Image forming device

Abstract

To improve the prediction accuracy of a density prediction model.SOLUTION: An image forming apparatus has an image forming unit that forms an image based on an image forming condition, an intermediate transfer body to which the formed image is transferred, a transfer unit that transfers the image on the intermediate transfer body to a sheet, and a measuring unit that measures an image for measurement on the intermediate transfer body. Variation correlation information is acquired that is correlated with a variation in the density of the image formed by the image forming unit. The image forming condition is created from the variation correlation information acquired based on a first determination condition. A first density of a first image for measurement to be formed is determined from the variation correlation information acquired based on the first determination condition. A second density of a second image for measurement to be formed is determined from the acquired variation correlation information based on a second determination condition different from the first determination condition. Whether to change the first determination condition used for creating the image forming condition to the second determination condition is controlled based on the determined first density, a result of measurement of the measured first image for measurement, the determined second density, and a result of measurement of the measured second image for measurement.SELECTED DRAWING: Figure 17

Description

Technical Field

[0001] The present invention relates to an image forming apparatus, and more particularly to an image forming apparatus using, for example, an electrophotographic process.

Background Art

[0002] Due to short-term variations caused by changes in the environment where the apparatus is installed or within the apparatus, and long-term variations caused by changes over time (aging degradation) of the photoreceptor and developer, etc., the density and density gradation of the output image may differ from the desired density and gradation. Therefore, in an image forming apparatus, it is necessary to correct the image forming conditions as needed in consideration of various such variations in order to match the density and gradation of the output image to the desired density and gradation.

[0003] Thus, the process of appropriately correcting changes in density and color tone is generally referred to as calibration. In calibration, for example, several density-uniform pattern images are formed on paper, a photoreceptor, or an intermediate transfer member, etc., the density of the formed pattern is measured and compared with its target value, and various conditions for forming an image are appropriately adjusted based on the comparison result.

[0004] Conventionally, in order to stabilize the density and gradation of the output image described above, for example, as in Patent Document 1, a specific correction pattern such as a gradation pattern is formed on paper. The formed pattern is read by an image reading unit, and the reading result (read data) of the gradation pattern by the image reading unit is fed back to image forming conditions such as γ (gamma) correction, thereby improving the stability of image quality.

[0005] In addition, as described above, the timing when calibration is required varies depending on the situation, including environmental changes and long periods of inactivity, and appropriate tone correction is necessary in various scenarios. For example, especially when powering on in the early morning when environmental changes are likely to occur, when resuming from the power-saving mode, or when the toner supply amount increases due to a high output image DUTY, or conversely, when jobs with a low output image DUTY are continuously performed. As a technology for performing such calibration, for example, a technology such as Patent Document 2 has been proposed. In Patent Document 2, density patch images of each color are formed on an intermediate transfer body or a transfer belt, read by a density detection sensor, and feedback is provided to high-voltage conditions and image processing conditions to adjust the maximum density and halftone gradation characteristics of each color.

[0006] In recent years, there has been an increasing demand for improving user-friendliness, particularly productivity by reducing waiting time and downtime, while maintaining image quality stability. There is also a strong demand for more rapid control of calibration control for image quality stabilization. As a technology to meet such demands, for example, as in Patent Document 3, a model that takes fluctuations in the external environment, image output conditions, and various sensor values as input values is created, and fluctuations in patches for calibration are predicted from the model. By doing so, a technology has been proposed that omits the image formation process of patches that consume a large amount of calibration time.

[0007] Furthermore, as a method of performing control in a model for predicting fluctuations so that an optimal operation value is obtained depending on the usage environment and usage situation, a technology such as Patent Document 4 has been proposed. Patent Document 4 proposes a technology that uses a neural network to learn the characteristics of an image forming apparatus and determines an operation amount from a state prediction value and a target value.

Prior Art Documents

Patent Documents

[0008]

Patent Document 1

Patent Document 2

Patent Document 3

Patent Document 4

Summary of the Invention

Problems to be Solved by the Invention

[0009] However, in such a calibration method that predicts variations in color and density using a model, the following problems occur. When performing calibration control for density adjustment using an optimal density prediction model that individually corresponds to the usage environment, output conditions, and usage status, it is necessary to modify the current prediction model. This is because, usually at the initial stage, it is common to use an average model that can cover a certain range of usage environments and situations, and it is not always optimal for individual usage environments.

[0010] To modify the prediction model, data that combines actual density variations with the environment, output conditions, etc. is required. Therefore, usually, control for actually forming a calibration patch and performing density adjustment is used in combination, and at the timing of executing calibration control using the patch, data for modifying the prediction model is simultaneously acquired.

[0011] However, when changes in internal disturbance / external disturbance factors of the image forming apparatus occur at the timing of acquiring correction data for modifying the prediction model and actually updating or switching the prediction model, the difference between the predicted density and the actual density may deteriorate more after modification than before modification. Examples of internal disturbance factors include replacement of parts of the image forming apparatus and change of image forming conditions. Examples of external disturbance factors include sudden changes in the temperature and humidity environment. From the characteristic of modifying the prediction model for each image forming apparatus, it is important to confirm that the prediction accuracy is reliably maintained and then perform prediction model modification from the viewpoint of ensuring quality.

[0012] The present invention has been made in view of the above circumstances. The object is to provide an image forming apparatus capable of achieving correction of an appropriate prediction model in the use environment in a short period in calibration for stabilizing color tone and density gradation control.

Means for Solving the Problems

[0013] In order to achieve the above object, the present invention has the following configuration. That is, an image forming means for forming an image based on image forming conditions, An intermediate transfer body onto which the image formed by the image forming means is transferred, Transfer means for transferring the image on the intermediate transfer body to a sheet, Measuring means for measuring a measurement image on the intermediate transfer body, Acquisition means for acquiring variation correlation information correlated with variation in density of an image formed by the image forming means, Generation means for generating the image forming conditions from the variation correlation information acquired by the acquisition means based on a first determination condition, Determining means for determining a first density of a first measurement image formed by the image forming means from the variation correlation information acquired by the acquisition means based on the first determination condition, and determining a second density of a second measurement image formed by the image forming means from the variation correlation information acquired by the acquisition means based on a second determination condition different from the first determination condition, Control means for controlling whether to change the first determination condition used by the generation means to generate the image forming conditions to the second determination condition based on the first density determined by the determination means, the measurement result of the first measurement image measured by the measurement means, the second density determined by the determination means, and the measurement result of the second measurement image measured by the measurement means. An image forming apparatus characterized by the above is provided.

Effects of the Invention

[0014] According to the present invention, it becomes possible to correct an appropriate prediction model in a usage environment in a short period of time.

Brief Description of the Drawings

[0015]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Mode for Carrying Out the Invention

[0016] Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. Note that the following embodiments do not limit the invention according to the claims. Although a plurality of features are described in the embodiments, not all of these plurality of features are essential for the invention, and the plurality of features may be arbitrarily combined. Furthermore, in the accompanying drawings, the same or similar configurations are denoted by the same reference numerals, and redundant descriptions are omitted.

[0017] [Embodiment 1] First, a first embodiment of the present invention will be described. In this embodiment, a method for solving the above problems will be described using an image forming apparatus of an electrophotographic system (or electrophotographic process). Although the description is made in the electrophotographic system, the characteristic points of the control, particularly the matters described in the claims, have the same problems in an inkjet printer, a sublimation printer, etc., and the problems can be solved using the methods described below. Therefore, it is claimed that each image forming apparatus is also included in the invention according to the above claims.

