Sheet binding device, sheet binding system and image forming system

Abstract

To provide a sheet binding device, a sheet binding system and an image forming system capable of obtaining a binding force desired by each user in binding a sheet bundle.SOLUTION: A sheet binding device 3 includes binding means 25 that binds a sheet bundle using a binding force obtained by applying pressure between the sheets, and control means 301 that controls the binding process of the binding means, and includes information acquisition means 31 that acquires physical property information of the sheets included in the sheet bundle to be bound and information on the number of sheets in the sheet bundle, estimation means 301 that estimates a binding force evaluation value after the binding process for the sheet bundle to be processed based on the acquired sheet physical property information and sheet number information using a trained model trained using multiple learning data including sheet physical property information of learning sheets, sheet number information of the learning sheet bundle, and a binding force evaluation value after the binding process for the learning sheet bundle, and determination means 301 that determines a control content of the control means in the binding process for the sheet bundle based on the estimated binding force evaluation value.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to a sheet binding device, a sheet binding system, and an image forming system. [Background technology]

[0002] Conventionally, a sheet binding device has been known that includes a binding means that binds a stack of sheets to be processed that includes multiple sheets to be processed by applying pressure between the multiple sheets to be processed using a bonding force, and a control means that controls the binding process of the binding means.

[0003] For example, Patent Document 1 discloses a sheet binding device that performs a binding process by applying pressure to and deforming a portion of a sheet bundle of multiple overlapping sheets into an uneven shape. This sheet binding device has pre-registered experimental data, such as the relationship between the thickness of the sheets and the binding pressure generated at the binding section under initial binding conditions (contents of the initial binding process) and the relationship between the number of sheets and the binding pressure. Then, before the final sheet to be bound reaches the binding section, the input sheet thickness and number of sheets are compared with the registered experimental data to determine the binding pressure for the binding process, and the number of binding operations for the binding process is set according to the determined binding pressure. Summary of the Invention [Problem to be solved by the invention]

[0004] In a sheet binding device that binds a stack of sheets to be processed using a binding force obtained by applying pressure between the sheets to be processed without using staples or other binding components, the binding process content required to achieve the desired binding force varies depending on conditions such as the physical properties of the sheets to be processed and the number of sheets to be bound. Furthermore, there are an enormous number of combinations of conditions, such as the physical properties of the sheets used by users and the number of sheets to be bound. Therefore, it has been difficult for each user to appropriately select the binding process content and achieve the desired binding force under their usage conditions. [Means for solving the problem]

[0005] In order to solve the above-mentioned problems, the present invention provides a sheet binding device comprising: a binding means for binding a processing target sheet bundle including a plurality of processing target sheets by using a binding force obtained by applying pressure between the plurality of processing target sheets; and a control means for controlling the binding process of the binding means, and is characterized by having: an information acquisition means for acquiring physical property information of the processing target sheets included in the processing target sheet bundle to be bound by the binding means and sheet number information of the processing target sheet bundle before the binding process is completed; an estimation means for estimating a binding force evaluation value after the binding process for the processing target sheet bundle based on the sheet physical property information and sheet number information acquired by the information acquisition means by having a computer execute a trained model trained using a plurality of training data including sheet physical property information of learning sheets, sheet number information of the learning sheet bundle including the learning sheets, and a binding force evaluation value after the binding process for the learning sheet bundle; and a determination means for determining the control content of the control means in the binding process for the processing target sheet bundle based on the binding force evaluation value estimated by the estimation means. [Effects of the Invention]

[0006] According to the present invention, it is possible to appropriately select the content of the binding process for a sheet stack under various usage conditions of each user, and obtain a desired binding strength. [Brief explanation of the drawings]

