server 300A notifies the pass if the change value in the inquiry from the image processing device 200 (see step S53 in FIG. 12A) is included in the pass history 443, and notifies the fail if it is included in the fail history 444. If the changed value is not included in either the pass history 443 or the fail history 444, the server 300A executes the test and notifies the image processing device 200 of the pass / fail result. Subsequently, the server 300A adds a change value (control parameter value) to the pass history 443 or the fail history 444 according to the pass / fail.

If the change value is not included in either the pass history 443 or the fail history 444, the server 300A instead of executing the test, the control parameter value of the pass history 443 closest to the change value, or the fail history 444. The control parameter value of may be notified together with the pass / fail. For example, suppose that the image processing device 200 whose model name is "C1234" inquires about the control parameter (frequency f of the AC component) of the "AC frequency" of 3.2 KHz. The server 300A searches for the control parameter value closest to 3.2 KHz among the control parameter values included in the pass history 443 and the fail history 444. The control parameter value closest to 3.2 KHz is 3.1 KHz included in the pass history 443, and the server 300A notifies the image processing device 200 of 3.1 KHz as a conditional pass control parameter. ..

Upon receiving the conditional pass of 3.1 KHz, the image processing apparatus 200 changes the frequency f of the AC component to 3.1 KHz (see step S22 and subsequent steps in FIG. 8B). If the change value of the inquiry is 4.3 KHz, the server 300A notifies the image processing device 200 of 4.3 KHz as a conditionally rejected control parameter. The image processing device 200 updates the action value with a negative reward for the change value of 4.3 KHz (see steps S25, S27, and S28 in FIG. 8B).

<< Modification 5: Server pass history parameter database >>
The pass value parameter database 410 (see FIG. 11) and the pass / fail history parameter database 440 (see FIG. 15) store the pass control parameter values. The server 300A may store the passed control parameter value in association with the surrounding environment and usage history of the image processing device 200.

FIG. 16 is a data configuration diagram of the pass history parameter database 450 stored in the server 300A according to the fifth modification of the present embodiment. The pass history parameter database 450 is tabular data, and stores the pass history in association with the surrounding environment and usage history of the image processing device 200 in addition to the model. The pass history parameter database 450 is configured to include attributes of model 451, parameter 452, environment 453, usage history 454, and pass history 455.

The model 451 and the parameter 452 and the pass history 455 are the same as the model 441, the parameter 442, and the pass history 443 of the pass / fail history parameter database 440 (see FIG. 15), respectively. The environment 453 indicates the surrounding environment of the image processing apparatus 200, and includes, for example, "high humidity" and "low humidity". The usage history 454 indicates the cumulative number of prints.
The record 459 is a pass history of the control parameter value of the "AC frequency" for the image processing device 200 in which the installed environment 453 is "high humidity" and the model 451 with a cumulative print count of less than 200,000 is "C1234". It shows that 455 is 2.9 KHz to 3.1 KHz.

When making an inquiry (see step S13 in FIG. 8A), the image processing device 200 transmits a changed value including its own environment and usage history in addition to the model. The server 300A notifies the pass if the change value is included in the pass history 455, and executes the test if it is not included.

<< Modifications 6 and 7: Firmware update >>
In the above-described embodiment, the image processing apparatus 200 changes its own control parameters after receiving the acceptance notification. In addition to this, the control parameters included in the firmware may be updated. The firmware is the firmware 223 (see FIG. 3) executed by the printer control CPU 221. The image processing device 200 downloads the firmware from the server and updates the firmware 223 periodically or whenever there is a master firmware update stored in the server.

FIG. 17 is a functional block diagram of the server 300B according to the sixth modification of the present embodiment. Compared to the server 300 (see FIG. 6), the server 300B further stores the average pass value parameter database 460 (see FIG. 18 below) and the firmware repository 470. The firmware repository 470 is a repository of the firmware 223 of the printer control unit 220, and stores the firmware 223 for each model of the image processing device 200.

FIG. 18 is a data configuration diagram of the average pass value parameter database 460 stored in the server 300B according to the modification 6 of the present embodiment. The average pass value parameter database 460 is tabular data, and includes the specified values of the control parameters included in the firmware in the current version and the average values of the control parameters that have passed the standard test for each model and control parameter. The average pass value parameter database 460 includes attributes of model 461, parameter 462, current default value 463, and pass average value 464.