[0018] (Image forming apparatus) (Reader unit) As shown in FIG. 1, the image forming apparatus 100 includes a reader unit A. The document placed on the document table glass 102 of the reader unit A is illuminated by a light source 103, and the reflected light from the document forms an image on a CCD sensor 105 via an optical system 104. The CCD sensor 105 consists of red, green, and blue CCD line sensor groups arranged in three columns, and generates red, green, and blue color component signals for each line sensor. These reading optical system units are moved in the direction of arrow R103 shown in FIG. 1 to convert the image of the document into an electrical signal for each line. On the document table glass 102, there are a positioning member 107 that abuts on one side of the document to prevent the document from being placed obliquely, and a reference white plate 106 for determining the white level of the CCD sensor 105 and performing shading correction in the thrust direction of the CCD sensor 105. The image signal obtained by the CCD sensor 105 is subjected to A / D conversion, shading correction using the reading signal of the reference white plate 106, and color conversion by a reader control unit 108, and then sent to the printer unit and processed by a printer control unit. In addition, an operation unit 20 and a display 218 for an operator to perform operations such as copy start and various settings are connected to the reader unit A. The reader unit A may also be provided with a CPU, a RAM 215, and a ROM 216 for control. These control the reader unit A.

[0019] (Printer unit) As shown in FIG. 1, the image forming apparatus 100 is a tandem type intermediate transfer full-color printer in which image forming units PY, PM, PC, and PK for yellow, magenta, cyan, and black are arranged along an intermediate transfer belt 6 which is an intermediate transfer member.

[0020] In the image forming unit PY, an electrostatic latent image is formed on a photosensitive drum 1Y, and a yellow toner image corresponding to the electrostatic latent image is formed and primarily transferred to the intermediate transfer belt 6. Similarly, in the image forming unit PM, a magenta toner image is formed on a photosensitive drum 1M and primarily transferred onto the yellow toner image on the intermediate transfer belt 6. In the image forming units PC and PK, a cyan toner image and a black toner image are formed on photosensitive drums 1C and 1K respectively, and are similarly sequentially overlaid and primarily transferred to the intermediate transfer belt 6.

[0021] The four-color toner images primary-transferred onto the intermediate transfer belt 6 are conveyed to the secondary transfer unit T2 and are batchwise secondary-transferred onto the sheet P. The recording material P onto which the four-color toner images have been secondary-transferred is conveyed by the conveyance belt 10, is heated and pressed by the fixing device 11 to fix the toner images on the surface, and is then discharged to the outside of the machine body.

[0022] The intermediate transfer belt 6 is supported by being looped around the tension roller 61, the drive roller 62, and the opposing roller 63, and is driven by the drive roller 62 to rotate in the direction of arrow R2 at a predetermined process speed.

[0023] The sheet P drawn out from the cassette 65 is separated one by one by the separation roller 66 and is sent out to the registration roller 67. The registration roller 67 receives and waits for the sheet P in a stopped state, and feeds the sheet P into the secondary transfer unit T2 in synchronization with the toner image on the intermediate transfer belt 6.

[0024] The secondary transfer roller 64 abuts on the intermediate transfer belt 6 supported by the opposing roller 63 to form the secondary transfer unit T2. By applying a DC voltage of positive polarity to the secondary transfer roller 64, the toner image charged negatively and carried on the intermediate transfer belt 6 is secondary-transferred onto the sheet P.

[0025] The image forming units PY, PM, PC, PK are substantially configured in the same manner except that the colors of the toners used in the developing devices 4Y, 4M, 4C, 4K are different, being yellow, magenta, cyan, and black, respectively. Hereinafter, when there is no particular need for distinction, in order to indicate that they are for any of the colors, the subscripts Y, M, C, K attached to the symbols are omitted and they will be described comprehensively.

[0026] As shown in FIG. 1, around the photosensitive drum 1, a charging device 2, an exposure device 3, a developing device 4, a primary transfer roller 7, and a cleaning device are arranged in the image forming unit.

[0027] The photosensitive drum 1 has a photosensitive layer with a negative charging polarity formed on the outer peripheral surface of an aluminum cylinder, and rotates in the direction of the arrow at a predetermined process speed. The photosensitive drum 1 is an OPC photosensitive member with a reflectivity of approximately 40% for near-infrared light (960 nm). However, an amorphous silicon-based photosensitive member or the like with a similar reflectivity may also be used.

[0028] The charging device 2 uses a scorotron charger, irradiates the photosensitive drum 1 with charged particles associated with corona discharge, and charges the surface of the photosensitive drum 1 to a uniform negative potential. The scorotron charger has a wire to which a high-voltage is applied, a shield portion connected to the ground, and a grid portion to which a desired voltage is applied. A predetermined charging bias is applied to the wire of the charging device 2 from a charging bias power source (not shown). A predetermined grid bias is applied to the grid portion of the charging device 2 from a grid bias power source (not shown). Depending on the voltage applied to the wire, the photosensitive drum 1 is charged to approximately the voltage applied to the grid portion.

[0029] The exposure device 3 scans a laser beam with a rotating mirror and writes an electrostatic image of an image on the surface of the charged photosensitive drum 1. A potential sensor (not shown), which is an example of potential detection means, can detect the potential of the electrostatic image formed by the exposure device 3 on the photosensitive drum 1. The developing device 4 attaches toner to the electrostatic image of the photosensitive drum 1 and develops it into a toner image.

[0030] The primary transfer roller 7 presses against the inner surface of the intermediate transfer belt 6 to form a primary transfer portion T1 between the photosensitive drum 1 and the intermediate transfer belt 6. By applying a positive direct current voltage to the primary transfer roller 7, the negative-polarity toner image carried on the photosensitive drum 1 is primarily transferred to the intermediate transfer belt 6 passing through the primary transfer portion T1.

[0031] The image density sensor (patch sensor) 200 is arranged to face the intermediate transfer belt and measures the image density of the unfixed toner. In this embodiment, it is arranged to face the intermediate transfer belt, but it can be arranged as appropriate including a configuration arranged to face the photosensitive drum. Also, the image density sensor arranged on the photosensitive drum, intermediate transfer belt, etc. is a sensor that measures the image density of the unfixed toner. It is also possible to arrange an image density sensor that measures the pattern image after fixing downstream of the fixing device, and it is not limited to the image density sensor described in this embodiment.

[0032] The cleaning device rubs the cleaning blade against the photosensitive drum 1 to collect the residual transferred toner remaining on the photosensitive drum 1 that has escaped transfer to the intermediate transfer belt 6.

[0033] The belt cleaning device 68 rubs the cleaning blade against the intermediate transfer belt 6 to collect the residual transferred toner remaining on the intermediate transfer belt 6 that has escaped transfer to the sheet P and passed through the secondary transfer section T2.

[0034] Note that a potential sensor for measuring the potential on the surface of the photosensitive drum 1 for each color component may be provided and configured to output a signal indicating the potential.

[0035] (Image processing unit) FIG. 2 is a diagram showing the configuration of the printing system according to the present invention. In the figure, 301 is a host computer, and 100 is an image forming apparatus. The host computer 301 and the image forming apparatus 100 are connected by a communication line such as USB2.0 High-Speed, 1000Base-T / 100Base-TX / 10Base-T (compliant with IEEE 802.3).

[0036] In the image forming apparatus 100, the printer controller 300 controls the operation of the entire printer. The printer controller 300 has the following configuration. A host I / F unit 302 that controls input / output with the host computer 301. An input / output buffer 303 for transmitting and receiving control codes from the host I / F unit 302 and data from each communication means. A printer controller CPU 313 that controls the operation of the entire controller 300. A program ROM 304 in which a control program and control data of the printer controller CPU 313 are built-in. A RAM 309 used as a work memory for interpreting the control codes and data, performing calculations necessary for printing, or processing print data. An image information generation unit 305 that generates various image objects based on the settings of data received from the host computer 301. A RIP (Raster Image Processor) unit 314 that expands the image object into a bitmap image. A color processing unit 315 that performs color conversion processing for multi-color. A tone correction unit 316 that executes tone correction for single color. A pseudo halftone processing unit 317 that executes pseudo halftone processing such as a dither matrix or an error diffusion method. An engine I / F unit 318 that transfers the converted image to the image forming engine unit. An image forming engine unit 101 that forms the converted image data as an image. The above is the flow of image processing of the printer controller during basic image formation, which is indicated by thick solid lines.