[0007] [Figure 1] 1 is a schematic diagram illustrating the overall configuration of an image forming system including a sheet binding device according to an embodiment. [Figure 2] FIG. 2 is an explanatory diagram for explaining the configuration of an image forming apparatus and a post-processing apparatus that constitute the image forming system. [Figure 3] 10 is an explanatory diagram showing an example in which a sheet physical property value sensor for detecting sheet physical property information is provided inside the post-processing device. FIG. [Figure 4] FIG. 10 is an explanatory diagram showing an example in which a sheet physical property value sensor for detecting sheet physical property information is provided in the image forming apparatus. [Figure 5]10 is a flowchart showing the flow of the learning phase of the trained model executed by the control unit in the post-processing device. [Figure 6] FIG. 4 is an explanatory diagram showing an example of a pressure binding processing section of the post-processing device. [Figure 7] FIG. 10 is an explanatory diagram showing another example of the pressure binding processing section of the post-processing device. [Figure 8] FIG. 10 is an explanatory diagram showing a connection point between a first arm portion and a second arm portion in a pressure binding processing portion according to another example of the same. [Figure 9] FIG. 10 is a perspective view schematically showing three cams present in the pressure binding processing section. [Figure 10] 1A is an explanatory diagram showing the operation of the divided arm portion in the pressure binding processing section when the arm fixing pin is not positioned, and FIG. 1B is an explanatory diagram showing the operation of the divided arm portion in the pressure binding processing section when the arm fixing pin is positioned. [Figure 11] FIG. 10 is an explanatory diagram showing an example of a post-processing device that has a configuration for acquiring characteristic data of the surrounding environment detected by an environmental sensor in addition to sheet physical property information. [Figure 12] An explanatory diagram showing an example of the relationship between the standard value, the estimated value of the binding strength output by the trained model, and the binding strength after adjustment. [Figure 13] FIG. 1 is an explanatory diagram illustrating an example of a post-processing device that can communicate with a cloud system via a network. [Figure 14] An explanatory diagram showing an example in which a trained model is implemented in a processing unit of a cloud system with which a post-processing device can communicate via a network. [Figure 15] FIG. 10 is an explanatory diagram showing an example in which sheet physical property information, sheet number information, etc. included in learning data are collected by a cloud system from a plurality of post-processing devices. [Figure 16] 10 is a flowchart showing an example of the flow of a pressure binding process in the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0008] Hereinafter, an embodiment of an image forming system using a sheet binding device according to the present invention as a post-processing device of an image forming apparatus will be described. The sheet binding device of the present invention is an example of a post-processing device that binds a stack of sheets, with sheets on which images have been formed by an image forming device as the processing target sheets, but it is applicable to binding stacks of any type of sheets, not just sheets on which images have been formed.

[0009] FIG. 1 is a schematic diagram showing the overall configuration of an image forming system equipped with a sheet binding device according to this embodiment. The image forming system 1 of this embodiment comprises an image forming apparatus 2 and a post-processing apparatus 3 as a sheet binding apparatus. In this image forming system 1, the post-processing apparatus 3 is connected to the rear stage of the image forming apparatus 2. The post-processing apparatus 3 receives sheets on which images have been formed from the image forming apparatus 2 and performs a predetermined binding process.

[0010] FIG. 2 is an explanatory diagram for explaining the configuration of the image forming apparatus 2 and post-processing apparatus 3 that constitute the image forming system 1 of this embodiment. The image forming apparatus 2 is an apparatus that forms an image on a sheet by a known electrophotographic process. The image forming apparatus 2 of this embodiment mainly includes a display unit 201, an operation unit 202, a paper feed unit 203, an image forming unit 204, a fixing unit 205, and a control unit 206.

[0011] The display unit 201 notifies the user of the status of various devices and operation details. The operation unit 202 is a user interface that allows the user to perform setting operations such as the operating mode and the number of copies to be printed. The paper feed unit 203 stocks a large number of sheets and separates and feeds the sheets one by one. The image creation unit 204 forms a latent image on the surface of a photosensitive member serving as a latent image carrier in accordance with input image data, and transfers a toner image (image) obtained by developing the latent image onto a sheet fed from the paper feed unit 203. The fixing unit 205 fixes the toner image onto the sheet onto which the toner image has been transferred by the image creation unit 204. The control unit 206 controls each unit of the image forming apparatus 2.

[0012] The image forming method in the image forming apparatus 2 is not limited to the electrophotographic method, but may be any known image forming method such as an inkjet method or a thermal transfer method.