The model 461 includes the model name of the image processing apparatus 200, and the parameter 462 includes the name of the control parameter. The current specified value 463 is the specified value of the control parameters included in the firmware in the current version, and the passing average value 464 is the average value of the control parameters that have passed the standard test.
In record 469, for the image processing device 200 whose model is "C1234", the specified value of the control parameter of "AC frequency" included in the current firmware 223 is 3.0 KH, and the average value that passed the test is 3. It indicates that it is 1 KHz.

The server 300B executes the test, and if the result is a pass, updates the pass average value 464. The server 300B periodically updates the parameter 462 included in the firmware for the model 461 stored in the firmware repository 470 to the pass average value 464. Next, the server 300B updates the current specified value 463 to the passing average value 464. After the update, the control parameters included in the firmware downloaded from the server 300B by the image processing apparatus 200 are the updated pass average values.

The server 300B may be updated when the difference between the current specified value 463 and the passing average value 464 becomes larger than a predetermined value. Further, the average pass value parameter database 460 may store the current specified value and the pass average value in association with the environment and usage history. In this case, the firmware repository 470 stores the firmware 223 in association with the environment and usage history in addition to the model.

FIG. 19 is a data configuration diagram of the average pass value parameter database 460A stored in the server 300B according to the modification 7 of the present embodiment. Compared to the average pass value parameter database 460 (see FIG. 18), the attributes of environment 465 and usage history 466 have been added. The environment 465 and the usage history 466 are the same as the environment 453 and the usage history 454 of the pass history parameter database 450 (see FIG. 16), respectively. The server 300B updates the parameter 462 included in the firmware associated with the model 461, the environment 465, and the usage history 466 stored in the firmware repository 470 to the pass average value 464.

<< Modification 8: Data transmitted from the image processing device to the server >>
In the above-described embodiment, the data transmitted from the image processing device 200 to the server 300 is a model name and three changed values (see step S13 in FIG. 8A). Further, the image processing device 600 (test target machine) did not store the machine learning model 213.
On the other hand, the image processing device 600 and the image processing device 200 (user machine) may have the same configuration. Specifically, the image processing device 600 and the machine learning model 213 are stored in the same manner as the image processing device 200. Further, both the changed value (control parameter value) and the machine learning model 213 are transmitted from the image processing device 200 to the server, and further from the server to the image processing device 600 (see steps S13 and S15). In this way, the image processing device 600 and the image processing device 200 (user machine) may have the same configuration and then be tested (see steps S15 to S18).
By doing so, the image processing device 200 (user machine) and the image processing device 600 (test target machine) have the same configuration, and a more accurate test can be executed.

<< Modification 9: Reinforcement learning for multiple control parameters >>
In the above-described embodiment, the targets of reinforcement learning are the voltage V1 of the DC component, the amplitude V2 of the AC component, and the frequency f (see FIG. 4), and the reinforcement learning is executed for each control parameter. This is because the voltage V1 of the DC component, the amplitude V2 of the AC component, and the frequency f are independent, and the change of one control parameter does not affect the other. For example, changing the voltage V1 does not affect the halftone density.

If there is a mutual influence between a plurality of control parameters in other control parameters, reinforcement learning may be executed with these control parameters as a set. In this case, the state is a set of control parameter values, and the action is a change of a plurality of control parameter values. Further, instead of performing the test for each changed value (steps S14 to S20 in FIG. 8A), the test is performed after changing a plurality of control parameters of the image processing apparatus 600 (test target machine) at the same time.

≪Modification example 10: Reinforcement learning for each paper type≫
In the above-described embodiment, the paper on which the test chart is printed is not distinguished. In order to determine the optimum value of the control parameter according to the paper type (high-quality paper, matte paper, etc.), the image processing apparatus 200 may execute reinforcement learning for each control parameter and the paper type. The machine learning model 213 (behavioral value) exists for each control parameter and paper type.

≪Other variants≫
In the above-described embodiment, the image processing apparatus 200 prints a test chart every time a predetermined number of sheets are output to execute reinforcement learning (see steps S33 to S37 in FIG. 9A). The image processing device 200 may execute reinforcement learning at another timing. For example, the image processing device 200 may execute the control parameter change processing during the idling time when there is no print job to be executed. Alternatively, in the modification 10, the image processing apparatus 200 may execute the control parameter change processing at the start of the print job using the new paper type.