[0037] The printer controller 300 is in charge of not only image formation but also various control operations. It has a control program in the program ROM 304. The following are included as the control program and data. · A maximum density condition determination unit 306 that performs maximum density adjustment. · A predicted density calculation unit 307 that predicts the density based on the output value from the sensor or the like. · A tone correction table generation unit (γLUT) 308 that performs density tone correction. The generated tone correction table includes, for example, an output density value corresponding to the input density value as a correction value. · A prediction model correction unit 350 that corrects the model for calculating the predicted density. Details of various control operations in the printer controller will be described later.

[0038] Note that the gradation correction table may also be referred to as an image correction condition. Also, since prediction is to perform a given operation based on given parameters and determine a target value, prediction may also be referred to as determination. The value obtained by prediction may be referred to as a prediction result or a determination result.

[0039] In addition, it has a table storage unit 310 for temporarily storing the adjustment results from the maximum density condition determination unit 306 to the gradation correction table generation unit 308. Furthermore, it has an operation panel 218 for giving execution instructions to the operation of the printing apparatus and the above correction process, and a panel I / F unit 311 connecting the printer controller 300 and the operation panel 218. Furthermore, it is composed of an external memory unit 181 used for storing print data and information of various printing apparatuses, a memory I / F unit 312 connecting the controller 300 and the external memory unit 181, and a system bus 319 connecting each unit.

[0040] Furthermore, the image forming apparatus 100 includes an image forming engine unit 101, which is controlled by an engine control CPU 1012. In addition, the image forming engine unit 101 also includes a first density sensor 200, a second density sensor 500, a timer 201, a counter 202, and the like.

[0041] (Density Prediction Unit) Next, the predicted density calculation unit in the printer controller 300 will be described with reference to FIG. 3. Various signal values from the image density sensor 200, timer 201, counter 202, and the current image formation conditions 203 included in the image forming apparatus 100 are input to the predicted density calculation unit 307 in the printer controller 300. The image formation conditions 203 include the current exposure intensity (hereinafter LPW) and charging potential (hereinafter Vd) in the image forming apparatus 100, etc. Further, it may include the temperature inside the machine, etc. At this time, first, the signal values are input to the input signal value processing unit 320 in the predicted density calculation unit 307. The input signal value processing unit 320 includes a signal value storage unit 321 that stores basic signal values, and a difference calculation unit 322 that calculates the difference between the input signal values and the signal values stored in the signal value storage unit 321.

[0042] The signal values processed by the input signal value processing unit 320 are input to the density prediction unit 330. The density prediction unit 330 includes a density storage unit 331 that stores basic densities, and a prediction function unit 332 that predicts the density from the input values from the input signal processing unit 320. The prediction function unit 332 has an image density prediction model (also referred to as a prediction model) 3321 that calculates the density change amount from the basic density from the input values. The calculated density change amount and the basic density stored in the density storage unit 331 are added together to calculate the current predicted density. The image density prediction model 3321 will be described later. Also, the acquisition of the basic signal values and the acquisition of the basic densities will be described later.

[0043] The calculated predicted density is input to the gradation correction table generation unit 308. The gradation correction table generation unit 308 creates a γLUT for input to the gradation correction unit 316 based on the predicted density. The gradation correction method will be described later.

[0044] (Prediction Model Correction Unit) Next, the prediction model correction unit 350 that corrects the model for calculating the predicted concentration will be described with reference to FIG. 4. The correction of the prediction model is performed, as will be described later, by adding correction data to the data used to create the current model. That is, correction data is added to create a corrected model. Therefore, the model creation data storage unit 351 that stores the data when the current model is created includes a signal value storage unit that stores signal values such as sensors and conditions for model creation, and a concentration value storage unit that stores the corresponding concentration values. Note that the current model refers to the basic model (or initial model) initially created when no correction has been made, and the latest corrected model when correction has been made.

[0045] Also, the model correction data storage unit 352 stores the newly acquired correction data. The model correction data storage unit 352 includes a signal value storage unit that stores signal values, and a concentration value storage unit that stores the concentration values corresponding to the stored signal values.

[0046] Furthermore, the model calculation unit 353 that determines a new model using these data includes a calculation unit that newly creates a model, and a model storage unit that stores the created model. Note that when the model correction is completed, the relationship between the signal value and the concentration value is stored as a data set in the model creation data storage unit. The prediction model correction unit 350 described here can be realized by being included in the image forming apparatus or in a device connected to the image forming apparatus via a network.

[0047] (Basic signal value, basic concentration acquisition) Next, the method for acquiring the basic signal value stored in the signal value storage unit 321 and the basic concentration stored in the concentration storage unit 331, which were described in the concentration prediction unit 330 described above, will be explained. As an example, the basic concentration used in this embodiment is obtained by automatic gradation correction using the output image (toner image after fixing) formed on the paper that is periodically performed as shown in FIG. 5. Note that in this embodiment, a system having a potential sensor that measures the potential on the drum surface will be described, but the present invention is not limited to this.

[0048] (Potential control) When the automatic gradation correction control is started at the user's discretion, first, the potential control process (S201) starts. Before printing on the sheet, the engine control unit CPU1012 determines the target charging potential (VdT), grid bias (Y), and development bias (Vdc) by potential control. The potential control process can determine the charging potential and the like according to the environmental conditions (including temperature and humidity conditions) in which the image forming apparatus 100 is installed. Note that the engine control unit CPU1012 may also be referred to as the engine control unit 1012.

[0049] In the present embodiment, the engine control unit 1012 performs potential control called two-point electric control. FIG. 6 is a diagram for explaining the concept of potential control by two-point electric control. In FIG. 6, the horizontal axis represents the grid bias, and the vertical axis represents the photoreceptor surface potential. VD1 represents the charging potential under the first charging condition (grid bias 400V), and Vl1 represents the potential of the exposed portion formed with the standard laser power. Also, Vd2 represents the charging potential under the second charging condition (grid bias 800V), and Vl2 is the potential of the exposed portion formed with the standard laser power at that time. At this time, the contrast potentials (Cont1, Cont2) at grid biases of 400V and 800V can be calculated from equations (1) and (2).

[0050] (Cont1)=(Vd1―Vl1) ···(1) (Cont2)=(Vd2-Vl2) ···(2) Here, the increase amount (ContΔ) of the contrast potential per 1V of the charging potential can be calculated by equation (3) based on the results of equations (1) and (2).

[0051] (ContΔ)=((Cont2-Cont1) / (Vd2-Vd1))···(3) On the other hand, an environmental sensor (not shown) is provided inside the image forming apparatus 100, and the environmental sensor measures the environmental conditions such as the temperature and humidity inside the image forming apparatus 100. The engine control unit 102 obtains the environmental conditions (for example, absolute moisture content) inside the image forming apparatus 100 based on the measurement results of the environmental sensor. Then, the target contrast potential (ContT) corresponding to the environmental conditions is referred to from the pre-registered environmental table.

[0052] The relationship between the target contrast potential (ContT) and the increase amount of the contrast potential (ContΔ) can be calculated by the formula (4). ContT = Cont1 + X·ContΔ ···(4).

[0053] If the parameter "X" that satisfies the relationship of the formula (4) is calculated, the target charging potential (VdT) (hereinafter, this is also referred to as the "target potential") can be calculated by the formula (5). VdT = Vd1 + X ···(5).