[0013] The post-processing device 3 performs post-processing on the sheets P on which images have been formed by the image forming device 2. The post-processing performed by the post-processing device 3 includes a binding process, which is a "pressure binding process" that binds a stack of multiple sheets P on which images have been formed (sheet stack Pb) without using staples (binding components). Note that the post-processing performed by the post-processing device 3 also includes another post-processing, a "staple binding process" that binds a stack of multiple sheets P on which images have been formed (sheet stack Pb) using staples.

[0014] More specifically, the pressure binding process according to this embodiment is a process of applying pressure to the binding portion of the sheet bundle Pb to deform (pressure-deform) the binding portion and bind the sheet bundle Pb, which is called pressure binding. Note that the binding processes (pressure binding process and staple binding process) that can be performed by the post-processing device 3 include an end binding process that binds the end portion of the sheet bundle Pb and a saddle binding process that binds the center portion of the sheet bundle Pb.

[0015] The post-processing device 3 includes conveying roller pairs 10-19 (conveying section) and a switching claw 20. The conveying roller pairs 10-19 convey the sheet P supplied from the image forming device 2 inside the post-processing device 3. More specifically, the conveying roller pairs 10-13 convey the sheet P along a first conveying path Ph1. Furthermore, the conveying roller pairs 14-15 convey the sheet P along a second conveying path Ph2. Furthermore, the conveying roller pairs 16-19 convey the sheet P along a third conveying path Ph3.

[0016] The first conveying path Ph1 is a path that leads from a supply port of the sheet P from the image forming device 2 to the first discharge tray 21. The second conveying path Ph2 is a path that branches off from the first conveying path Ph1 between the pair of conveying rollers 11 and 14 in the conveying direction, and leads to the second discharge tray 26 through the internal tray 22 (placing section). The third conveying path Ph3 is a path that branches off from the first conveying path Ph1 between the pair of conveying rollers 11 and 14 in the conveying direction, and leads to the third discharge tray 30.

[0017] The switching claw 20 is disposed at a branching position of the first conveying path Ph1 and the second conveying path Ph2. The switching claw 20 is configured to be switchable between a first position where the sheet P is discharged onto the first discharge tray 21 through the first conveying path Ph1 and a second position where the sheet P conveyed along the first conveying path Ph1 is guided to the second conveying path Ph2.

[0018] Furthermore, when the trailing edge of the sheet P that has entered the second conveying path Ph2 passes through the pair of conveying rollers 11, the pair of conveying rollers 14 is rotated in the reverse direction, so that the sheet P is guided to the third conveying path Ph3. The post-processing device 3 also includes a plurality of sensors (shown as black triangles in FIG. 2) that detect the position of the sheet P on each of the conveying paths Ph1, Ph2, and Ph3.

[0019] The post-processing device 3 includes a first discharge tray 21. The sheets P discharged through the first conveying path Ph1 are placed on the first discharge tray 21. Of the sheets P supplied from the image forming device 2, those sheets P that are not to be bound are discharged to the first discharge tray 21.

[0020] The post-processing device 3 also includes an internal tray 22 as a loading tray, an end fence 23, side fences 24L and 24R, a pressure binding processing unit 25, a staple binding processing unit 155, and a second discharge tray 26. The internal tray 22, the end fence 23, the side fences 24L and 24R, the pressure binding processing unit 25, and the staple binding processing unit 155 perform edge binding processing on a sheet bundle Pb made up of a plurality of sheets P transported through the second transport path Ph2. The sheet bundle Pb that has been edge-stitched is discharged to the second discharge tray 26 from among the sheets P supplied from the image forming device 2.

[0021] The "edge binding process" referred to here includes "parallel binding process" in which binding process is performed along one side of the sheet bundle Pb that is parallel to the main scanning direction, "diagonal binding process" in which binding process is performed at a corner of the sheet bundle Pb, and "vertical binding process" in which binding process is performed along one side of the sheet bundle Pb that is parallel to the conveying direction.

[0022] In this embodiment, the direction from the conveying roller pair 15 toward the end fence 23 is defined as the "conveying direction" of the sheet P. That is, the "conveying direction" in this specification corresponds to the direction in which the sheet P, discharged from the image forming apparatus 2, moves toward the second discharge tray 26 by the conveying roller pair 10 and the like, and then moves toward the end fence 23 by the conveying roller pair 15. In addition, the direction perpendicular to the conveying direction is defined as the "main scanning direction (width direction of the sheet P)."