The above-described embodiment is an example of an image processing system 100 including an image processing device 200. Not only the image processing device 200, but also a manufacturing device having control parameters, a medical device, a home appliance, and the like, it is possible to optimize the control parameters of the device within the scope of the regulation. Specifically, in a control management system including a device and a server, the device determines a change value of a control parameter using machine learning and transmits it to the server. The server sets the control parameters of the device of the same type as the device to the changed value, executes the standard test, and notifies the device of the pass / fail result. The device changes the control parameters if it passes.

In the above-described embodiment, a machine learning technique is used to determine a change value of a control parameter, and it is tested whether the change value satisfies the regulation. The change value of the control parameter may be determined by using a technique different from the machine learning technique. For example, a local search algorithm such as a hill climbing method may be used to determine the optimum change value of the control parameter.

In the modified examples 6 and 7, the server 300B is a server for determining the pass / fail of the change value of the control parameter, and stores the firmware repository 470. On the other hand, the server that determines the pass / fail of the change value of the control parameter and the server that stores the firmware repository 470 and transmits the firmware to the image processing device 200 may be different servers.

Although the embodiments of the present invention and variations thereof have been described above, these embodiments are merely examples and do not limit the technical scope of the present invention. The present invention can take various other embodiments, and various modifications such as omission and substitution can be made without departing from the gist of the present invention. These embodiments and modifications thereof are included in the scope and gist of the invention described in the present specification and the like, and are also included in the scope of the invention described in the claims and the equivalent scope thereof.

100 Image processing system (control management system)
200 Image processing device (user machine, device)
211 Overall control CPU (control unit, machine learning execution unit, provisional determination unit, update unit, machine learning processing execution stage, provisional determination means, update means)
212 Storage unit 213 Machine learning model 214 Program 219 Network interface (communication unit, communication means)
223 Firmware 260 Image forming unit 300, 300A, 300B Server 310 Control unit (pass / fail result transmission unit, pass / fail determination means)
320 Storage unit 330 Test image processing device database 340 Communication unit 410 Pass value parameter database 420 Parameter type database 430 Test exemption parameter database 440 Pass / fail history parameter database 450 Pass history parameter database 460, 460A Average pass value parameter database 470 Firmware repository (firmware) )
550 standard tester 600 image processing device (test target machine)

Claims (20)