[0054] The change amount of the charging potential per 1V of the grid bias (VdΔ) can be calculated by the formula (6).

[0055] (VdΔ) = (Vd2 - Vd1) / (800 - 400) ···(6).

[0056] The grid bias (Y) that gives the target potential (VdT) can be calculated from the formula (7).

[0057] Target VdT = 400 + Y·VdΔ ···(7).

[0058] In the formula (7), VdΔ can be calculated by the formula (6), and VdT can be calculated from the formula (5). Therefore, by substituting the known potentials from the formulas (5) and (6), the grid bias (Y) that satisfies the relationship of the formula (7) can be finally determined.

[0059] Through the above processing, the target potential (VdT) and the grid bias (Y) according to the environmental conditions can be determined. The development bias (Vdc) has a specified potential difference with respect to the target potential (VdT), and it is possible to calculate by subtracting the specified potential from the determined target potential (VdT).

[0060] Perform subsequent image formation with the determined development bias (Vdc). Although the potential on each drum is negative, the negative sign is omitted here for the sake of clarity in the calculation process.

[0061] Through the above processing, the potential control process in step S201 of FIG. 5 is terminated. (Maximum toner loading adjustment) Next, proceed to step S202, and form a patch image for adjusting the maximum toner loading using the grid bias (Y) determined by the potential control in the previous step S201 and the development bias (Vdc) (S202).

[0062] In printers that prioritize productivity, the following flow is omitted, and a flow for adjusting the maximum loading amount only by potential control is also disclosed. However, since the charge holding amount of the coloring material in the developing device, the mixing ratio of toner and carrier, etc. also change depending on the environment and durability, control only by potential has low accuracy. Therefore, in this embodiment, a patch image with the exposure intensity (hereinafter, LPW) changed in several steps is formed, and the LPW used for normal image formation is determined.

[0063] The image forming apparatus 100, in which the grit bias (Y) and the developing bias (Vdc) are determined, forms five patch images ((1) to (5)) for each of black, cyan, yellow, and magenta, as shown in FIG. 7, in order to adjust the maximum loading amount. Note that the number of patches is not limited to this. The formation conditions of the five patch images are different in LPW, and are LPW1, LPW2, LPW3 (corresponding to the standard laser power used for potential control), LPW4, and LPW5 in order from the left. The laser power increases from LPW1 to LPW5 in order. Also, the number of patch colors may follow the number of color components used in the image forming apparatus 100 and is not limited to four colors.

[0064] The output image is set by the user in the reader unit, and the density of the image pattern is automatically detected (S203). FIG. 8 is a diagram showing the relationship between the density value of each patch image and LPW. By controlling LPW so that the detected density value matches the target density target value (hereinafter, also referred to as the "maximum loading amount target density value"), it is possible to adjust the toner loading amount. (Tone correction and basic value acquisition) When the adjustment of the maximum toner loading amount is completed, next, tone correction is performed. Here, using the grid bias (Y) determined previously, the developing bias (Vdc), and the LPW level, an image pattern of 64 tones for each color is formed and output onto paper (S204). Note that the number of tones is not limited to this.

[0065] The output image is set by the user in the reader unit, and the density of the image pattern is automatically detected (S205).

[0066] Interpolation processing and smoothing processing are performed on the density obtained from the image pattern to obtain the engine γ characteristic in the entire density range. Next, using the obtained engine γ characteristic and the tone target set in advance, a tone correction table for converting the input image signal into an output image signal is created (S206). In the present embodiment, as shown in FIG. 9, an inverse conversion process is performed so as to match the tone target to create a tone correction table.

[0067] When this operation is completed, the density on the paper will match the target density in the entire density range for the gradation target.

[0068] Apply the target LPW determined by the above procedure, and using the gradation correction table, form a toner image pattern including test images (also referred to as measurement images) of multiple gradations for each color component (S207). If the density of the test image is detected using the image density sensor 200 on the intermediate transfer medium (S208), the density value becomes the target density on the intermediate transfer medium and is stored in the density storage unit 331 as the basic density (S209). In this embodiment, after the gradation correction table is created, test images of 10 gradations are formed for each color component, the test images are measured using the image density sensor 200, and the results (for example, measurement values) are stored in the density storage unit 331 as the basic density. The measurement results of the density sensor 200 that vary according to the density of the test image are stored in the density storage unit 331. In this case, the data stored in the density storage unit 331 is the density value of the test image. Note that the density value may be stored together with the density value before or after gradation correction corresponding to the density. However, it is necessary to determine which one it is. Also, if the test images to be formed are determined in advance, the density value for each detected test image without being associated with the density value may be stored. The basic density value is referred to during calibration.

[0069] Also, when performing this automatic gradation correction and obtaining the basic density, the sensor, counter, timer values, and image formation conditions such as grid bias, development bias, and LPW level are stored in the signal value storage unit 321 as the basic signal values (S210). Referring to the basic density, engine γ characteristics, and basic signal values obtained in this way, the gradation correction table (LUT) is updated in the manner described below.

[0070] In the present embodiment, since the image density prediction model is a model for predicting the density of a test image such as a patch on an intermediate transfer body, the basic density value is the density value measured on the intermediate transfer body and saved. However, when, for example, the model is for predicting the density of a test image on a sheet, the density of the test image on the sheet is measured by the reader unit A and saved as the basic density value (basic density). The basic density may be appropriately selected depending on which patch density the image density prediction model deals with, and is not limited to the above. Note that a density sensor provided in the sheet conveyance path may be used instead of the reader unit A.

[0071] (Density correction control) (Outline of control timing of actual measurement control and prediction control) The gradation correction table serving as a basis is created according to the procedure of FIG. 5, and the basic density and basic signal value are saved. The gradation correction table needs to be updated according to changes in color tone and density that occur according to the degree of use of the image forming apparatus. Therefore, in the present embodiment, density correction by actual measurement control and density correction by prediction control are used in combination.

[0072] The density correction sequence by actual measurement control in which a density patch is formed on the intermediate transfer belt and the density patch is read by an image density sensor such as a density sensor generally interrupts the image formation sequence which is a printing operation and is often implemented, which is one of the causes of productivity reduction. On the other hand, performing actual measurement control at a low frequency for fear of productivity reduction leads to deterioration of image quality due to neglect of fluctuations in color tone, density, or any of them. Based on this background, in a conventional image forming apparatus, the control timing of actual measurement control is set in consideration of the balance between color tone / density fluctuations and productivity. Depending on the main body configuration, it is also possible to improve the frequency of actual measurement control by forming a density patch outside the image formation range, but since performing actual measurement control at a high frequency may lead to an increase in toner usage, that is, an increase in cost, it is difficult to increase the frequency of actual measurement control at present.

[0073] However, by implementing predictive control of the density, it becomes possible to compensate for the density correction between the controls of the actual measurement control and suppress the color tone and density fluctuations. For example, while performing density correction by actual measurement control periodically (e.g., at a second frequency), density correction by predictive control is performed at a higher frequency (e.g., at a first frequency) than the density correction by actual measurement control. By doing so, since density correction can be performed at a high frequency, it is possible to further suppress color tone fluctuations. Furthermore, in predictive control, since it does not involve the formation and reading of test images, it does not reduce productivity.

[0074] (Method for Creating (Updating) LUT at the Time of Predictive Density Correction) Next, a method of reflecting the calculated density value in the LUT in predictive control will be described. First, at the time of automatic gradation correction (Fig. 5) arbitrarily performed by the user, a gradation correction table (hereinafter referred to as the basic correction LUT) is formed in accordance with the engine γ characteristic so as to become a preset gradation target (hereinafter referred to as the gradation LUT). Thereafter, the basic density values of 10 gradations for each color described above are acquired. After automatic gradation correction, the input image data is converted by this initial correction LUT and input to the engine, and is output in combination with the engine γ characteristic so as to become the target gradation LUT.