[0023] The multiple sheets P conveyed in order via the second conveying path Ph2 are temporarily placed on the internal tray 22. The end fence 23 aligns the position in the conveying direction of the sheets P or the sheet stack Pb placed on the internal tray 22. The side fences 24L and 24R align the position in the main scanning direction of the sheets P or the sheet stack Pb placed on the internal tray 22.

[0024] The pressure binding processing unit 25 and the staple binding processing unit 155 perform edge binding processing on the end of the sheet bundle Pb aligned by the end fence 23 and the side fences 24L, 24R. Then, the conveying roller pair 15 discharges the sheet bundle Pb that has been edge-stitched onto the second discharge tray 26.

[0025] The post-processing device 3 further includes an end fence 27, a saddle stitching section 28, a paper folding blade 29, and a third discharge tray 30. The end fence 27, the saddle stitching section 28, and the paper folding blade 29 perform saddle stitching on a sheet bundle Pb made up of a plurality of sheets P transported through the third transport path Ph3. The sheet bundle Pb that has been saddle stitched is discharged to the third discharge tray 30 from among the sheets P supplied from the image forming device 2.

[0026] The end fence 27 aligns the positions in the conveying direction of the multiple sheets P conveyed in order through the third conveying path Ph3. The end fence 27 is also configured to be movable between a binding position where the center of the sheet stack Pb faces the saddle stitching processing unit 28, and a folding position where the center faces the paper folding blade 29. The saddle stitching processing unit 28 stitches the center of the sheet stack Pb aligned by the end fence 27 at the binding position. The paper folding blade 29 folds the sheet stack Pb placed on the end fence 27 at the folding position in half and causes it to be sandwiched between the conveying roller pair 18. The conveying roller pair 18, 19 discharges the sheet stack Pb that has been saddle stitched onto the third discharge tray 30.

[0027] The post-processing device 3 is equipped with a control unit 301 that controls each unit of the post-processing device 3. The control unit 301 functions as a determination unit that determines control parameters for the pressure binding processing unit 25 in the pressure binding process, which is a stapleless binding process. The control unit 301 is also capable of communicating with the control unit 206 of the image forming device 2.

[0028] Next, the control of the pressure binding process in this embodiment will be described. In this embodiment, in a binding process (such as a pressure binding process) that binds a sheet stack Pb using a binding force obtained by applying pressure between the sheets P without using staples, the binding process content required to obtain a desired binding force varies depending on conditions such as differences in the physical properties of the sheets P and the number of sheets to be bound. For example, in the pressure binding process of this embodiment, the sheet stack Pb is compressed under pressure, causing loosened sheet fibers to become entangled, thereby obtaining a binding force. In such a pressure binding process, even when the pressure binding processing unit 25 performs the same binding process content using the same control parameters (pressure force, pressure time, number of pressure applications, etc.), the resulting binding force varies depending on the physical properties of the sheets P. Furthermore, even when the same binding process content is performed, the resulting binding force varies because the pressure applied between the sheets P varies depending on the number of sheets to be bound.

[0029] Furthermore, the obtained binding force does not simply increase by increasing the pressure, pressure time, number of times of pressure application, etc. of the sheets P (by increasing the binding force for binding the sheet stack). If the pressure, pressure time, number of times of pressure application, etc. of the sheets P (binding force) are increased too much, the sheets P will be damaged, and the binding force will actually decrease. Note that the "binding force" referred to here is an index value that increases or decreases depending on the pressure, pressure time, number of times of pressure application (number of binding operations), etc. applied to the sheet stack Pb by the pressure binding processing unit 25, and refers to the overall force acting on the sheet stack Pb from the pressure binding processing unit 25 during the pressure binding process.

[0030] Generally, the conditions for the crimp binding process, such as the physical properties of the sheets P and the number of sheets, vary widely depending on the user's usage situation, resulting in a vast number of combinations of conditions. Furthermore, some users use special types of sheets that are not widely available on the market. Therefore, it is practically difficult to determine the appropriate binding process (control parameters) by conducting tests in advance for every possible combination. Therefore, conventional methods of appropriately selecting binding process contents according to the appropriate binding process contents for each combination determined in advance through tests cannot accommodate the various usage situations of each user, and there are cases where the binding process contents cannot be appropriately selected. If the binding process contents cannot be appropriately selected, the desired binding strength may not be achieved, excessive binding force may be applied, causing damage to the sheets, or excessive binding may result in reduced productivity and wasteful power consumption.