An image processing system equipped with an image processing device and a server.
The image processing device is
Image forming part and
A machine learning execution unit that executes machine learning related to a predetermined control parameter of the image forming unit, and a machine learning execution unit.
A tentative determination unit that tentatively determines a new control parameter as a changed control parameter based on the result of the machine learning,
A communication unit that transmits the changed control parameter to the server before updating the changed control parameter as a control parameter to be executed by the image processing apparatus is provided.
The server
It is provided with a pass / fail result transmission unit that determines the pass / fail result of the standard test in the changed control parameter and transmits it to the image processing apparatus.
The image processing device further
It is provided with an update unit that updates the changed control parameter as a control parameter executed by the image processing apparatus according to the pass / fail result.
An image processing system characterized by this. An image processing device of the same type as the image processing device is connected to the server.
The pass / fail result transmission unit changes the control parameter of the image processing device of the same type to the changed control parameter, and causes the standard test to be performed using the image processing device of the same type.
The image processing system according to claim 1. Control parameters that have passed the standard test are stored in the storage unit of the server.
When the changed control parameter is included in the passed control parameter, the pass / fail result transmitting unit transmits that the standard test has passed.
The image processing system according to claim 1. In the storage unit of the image processing device, a control parameter group that affects the pass / fail of the standard test and a control parameter group that does not affect the standard test are stored.
The update part
The control parameters included in the control parameter group that affect the pass / fail of the standard test are changed after receiving the pass from the server by the communication unit.
The control parameters included in the control parameter group that does not affect the pass / fail of the standard test are changed regardless of the pass / fail result.
The image processing system according to claim 1. A range of control parameters that do not affect the standard test is stored in the storage unit of the image processing device.
The update part
Control parameters included outside the range of control parameters that do not affect the standard test are changed after receiving a pass from the server by the communication unit.
Control parameters included in the range of control parameters that do not affect the standard test are changed regardless of the pass / fail result.
The image processing system according to claim 1. In the storage unit of the server, a pass range, which is a range of control parameters that have passed the standard test, and a fail range, which is a range of control parameters that have failed the standard test, are stored.
The pass / fail result transmitter is
If the changed control parameter is included in the pass range, it is transmitted that the standard test has passed.
If the changed control parameter is included in the fail range, the standard test is transmitted as fail and the result is transmitted.
When the changed control parameter is not included in the pass range or the fail range, the control parameter value closest to the control parameter corresponding to the changed control parameter in the pass range is transmitted. Or, the control parameters of the image processing apparatus of the same type are changed so as to correspond to the changed control parameters, and the pass / fail result of the standard test is determined and transmitted.
The update part
When the communication unit receives the control parameter value, the control parameter value is changed to the control parameter value.
The image processing system according to claim 2, wherein the image processing system is characterized by the above. The communication unit of the image processing device transmits either one or both of the environmental information and the usage history information to the server together with the changed control parameters.
The storage unit of the server stores the control parameters when the standard test is passed in association with the environmental information and the usage history information.
The pass / fail result transmission unit sets the control parameter after the change to the control parameter when the control parameter passes the standard test associated with either one or both of the environment information and the usage history information transmitted by the image processing apparatus. If applicable, send that the standard test has passed.
The image processing system according to claim 1. In the storage unit of the server, control parameters and firmware that have passed the standard test are stored for each type of the image processing device.
The pass / fail result transmitter changes the firmware of the type so that the control parameter value that has passed the standard test in the type becomes the specified value of the firmware of the type.
The image processing system according to claim 1. In the storage unit of the server, control parameters and firmware when the standard test is passed are stored for each model of the image processing device, for each usage history, and for each environment.
The pass / fail result transmitter uses the firmware so that the type, the usage history, and the control parameter values that have passed the standard test in the environment become the specified values of the type, the usage history, and the firmware of the environment. To change,
The image processing system according to claim 1. The image processing system according to claim 1, wherein the control parameter is a physical parameter that controls the image processing apparatus. The image processing system according to claim 1, wherein the standard of the standard test is a standard corresponding to at least one of a legal regulation and a safety regulation. The image processing system according to claim 1, wherein the machine learning is reinforcement learning accompanied by a predetermined reward determination. It is an image processing device
Image forming part and
A machine learning execution unit that executes machine learning related to a predetermined control parameter of the image forming unit, and a machine learning execution unit.
A tentative determination unit that tentatively determines a new control parameter as a changed control parameter based on the result of the machine learning,
A communication unit that transmits the changed control parameter to the server before updating the changed control parameter as a control parameter to be executed by the image processing device.
An update unit that updates the changed control parameter as a control parameter to be executed by the image processing device according to the pass / fail result of the standard test in the changed control parameter received from the server.
An image processing device comprising. The image processing apparatus according to claim 13, wherein the control parameter is a physical parameter that controls the image processing apparatus. The image processing apparatus according to claim 13, wherein the standard of the standard test is a standard corresponding to at least one of a legal regulation and a safety regulation. It is a program to be executed by an image processing device that is a computer.
The steps to form an image and
A step of executing machine learning regarding control parameters related to the formation of the image, and
A step of tentatively determining a new control parameter as a changed control parameter based on the result of the machine learning, and
Before updating the changed control parameter as a control parameter to be executed by the image processing device, a step of transmitting the changed control parameter to the server and a step of transmitting the changed control parameter to the server.
A step of updating the changed control parameter as a control parameter to be executed by the image processing device according to a pass / fail result of a standard test in the changed control parameter received from the server.
A program to execute. The program according to claim 16, wherein the control parameter is a physical parameter that controls the image processing apparatus. The program according to claim 16, wherein the standard of the standard test is a standard corresponding to at least one of a legal regulation and a safety regulation. A control management system that manages changes in the control parameters of a device having control parameters.
A machine learning execution means for executing machine learning related to the control parameters, and
A tentative determination means for tentatively determining a new control parameter as a changed control parameter based on the result of the machine learning,
Pass / fail determination means for determining the pass / fail result of the standard test in the changed control parameters,
An update means for updating the changed control parameter as a control parameter to be executed by the apparatus according to the pass / fail result, and an update means.
Control management system equipped with. A machine learning execution unit that executes machine learning related to control parameters,
A tentative determination unit that tentatively determines a new control parameter as a changed control parameter based on the result of the machine learning,
Before updating the changed control parameter as a control parameter to be executed by the own device, a communication unit that transmits the changed control parameter to the server, and a communication unit.
An update unit that updates the changed control parameter as a control parameter to be executed by the own device according to the pass / fail result of the standard test in the changed control parameter received from the server.
A device equipped with.

Images