[0075] Thereafter, for example, when the power is turned on, when waking up from sleep, when the environment changes, at a preset timing, etc., when the activation condition of the density correction control is satisfied, the density value is acquired, and the LUT (hereinafter referred to as the composite correction LUT) at the time of image output is created using the acquired density value. The method for creating the composite correction LUT will be described with reference to Figs. 10, 11, 12, and 13. Fig. 10 is a flowchart for creating the composite correction LUT. The process of Fig. 10 is executed, for example, by the printer controller CPU 313. The density curve in the following description is a curve showing the correspondence between the input signal value representing the density and the recorded density value (or predicted density value). The density curve may be realized, for example, by a table associating the input value and the density value. The process of Fig. 10 is executed, for example, at a preset predictive control timing. Specifically, it may be executed every time printing on a predetermined number (or number of pages) of sheets is completed.

[0076] First, obtain the predicted density value of the test image (S301). The acquisition of the predicted density will be described later with reference to FIG. 15. Next, plot the obtained predicted density values for each gradation, and create a density curve (broken line) for the predicted density values shown by the ○ points in FIG. 11 (S302). Perform an inverse transformation on this density curve of the predicted density values in order to correct it to the initial density curve, and create a correction-time LUT as shown by the long broken line in FIG. 12 (S303).

[0077] Here, the initial density curve corresponds to the density curve at the time of basic density acquisition indicated by ● in FIG. 12. This may be realized by a table associating the input signal value with the basic density value stored in the density storage unit. Also, the curves of the initial correction LUTs shown in FIGS. 11 and 13 show the characteristic of correcting the input signal value so that the relationship between the input signal value and the density becomes the initial density curve when an image is formed based on the output signal value obtained by converting the input signal value by the initial correction LUT. On the other hand, the prediction-time LUT shown in FIG. 12 is a LUT for converting the predicted density curve (characteristic) corresponding to the input value into the basic density curve (characteristic).

[0078] Finally, create a combined correction LUT as shown by the long two-dot chain line in FIG. 13 by multiplying (i.e., combining) the prediction-time LUT and the initial correction LUT (S304). The created combined correction LUT is passed to, for example, the gradation correction unit 316 and used for gradation correction. The input signal is converted into an output signal by this combined correction LUT after synthesis, and is output after being reflected in the output image. Note that the method of creating the density curve may be a generally used approximation method such as using an approximation formula connecting 10 points.

[0079] (Predicted Density Calculation) On the other hand, the flow for calculating the predicted density value in S301 is as shown in FIG. 15. Here, in the method of FIG. 5, a flow for predicting the density when the main body is started in a state where the basic signal value and the basic density have been acquired in advance will be described.

[0080] First, when the main body is activated, information such as environmental values at startup, idle time, and the number of toner replenishments, and information on image formation conditions for performing image formation are acquired as input signal values from sensors, timers, and counters provided in the image forming apparatus (S401). The difference between the acquired signal value and a basic signal value stored in advance is extracted (S402).

[0081] Next, the extracted difference value is substituted into an image density prediction model formula created in advance based on considerations (S403), and a difference value from the basic density at the current time is calculated as a predicted value (S404). The predicted density value at the current time is calculated from the sum of this difference predicted value and the basic density value, and the γ characteristic is obtained (S405). Note that the creation process of the image density prediction model will be described later with reference to FIG. 16. (Method for creating LUT during actual measurement density correction) Using FIGS. 19, 20, 21, and 22, a patch image for density correction is created, and a method for creating a composite correction LUT when the density is detected will be described. In this embodiment, a method of sequentially printing and correcting five patch images of 30H, 60H, 90H, C0H, and FFH will be described, but it is not limited to this.

[0082] The patch image is created by multiplying the correction LUT at the current time. After automatic gradation correction, it is created by multiplying a determined gradation for density correction, for example, the 30H pattern, by an initial correction LUT as shown in FIG. 20 obtained during automatic gradation correction (S901, S902). The created pattern is detected by a density detection sensor, and the detection result is plotted as the detection density of 30H (S903). At this time, it is newly plotted at the 30H portion of the initial target density value as shown by the ○ mark in FIG. 21. For the other 60H, 90H, C0H, and FFH, the density target values immediately after creating the initial correction LUT are used. Using the newly plotted 30H measured density value and the five points of the density values 60H, 90H, C0H, and FFH measured initially, a density curve as shown by the long two-dot chain line in FIG. 22 is created (S904). The method for creating this density curve may be a generally used approximation method such as using an approximation formula that connects the five points.

[0083] Next, an inverse transformation is performed to correct the current concentration curve created in S904 to the initial concentration curve, and a sequential correction LUT as shown by the dashed line in FIG. 22 is created (S905).

[0084] Finally, a composite correction LUT as shown by the solid line in FIG. 21, which is obtained by multiplying the sequential correction LUT and the initial correction LUT, is created (S906), and is reflected in the output image and output.

[0085] After reflecting this composite correction LUT, the output image and the tone pattern for image density correction in the next intersheet portion are output with the image being multiplied by this composite correction LUT.

[0086] Thereafter, continuously, a pattern image of another tone is created, density detection is performed, and a composite correction LUT is sequentially created in the same procedure.

[0087] (Normal density calculation) First, in normal image correction control, a flow for acquiring the density value in the current image forming apparatus is shown in FIG. 14. FIG. 14 corresponds to S901 - S903 in FIG. 20. When the main body is activated, information such as environmental values at startup, idle time, and toner replenishment times, and information on image forming conditions for performing image formation are acquired as input signal values from sensors, timers, and counters provided in the image forming apparatus (S501).

[0088] Next, a plurality of toner image patterns are formed under image forming conditions according to the acquired information (S502). In this embodiment, patterns of 10 tones for each color are formed, but it is not limited thereto.

[0089] Next, the formed patch image is subjected to density detection (S503) using the image density sensor 400 on the intermediate transfer body, and the density value (γ characteristic) at the correction time is acquired.

[0090] (Density prediction model creation) The image density prediction model is obtained by formulating it based on experimental results, taking information correlated with density fluctuations in the image (fluctuation correlation information) as input information and image density information as output information. For this reason, this mathematical formula itself may also be referred to as a prediction model, a prediction condition, or a determination condition. The input information includes environmental information that can be obtained from the sensor 200 immediately after the image forming apparatus is powered on or immediately after it resumes. Furthermore, the input information includes time information such as the elapsed time since the previous print obtained from the timer 201, count information such as the number of toner replenishments and the number of idle rotations obtained from the counter 202, the image forming conditions 203 before the image forming apparatus is left unattended, and so on. The density prediction model in the embodiment is a multiple regression model, and in its creation, the partial regression coefficients corresponding to each of the predetermined input information, which are explanatory variables, are determined.

[0091] Hereinafter, the procedure for pre-creating the image density prediction model used in this embodiment will be described with reference to the flowchart of FIG. 16. The procedure of this flowchart may be executed by the prediction model modification unit 350 in software (or firmware). Hardware-wise, it is executed by the printer controller CP313. In this description, for the sake of example, the explanation will proceed using a multiple regression model, but the present invention is not limited to this multiple regression model, and a regression model by other means may also be used. Also, as input values (explanatory variables), some of the variables exemplified may be used, or other variables may be included. The creation of the prediction model may be performed individually for each machine body, or may be performed for the machine bodies serving as samples. In the latter case, a prediction model may be created for the sample machine bodies and applied as an initial prediction model to image forming apparatuses of the same type. Also, there may be a plurality of sample machine bodies. Note that the prediction model created here is an initial prediction model installed in a newly produced image forming apparatus in which parts and the like are not worn out, and is updated in the manner described later.