[0031] Therefore, in this embodiment, physical property information of sheet P, which is a target sheet for the pressure binding process, and information on the number of sheets in a sheet bundle Pb including sheet P are acquired in advance, and a binding strength evaluation value after the pressure binding process for the sheet bundle Pb is estimated in advance based on this information. Then, control parameters (control details of the control unit 301) for the pressure binding process for the sheet bundle Pb are determined based on the previously estimated binding strength evaluation value, and the pressure binding processing unit 25 for the sheet bundle Pb is controlled using the determined control parameters. In this way, a binding strength evaluation value related to the binding strength of the sheet bundle Pb obtained by the pressure binding process is estimated in advance, and it is possible to appropriately adjust the details of the pressure binding process for the sheet bundle Pb in accordance with the binding strength evaluation value.

[0032] Furthermore, in the present embodiment, a trained model created by machine learning or the like is used as an estimation program for the advance estimation of the bonding strength evaluation value. Specifically, a trained model is used that has been trained using a plurality of pieces of training data including sheet physical property information of the learning sheet, sheet count information of the learning sheet bundle including the learning sheet, and a bonding strength evaluation value for the learning sheet bundle.

[0033] In the stage of creating the trained model of this embodiment (learning phase), a pressure binding process is performed in advance using various learning sheets with predetermined binding process content (control parameters fixed by initial setting), and a binding strength evaluation value for each binding strength is obtained. Then, machine learning or the like is performed using a learning dataset that collects multiple (large) learning data sets containing the respective sheet physical property information, sheet number information, and binding strength evaluation values, to obtain a trained model. The trained model obtained in this way is an estimation program that, when a pressure binding process is performed on a sheet P (inference phase), uses the sheet physical property information and sheet number information for the sheet P as input data and outputs the desired binding strength evaluation value after the pressure binding process.

[0034] According to the trained model of this embodiment, when sheets P are actually subjected to a pressure binding process (inference phase), even if the physical properties or number of sheets P do not match the combination of sheet physical property information and sheet number information used as training data, it is possible to accurately estimate a binding strength evaluation value related to the binding strength after the pressure binding process. Therefore, even if the physical properties and number of sheets P to be subjected to the pressure binding process vary widely depending on the user's usage situation, it is possible to predict the binding strength after the pressure binding process under each user's usage situation and appropriately adjust the binding process content. Therefore, it is possible to appropriately select the binding process content for the sheet stack Pb under each user's usage situation and obtain the desired binding strength without performing excessive binding process content (without damaging the sheets or causing a decrease in productivity or energy saving).

[0035] Here, in order for the control unit 301 to execute a trained model (estimation program) to pre-estimate the binding strength evaluation value after the pressure binding process in the pressure binding processing unit 25, it is necessary to obtain sheet property information and sheet number information in advance.

[0036] One method for acquiring the sheet physical property information and sheet number information is to acquire them from the user's operation details input via the operation unit 202 of the image forming apparatus 2, or from information included in an output instruction from an externally connected PC (personal computer). Another method is to provide a sensor for detecting the sheet physical property information or sheet number information, and acquire the sheet physical property information or sheet number information from the detection results of the sensor.

[0037] FIG. 3 is an explanatory diagram showing an example in which a sheet physical property sensor 31 for detecting sheet physical property information is provided inside the post-processing device 3. As shown in FIG. 3, the sheet physical property sensor 31 is provided on the conveying path immediately after the pair of conveying rollers 10 inside the post-processing device 3. However, the detection position of the sheet physical property sensor 31 may be any position upstream of the pressure binding processing unit 25 in the conveying direction.

[0038] Examples of the sheet physical property sensor 31 include a sensor that measures the thickness of the sheet, a sensor that measures the density of the sheet, a sensor that measures the surface smoothness of the sheet, and a sensor that measures the strength of the sheet. Sensors that measure the thickness of the sheet include ultrasonic sensors and laser sensors. Sensors that measure the density of the sheet include X-ray densitometers and gamma ray densitometers. Sensors that measure the surface smoothness of the sheet include surface profilers and laser triangulation sensors. Sensors that measure the strength of the sheet include stress sensors and strain gauges.

[0039] FIG. 4 is an explanatory diagram showing an example in which the image forming apparatus 2 is provided with a sheet physical property sensor 31 that detects sheet physical property information. It is also possible to employ a method in which a sensor provided in the image forming apparatus 2 is used, and the control unit 301 of the post-processing device 3 acquires the detection result of the sensor via communication from the control unit 206 of the image forming apparatus 2. In this case, the post-processing device 3 does not need to be provided with a sensor for detecting sheet physical property information, which simplifies the configuration of the post-processing device 3. In particular, since the image forming apparatus 2 is often provided with various sensors for detecting sheet physical property information for the image forming operation, using the sensors of the image forming apparatus 2 also leads to simplification of the entire system.

[0040] FIG. 5 is a flowchart showing the flow of the learning phase of the trained model executed by the CPU of the control unit 301 in the post-processing device 3. The trained model of this embodiment can be created by supervised learning (machine learning), for example, using an external PC or cloud service capable of generating trained models. The created trained model is a type of computational algorithm, and is implemented, for example, as a module as part of a control program (estimation program) executed by a CPU.

[0041] As an example of supervised learning, learning sheets with various physical properties (paper thickness, density, strength, etc.) are used, and pressure binding processing is performed for various numbers of sheets with predetermined binding processing contents (control parameters fixed by initial setting), and the binding strength of the bound sheet stack is measured. As a result, a data set of teacher data is created (S1) in which the binding strength (binding strength evaluation value) when pressure binding processing is performed for various numbers of sheets using learning sheets with various physical properties is used as the correct answer data.

[0042] Although the present embodiment is an example of creating a trained model by machine learning using supervised learning, the trained model may also be created by adopting other machine learning methods such as unsupervised learning or reinforcement learning. Furthermore, in the present embodiment, the value of the bonding force itself after the crimp binding process is used as the bonding force evaluation value, but other index values ​​related to the bonding force (for example, binding force, etc.) may also be used as the bonding force evaluation value.

[0043] Although the training data collected in processing step S1 may be used for machine learning as is, in this embodiment, feature engineering processing is added (S2). This feature engineering is processing that extracts or generates features from the collected training data in a format that can be understood by a machine learning algorithm, or extracts or generates indicators that indicate the statistical characteristics or relative importance of the physical properties of the sheet, and adds them to the training data.

[0044] After that, model training (learning) is performed to generate a trained model from the processed training data obtained after the feature engineering process (S3). During this training (learning), for example, patterns are extracted from the training data, and learning is performed to determine which sheet physical property values ​​correspond to which degree of binding strength.

[0045] Once a trained model has been generated by model training, the trained model is implemented in the control unit 301 of the post-processing device 3 and operated (inference phase). The trained model implemented in the control unit 301 of the post-processing device 3 uses, as input data, sheet property information and sheet count information of the sheets P acquired by the post-processing device 3, and outputs, as output data, a binding strength evaluation value (estimated value of binding strength) after the pressure binding process that is to be estimated. Then, based on the estimated value of binding strength output by the trained model in this way, the control unit 301 of the post-processing device 3 determines control parameters (pressure force, pressure time, number of pressure applications, etc.) of the pressure binding processing unit 25 so as to obtain a target binding strength.

[0046] Next, the configuration and operation of the pressure binding processing section 25 will be described. The pressure binding processing unit 25 of this embodiment has a function that can change at least one parameter from among the pressure force, pressure time, and number of pressure applications (number of binding operations) instructed by the control unit 301. By changing these parameters, it becomes possible to adjust the binding force by the pressure binding processing unit 25, and it becomes possible to perform the binding operation with an optimal binding force.