[0092] First, prepare a large number of variation patterns of environmental conditions and variation patterns of image formation conditions. Print a predetermined test image under these conditions, and measure the environmental conditions and image density (S101). The test image may be an image in which patterns of predetermined densities are arranged in a predetermined layout. The environmental conditions include the toner density in the developing unit during printing (the ratio of toner to carrier in the developing unit), the temperature and humidity at various locations, the toner density in the developing unit during the previous printing, the elapsed time since the previous printing, etc. These are environmental information that can be obtained immediately after power-on. The image formation conditions include the charging potential on the photosensitive drum (hereinafter Vd), the exposure intensity (hereinafter LPW), the development contrast in the developing unit (hereinafter Vcont), etc. The environmental conditions and image formation conditions to be measured may be those selected as explanatory variables. Also, the image density is the density of the test image, and for example, it may be any of the density on the photoreceptor of the toner patch, the density on the intermediate transfer body, or the density on the sheet as the recording medium. In this example, for the creation of the initial prediction model, an example using the density on the intermediate transfer body will be described, but other densities may also be used. Alternatively, a combination of measured values of multiple densities, such as a combination of the density on the intermediate transfer body and the density on the recording medium, may be used.

[0093] Next, using the first measurement data of each experimental day as a reference value, calculate the variation from the reference value for the values of each measurement item of environmental variation, image formation condition change, and image density variation (S102).

[0094] Next, use the measured values of each environmental condition and each image formation condition classified as identification data as input data (explanatory variables), and use the actually measured density measured under each environmental condition and each image formation condition as teacher data to obtain the coefficients of the multiple regression model (S103). The coefficients to be obtained are the coefficients such that the error between the predicted density value and the actually measured density value is minimized when obtaining the predicted density value with each environmental condition and each image formation condition included in the identification data as explanatory variables. In other words, in step S103, curve fitting of the regression model is performed. A detailed explanation of this will be given after the explanation of FIG. 16. In this way, the multiple regression model is created.

[0095] ● Determination of coefficients For example, in the following, the charging potential Vd during printing, the exposure intensity LPW, the toner concentration in the developing device, and the environmental temperature are described as variation correlation information correlated with the variation of the image density as input signal values, but it is not limited thereto. Note that the items included in this variation correlation information may be included in the basic signal values or may be the same. Further, although the description is made up to a four-input linear function model for the above sensor input values, even when five or more sensor inputs and image formation condition inputs are used, a regression model can be created by performing the same processing, and it is not limited thereto.

[0096] Create a linear function model that predicts the image density variation y as the output variable from combinations of four types of input variables. Here, the input variable x n_train is the LPW variation x i(n) during printing, the charging potential variation x 1(n) , the toner concentration variation x 2(n) in the developing device, and the environmental temperature variation x 3(n) . 4(n)

[0097] Four-input model: y n_train = a1 × x 1(n) + a2 × x 2(n) + a3 × x 3(n) + a4 × x 4(n) (i = 1, 2, 3, 4, n = number of data) More generally, y n_train = Σ i a i × x i(n)

[0098] For this input model, the actually measured data of the image density variation, which is the output variable, is used as the teacher data y n_teach , and curve fitting is performed on this value. As an example of the curve fitting method, for the coefficients (a1, a2, a3, a4) of the linear function model, the sum of squares L of the prediction error represented by the following formula between the predicted value and the actually measured value is calculated, and the coefficients that minimize this are derived. ​​The derivation method will be described. First, each variable is represented by a matrix as follows.

[0099]

Number

[0100] Then, as described above, the sum of the squares of the differences between the predicted values and the measured values is obtained.

[0101]

Number

[0102] This expansion formula is the sum of the squares of the prediction errors L, and the purpose is to obtain the matrix a that minimizes this L, that is, the coefficients (a1, a2, a3, a4) of the linear function model. That is, y = y n_teach , y n_train = xa, L = y T y - 2y T xa + a T x T xa Find the coefficient matrix a that gives the minimum value of. Therefore, taking L as the target variable, setting the equation obtained by differentiating L with respect to a to 0 and solving it to derive the optimal coefficients of the regression model. First, find the differential formula.

[0103]

Number

[0104] Then, set this solution to 0.

[0105] ―2y T X + a T (X T X + (X T X) T ) = 0 Then, by performing the formula expansion with a on the left side, a is obtained as follows.

[0106] a = ((X T X) T X T y n_teach ) In this way, by obtaining the coefficient a matrix of the multiple regression model as an example of the image density prediction model, the multiple regression model can be created.

[0107] In this embodiment, the input variables are set as simple ones such as x 1(n) , x 2(n) , x 3(n) , x 4(n) . However, by preparing products or quotients of environmental conditions and image formation conditions such as x 1(n) ×x 2(n) , complex models can also be considered. For example, an input variable that can represent the change in toner charge amount considering the toner density in the developing device and the standing time can be created, and the prediction model can be considered.

[0108] (Modification of the density prediction model) As described above, when performing calibration control for density adjustment using an optimal density prediction model that individually corresponds to the usage environment, output conditions, and usage status, it is necessary to modify the current prediction model. This is because, usually at the initial stage, it is common to use an average model that can cover a certain range of usage environments and situations, and it is not always optimal for individual usage environments.

[0109] To modify the prediction model, data (measurement values) that combines the actual density fluctuations with environmental conditions, image formation conditions, etc. is required. Therefore, usually, control for actually forming a calibration patch and performing density adjustment is used in combination, and data for modifying the prediction model is acquired simultaneously at the timing of executing the control by the patch (actual measurement control).

[0110] Then, the number of data n is increased and added to the following matrix data.

[0111]

Equation

[0112] Then, using each of the additionally updated variables, the coefficient a of the multiple regression model is obtained again in the same manner as the above-described process (i.e., the method shown in FIGS. 16 and equations 1 to 4), thereby creating a new multiple regression model adapted to the changed environment or the like.

[0113] In addition, when correcting the density prediction model, it is also conceivable to correct the model that is actually performing density prediction at any time, or to have a plurality of density prediction models and separately have a model that actually performs density prediction and a density prediction model for which correction is in progress.

[0114] Also, as described above, the data accumulation for correcting the density prediction model and the execution of the calculation for actually obtaining the corrected density prediction model can be realized by performing them in the image forming apparatus or in an apparatus connected to the image forming apparatus via a network. The location of the calculation does not limit the present invention.

[0115] (Judgment of switching to the corrected density prediction model) In the present embodiment, a method of switching between a model that has a plurality of density prediction models and performs density prediction as necessary and a model for which correction is in progress will be described. In the present embodiment, the first model is an average density prediction model A (or also referred to as the first determination condition) that can cover a certain range of usage environments and situations. Further, the second model is a corrected density prediction model B (or also referred to as the second determination condition) that is corrected at any time in response to the results of correction data, and a configuration in which these two models are properly used will be described. That is, in the present embodiment, as the prediction model 3321, there are a prepared stationary prediction model A and a prediction model B that is adaptively corrected.

[0116] The density prediction model A may be a pre-prepared model. For example, in various usage environments, the model created by the above-described procedure may be created and used as the model A. The usage environment may include, for example, the temperature and humidity at various locations inside and outside the machine, the standing time since the previous printing, and the like. The density of the test images formed by varying the values of these environments is measured, and a function or table for obtaining the density with the environmental parameters is created and used as the density prediction model A. Of course, using machine learning, the parameter values may be associated with the measured density values serving as the teacher data, and a learned model may be created and used as the density prediction model A. Also, the model A may be used as the initial model of the density prediction model, and a corrected density prediction model B may be created based on it.