[0047] FIG. 6 is an explanatory diagram showing an example of the pressure binding processing section 25 in this embodiment. The crimp binding processing unit 25 is configured to support each component via a base 109, with upper concave-convex teeth 102 arranged on the upper tooth arm 104 and lower concave-convex teeth 101 arranged on the lower tooth arm 103. The upper tooth arm 104 is configured to be rotatable around an arm rotation center 104a. In the crimp binding processing unit 25, a motor 110 serving as a drive source rotates a gear 105, causing a cam drive gear 106 to rotate, which in turn rotates a cam 107 fixed thereto, thereby pushing up the cam abutment surface of the upper tooth arm 104. As a result, the upper concave-convex teeth 102 of the upper tooth arm 104 move downward and mesh with the lower concave-convex teeth 101 of the lower tooth arm 103. The crimp binding processing unit 25 also includes a return spring 108 for returning the upper and lower concave-convex teeth 101, 102 to a state spaced apart from each other.

[0048] In the compression binding processing unit 25 of this embodiment, the pressure, pressure time, and number of pressurizations of the sheet P can be adjusted by controlling the rotational movement of the cam 107. Specifically, by controlling the rotational position of the cam 107, the meshing amount of the upper and lower uneven teeth 101, 102 can be adjusted, and a sheet pressure corresponding to the meshing amount can be applied to the sheet P. Furthermore, by controlling the rotational movement timing and rotational speed of the cam 107, the meshing time (meshing speed, meshing start timing, meshing end timing, etc.) of the upper and lower uneven teeth 101, 102 can be adjusted, and the sheet pressurization time corresponding to the meshing time can be adjusted. Furthermore, by controlling the number of rotations of the cam 107, the number of meshings of the upper and lower uneven teeth 101, 102 can be adjusted, and the number of pressurizations of the sheet can be adjusted.

[0049] 7 to 10 are explanatory diagrams showing other examples of the pressure binding processing section 25 in this embodiment. In this example, three cams (cam 107a, cam 107b, cam 107c) are provided that push up the upper teeth arm 104 for different times when the cams rotate. The upper teeth arm 104 is divided into two arm parts 104A and 104B that can be opened and closed around an arm movable shaft 104b provided in the middle part.

[0050] Of the two arm portions 104A, 104B, the tip of the first arm portion 104A is provided with upper concave and convex teeth 102. The second arm portion 104B is composed of three divided arm portions 104Ba, 104Bb, and 104Bc that can rotate around the arm movable shaft 104b independently for each of the three cams 107a, 107b, and 107c.

[0051] An arm fixing pin 111 is disposed between the first arm portion 104A and the second arm portion 104B. The arm fixing pin 111 is configured to be movable along the axial direction of the arm movable shaft 104b. When the arm fixing pin 111 is positioned between one of the three divided arm portions 104Ba, 104Bb, and 104Bc that make up the second arm portion 104B and the first arm portion 104A, the divided arm portion is connected and fixed to the first arm portion 104A.

[0052] In the illustrated example, the arm fixing pin 111 is located between the central split arm portion 104Bb of the three split arm portions 104Ba, 104Bb, and 104Bc and the first arm portion 104A. When the cam drive gear 106 rotates, and the three cams 107a, 107b, and 107c fixed thereto rotate, each cam pushes up the three split arm portions 104Ba, 104Bb, and 104Bc.

[0053] At this time, in a split arm portion (for example, split arm portion 104Ba) where the arm fixing pin 111 is not located, only the split arm portion 104Ba rotates around the arm movable shaft 104b, and no force from the cam 107a is transmitted to the first arm portion 104A, as shown in FIG. 10(a). In contrast, in a split arm portion 104Bb where the arm fixing pin 111 is located, the split arm portion 104Ba and the first arm portion 104A are connected and fixed together by the arm fixing pin 111, as shown in FIG. 10(b). Therefore, the force from the cam 107b is also transmitted to the first arm portion 104A, and the first arm portion 104A rotates around the arm rotation center portion 104a in response to the operation of the cam 107b, causing the upper and lower concave and convex teeth 101, 102 to mesh with each other.

[0054] According to this example, by controlling the position of the arm fixing pin 111 (the position in the axial direction of the arm movable shaft 104b), it is possible to select a binding operation according to the cam shapes of the cams 107a, 107b, and 107c, and perform the pressure binding process on the sheet bundle Pb. Therefore, by appropriately selecting the cam shapes of the cams 107a, 107b, and 107c, it is possible to realize various contents (pressure force, pressurizing time, number of pressurizing times, etc.) of the binding operation (binding process) in the pressure binding processing unit 25.