[0117] As described above, at the timing of obtaining the correction data set for correcting the density prediction model B and at the timing of updating or switching the density prediction model to the corrected density prediction model B, changes in internal and external disturbance factors of the image forming apparatus may occur. In such a case, the difference between the predicted density and the measured density may deteriorate more in the corrected second model B than in the average model A which is the first model.

[0118] Therefore, in the present embodiment, before switching the density prediction model to the second corrected model B, the predicted densities of the two models, namely the density prediction model A and the corrected density prediction model B, are each compared with the actually measured density detected by forming and detecting a toner patch. Then, it is determined whether or not to switch to the corrected model B according to the result obtained by the comparison.

[0119] The following will be described with reference to FIG. 17. In the present embodiment, the data acquired at the time of the first calibration is regarded as the first one, and the corrected density prediction model is created for the first time at the timing when the predetermined number of data m = 100 is accumulated. When the density prediction control using the density prediction model starts (S701), the collection of the data set for correcting the prediction model is started. A data set that combines the actual density variation with the environment, output conditions, etc. is accumulated each time the control for actually forming a calibration patch and performing density adjustment is carried out (S702). That is, along with the operation of the image forming apparatus, the variation correlation field information correlated with the density variation and the corresponding measured density value are accumulated. From the first time to the m-th time, calibration may be performed using the density prediction model A.

[0120] At the first correction timing, that is, after starting the collection of a predetermined number m (for example, 100) of data sets, when it is determined that m data sets have been acquired (S703), the corrected density prediction model B is created (S704). Here, the density prediction model is created from m data sets by the model creation method described above.

[0121] After the corrected density prediction model B is created, in order to verify the accuracy of the previously prepared average density prediction model A and the corrected density prediction model B created this time, the accumulation of prediction accuracy data is performed (S705). The prediction accuracy data is the difference between the actual density of the toner patch and the predicted density obtained from the result of inputting information such as the environment and output conditions into the density prediction model. n = 30 differences are obtained for each of the density prediction models A and B, and the process proceeds to accuracy verification (S706 - YES). Here, the sum of the squares of the differences described above is obtained for each density prediction model.

[0122] For example, a test image including 30 patches with different densities can be formed on an intermediate transfer belt, for example, and prediction accuracy data can be obtained by detecting it with the density sensor 200. Alternatively, each time calibration is performed by actual measurement, the predicted value by the prediction model A can be accumulated together with the actually measured value, and this can be repeated until 30 sets of data are accumulated with one set being a combination of an actual measurement value and a predicted value. The test image may be stored in the image forming apparatus in advance. Note that since the test image is formed for each color, the total number of patches becomes 120. Then, the sum of the squares L of the differences between the predicted density and the actually measured density is obtained for each prediction model as follows.

[0123] L(A)=Σ n=1 m (y n_teach - y n_train_A ) 2 L(B)=Σ n=1 m (y n_teach - y n_train_B ) 2 Here, L(A) is the sum of the squares of the differences for the prediction model A, and L(B) is the sum of the squares of the differences for the prediction model B. y n_teach is the actually measured density value for, for example, the nth color (patch). y n_train_A is the predicted density value for, for example, the nth color (patch) by the prediction model A, and y n_train_B is the predicted density value for, for example, the nth color (patch) by the prediction model B. Also, m is set to 30 here.

[0124] If a predetermined number of prediction accuracy data has been accumulated (S706 - YES), the magnitude relationship between L(A) and L(B), which is the sum of the squares of the predicted concentration and the measured concentration, is determined (S707). If the relationship is L(A) > L(B), the concentration prediction model is switched to the corrected prediction model B (S708). That is, if the difference between the predicted concentration and the measured concentration is larger for prediction model A, prediction model B is selected, and prediction model B is used for calibration using future predicted values. Conversely, if the difference between the predicted concentration and the measured concentration is larger for prediction model B, prediction model A is selected, and prediction model A is used for calibration using future predicted values. If the differences are equal, the currently used prediction model can be continuously selected, or the previously determined prediction model can be selected. The switching of the prediction model may be achieved, for example, by switching the set of coefficients used in each model.

[0125] In the verification of prediction accuracy, if the result is that the accuracy of the corrected concentration prediction model with L(A) < L(B) deteriorates, the process returns to the flow of continuing to accumulate the correction data set (S702). Then, the accumulation of data is started again up to a predetermined data set of m = 100. Here, it is also possible to divert the n = 30 prediction accuracy verification data accumulated in the above-mentioned flow S706 as the correction data for S702.

[0126] From the timing of the previous accuracy verification comparison (S707), if m = 100 pieces of data for modifying the model again are newly acquired, the following modified model B is created. For example, the concentration prediction model has been modified and the prediction accuracy verified once before, but no improvement in prediction accuracy was seen at that time. There may also be cases where the modification of the prediction model is the second flow. In such cases, it is also possible to add 100 pieces of the newly acquired second dataset to the first 100 pieces of the dataset. At this time, taking the upper limit of the total dataset as 1000 as an example, for example, if data accumulation of more than that is performed, data that is far from the current data acquisition timing is removed from the modification dataset. That is, starting from the oldest data, the data is removed from the modification dataset until the number of datasets is within the upper limit. In this way, a concentration prediction model corresponding to the more recent image formation state can be created.

[0127] In this way, the method of adding newly acquired data to the data acquired in the past has a drawback that the difference between the predicted density and the actual density is not created according to the data obtained from the most recent engine state. On the other hand, it also has an aspect that it is possible to create a relatively general-purpose model corresponding to long-term and multiple datasets.

[0128] In this way, the concentration prediction model is switched from the average concentration prediction model A with a pre-existing concentration prediction model to the modified concentration prediction model B obtained from the dataset information obtained from an individual image forming apparatus. In such a configuration, after verifying the accuracy of the concentration prediction model A and the concentration prediction model B, the model is switched. By doing this, it is possible to more reliably adopt a model with high prediction accuracy and perform concentration prediction. And by performing this concentration prediction, it is possible to provide an image forming apparatus with higher density stability.

[0129] [Embodiment 2] In Embodiment 1, a method of switching the density prediction model to the corrected density prediction model only once was described. However, since the state of the image forming apparatus is constantly changing, the correction of the density prediction model can be continuously executed to keep up with the state change and continuously execute highly accurate density prediction.

[0130] In this embodiment, a configuration for continuously updating the corrected density prediction model at all times will be described. To realize this configuration, in this example, the average density prediction model A prepared in advance described in Embodiment 1 and two models B1 and B2, which are density prediction models for correcting the model according to the state of the image forming apparatus, will be used for the description. Model B1 may also be referred to as the first determination condition, and model B2 may be referred to as the second determination condition. In this embodiment, for example, as the initial value of the prediction model B1, a prediction model corresponding to the prediction model A of Embodiment 1 may be prepared in advance. In this case, the prediction model A is not updated in Embodiment 1, but in this embodiment, among the two prediction models B1 and B2, the unused model, that is, the non-selected model, is updated. Thereby, even when a specifically deviated prediction model is created (or updated), by adopting a more accurate prediction model, the quality of calibration and thus the quality of the formed image can be maintained.

[0131] In Embodiment 1, a flow of switching the density prediction model used for density correction control from the average density prediction model previously provided in the image forming apparatus to the corrected density prediction model provided in the individual image forming apparatus was described. In this embodiment, after switching to the corrected density prediction model B1, a flow of continuously creating an appropriate optimal corrected density prediction model using a different corrected density prediction model B2 and selecting the corrected density prediction model each time will be described with reference to FIG. 18.

[0132] First, start selecting the correction concentration prediction models B1 and B2 (S801). Then, first, determine the concentration prediction model currently used for concentration correction control (S802). If the prediction model being used is not model B1, it is determined that the prediction model being used is model B2 (S802-No), and the process branches to S812. If it is determined in S802 that model B1 is being used (S802-YES), the process branches to S803. Hereinafter, the case where the concentration prediction model B1 is currently being used will be described.

[0133] In this case, start creating the concentration prediction model B2 (S803), and accumulate a dataset of the teacher data, which is the concentration information of the actually measured toner patches, the input data combined with the environment and output conditions when forming the image, and the predicted concentration data by model B2. Control for actually forming calibration patches and performing concentration adjustment is used in combination, and at the timing of executing the control by the patches, data for correcting the prediction model is acquired simultaneously.

[0134] Then, when the number of pieces of data for model correction to be newly added reaches m = 100 from the timing immediately before the accuracy verification of the prediction model (S805), update the concentration prediction model B2 (S806). Even if it is called an update, if the concentration prediction model B2 does not yet exist at this time, it will be newly created.

[0135] At this time, although the newly acquired dataset has m = 100, the number of data used for predicting model correction may not only be the newly acquired m = 100 data, but also the data for model correction acquired previously. However, since it is difficult to implement storing data with an infinite increase in the number of data, as an example, when exceeding a certain upper limit, it is desirable to configure to replace old data acquired at a timing far from the current time with newly acquired data. For example, 1000 data may be set as the upper limit of the total accumulated data for model correction. As a result, it becomes possible to correct the model while securing the number of data necessary for obtaining a stable model including the state of the latest image forming apparatus. Further, the data accumulated for this model correction can be shared as data when updating the density prediction models B1 and B2.

[0136] After the density prediction model B2 is updated, data accumulation for verifying the prediction accuracy of the density prediction model B1 currently used for density correction control and the density prediction model B2 created this time is started (S807).

[0137] The prediction accuracy data is the same as the content described in the first embodiment. In this embodiment, when n = 30 are obtained for each of the density prediction model B1 and the density prediction model B2, the process proceeds to accuracy verification (S808). Here, the following two types of sums of squares are obtained for each density prediction model. This may be obtained by simply replacing the prediction model A and the prediction model B in the first embodiment with the density prediction model B1 and the density prediction model B2, respectively.

[0138] L(B1)=Σ n=1 m (y n_teach - y n_train_B1 ) 2 L(B2)=Σ n=1 m (y n_teach - y n_train_B2 ) 2 Here, L(B1) is the sum of squares of differences for the prediction model B1, and L(B2) is the sum of squares of differences for the prediction model B2. yn_teach is the measured value of the density for, for example, the nth color (patch). y n_train_B1 is the predicted value of the density for, for example, the nth color (patch) by the prediction model B1, y n_train_B2 is the predicted value of the density for, for example, the nth color (patch) by the prediction model B2. Also, m is set to 30 here.

[0139] If a predetermined number of prediction accuracy data has been accumulated (S706 - YES), the magnitude relationship between L(B1) and L(B2), which are the sum of squares of the predicted density and the measured density respectively, is determined (S809). If the relationship is L(B1)>L(B2), the density prediction model is switched to the corrected prediction model B2 (S810). Then, the correction of the current density prediction model is completed, and the correction of the next density prediction model is started (S811). In this explanation, the density prediction model B2 was corrected, and since the density prediction accuracy was improved, the model used for density correction control was switched to the density prediction model B2. However, if no improvement in prediction accuracy is seen in the current model correction in S809, the collection of the input / output data and the teacher data dataset is continued, and the flow for improving the model is advanced. (S804~S809).

[0140] Also, in this case, the prediction model B1 is used for density correction control, and the flow for updating the density prediction model B2 (S803~S810) has been explained. Another case, where the prediction model B2 is used for density correction control and the flow for updating the prediction model B1 (S812~S819) can be explained with the same content.

[0141] In this way, in order to correct the density prediction model, two models, the corrected prediction model B1 and the prediction model B2, are provided, and the model is corrected at any time in parallel with the model used for density correction control. By doing so, the density prediction model can be corrected with data including the latest state of the image forming apparatus. Then, after verifying the accuracy of the updated density prediction model and switching the model, it becomes possible to adopt a model with higher prediction accuracy more reliably and perform density prediction, and it becomes possible to provide an image forming apparatus with higher density stability.

[0142] [Other Embodiments] The present invention can also be realized by supplying a program that realizes one or more functions of the above-described embodiments to a system or apparatus via a network or a storage medium, and causing one or more processors in a computer of the system or apparatus to read and execute the program. It can also be realized by a circuit (for example, ASIC) that realizes one or more functions.

[0143] The invention is not limited to the above-described embodiments, and various changes and modifications can be made without departing from the spirit and scope of the invention. Therefore, the claims are attached to disclose the scope of the invention.

Description of Reference Numerals

[0144] 1: Photoconductor drum, 2: Charging device, 3: Exposure device, 4: Developing device, 6: Transfer device, 200: Intermediate transfer image density sensor, 300: Printer controller

Claims

1. Image forming means for forming an image based on image forming conditions, An intermediate transfer body onto which the image formed by the image forming means is transferred, Transfer means for transferring the image on the intermediate transfer body to a sheet, Measuring means for measuring a measurement image on the intermediate transfer body, Obtaining means for obtaining variation correlation information correlated with variation in the density of the image formed by the image forming means, Generating means for generating the image forming conditions from the variation correlation information obtained by the obtaining means based on a first determination condition, Determining means for determining a first density of a first measurement image formed by the image forming means from the variation correlation information obtained by the obtaining means based on the first determination condition, and determining a second density of a second measurement image formed by the image forming means from the variation correlation information obtained by the obtaining means based on a second determination condition different from the first determination condition, Control means for controlling whether to change the first determination condition used by the generating means to generate the image forming conditions to the second determination condition based on the first density determined by the determining means, the measurement result of the first measurement image measured by the measuring means, the second density determined by the determining means, and the measurement result of the second measurement image measured by the measuring means. An image forming apparatus characterized by the above.

2. If the difference between the second density and the density corresponding to the measurement result of the second measurement image measured by the measuring means is smaller than the difference between the first density and the density corresponding to the measurement result of the first measurement image measured by the measuring means, the control means changes the first determination condition used by the generating means to generate the image forming conditions to the second determination condition. The image forming apparatus according to claim 1, characterized by the above.

3. Further comprising a sensor for measuring temperature, The image forming apparatus according to claim 1 or 2, characterized in that the variation correlation information includes information on the temperature measured by the sensor.

4. Further comprising a sensor for measuring humidity, The image forming apparatus according to any one of claims 1 to 3, characterized in that the variation correlation information includes information on the humidity measured by the sensor.

5. The image forming means includes a photoreceptor, a charging means for charging the photoreceptor, an exposure means for exposing the photoreceptor charged by the charging means to form an electrostatic latent image, and a developing means for developing the electrostatic latent image using toner. The image forming apparatus according to any one of claims 1 to 4, wherein the variation correlation information includes information on the charging potential of the photoreceptor charged by the charging means.

6. The image forming means includes a photoreceptor, a charging means for charging the photoreceptor, an exposure means for exposing the photoreceptor charged by the charging means to form an electrostatic latent image, and a developing means for developing the electrostatic latent image using toner. The image forming apparatus according to any one of claims 1 to 5, wherein the variation correlation information includes information on the exposure intensity of the exposure means when the exposure means exposes the photoreceptor.

7. The image forming means includes a photoreceptor, a charging means for charging the photoreceptor, an exposure means for exposing the photoreceptor charged by the charging means to form an electrostatic latent image, and a developing means for developing the electrostatic latent image using toner. The image forming apparatus according to any one of claims 1 to 6, wherein the variation correlation information includes information on the toner concentration of the developing means.

Images