[0055] Regarding the number of times pressure is applied (number of binding operations), generally, the more times pressure is applied, the stronger the binding force, but if the binding operation is performed with the binding position slightly shifted each time pressure is applied, the binding force may actually decrease. Therefore, when changing the number of times pressure is applied, it is preferable to either make the binding operation performed at the exact same position as the first time when pressure is applied from the second time onwards, or to completely shift the binding position so that the binding marks from the previous pressures do not overlap.

[0056] Furthermore, the learning data in the learning phase of the trained model may include environmental information including at least one of temperature, humidity, and air pressure. In this case, for example, in the post-processing device 3, as shown in FIG. 11, a configuration is added to acquire characteristic data (temperature, humidity, air pressure, etc.) of the surrounding environment detected by the environmental sensor 32 in addition to the seat physical property information. Then, when the trained model is executed in the post-processing device 3 (inference phase), the environmental information (temperature, humidity, air pressure in the surrounding environment) detected by the environmental sensor 32 serving as an environmental information detection means such as a temperature sensor or a humidity sensor is also used as input data for the trained model.

[0057] For example, even if the sheets have the same physical properties, differences in the environment in which the crimp binding process is actually performed can result in large errors in the bond strength evaluation value (estimated bond strength) output by the trained model. Therefore, using environmental information as part of the training data is useful for improving the estimation accuracy of the bond strength evaluation value by the trained model.

[0058] Furthermore, the control unit 301 of the post-processing device 3 determines control parameters (pressure force, pressure time, number of pressurizations, etc.) for the pressure binding process for the sheet bundle Pb based on the binding force evaluation value (estimated value of binding force) output by the trained model. The determination conditions at this time include, for example, if the estimated value of binding force output by the trained model is less than the target binding force, changing the control parameters to increase the binding force (increasing the pressure force, pressure time, number of pressurizations, etc.) and adjusting the content of the pressure binding process. On the other hand, for example, if the estimated value of binding force output by the trained model is equal to or greater than the target binding force, the control parameters are not changed (without adjusting the content of the pressure binding process) and the pressure binding process is performed.

[0059] However, the conditions for determining the control parameters (pressure force, pressure time, number of pressurizations, etc.) are not limited to the binding force, and other conditions may be added. For example, a condition for reducing the amount of power consumption required for the crimp binding process may be added. Specifically, as shown in FIG. 12, if the estimated value of the binding force output by the trained model (i.e., the estimated value of the binding force when the crimp binding process is performed with the control parameters set to the initial settings) is greater than a standard value (e.g., a necessary and sufficient binding force), the control parameters may be changed to reduce the binding force (reducing the pressure force, pressure time, number of pressurizations, etc.), thereby adjusting the content of the crimp binding process. In this case, the amount of power consumption required for the crimp binding process can be reduced while ensuring that the binding force is equal to or greater than the standard value, thereby reducing energy consumption.

[0060] Furthermore, as a condition for reducing the amount of power consumption required for the pressure binding process, for example, when the estimated binding force output by the trained model is less than the target binding force, when selecting which control parameter (pressure force, pressure time, number of pressure applications, etc.) to adjust, the control parameter that will result in the least increase in power consumption when changed may be selected and adjusted.

[0061] 13, the post-processing device 3 may be provided with a communication device as a communication means capable of communicating with a cloud system 600, which is an external device, via a network. This communication device may be a communication device of the image forming device 2 to which the post-processing device 3 is connected.

[0062] In this embodiment, by providing such a communication device, the binding strength evaluation value estimated by the trained model executed by the control unit 301 of the post-processing device 3 and the information (such as sheet physical property information) used to estimate the binding strength evaluation value can be stored in the cloud system 600. Post-processing devices 3 of the same type as the post-processing device 3 of this embodiment (for example, post-processing devices that employ a pressure binding processing unit 25 with the same configuration) are also communicatively connected to this cloud system 600. Therefore, in this embodiment, the binding strength evaluation value estimated by the trained model of each post-processing device 3 and the information (such as sheet physical property information) used to estimate the binding strength evaluation value from multiple post-processing devices 3 on the market are collected in the cloud system 600.

[0063] Generally, even among the multiple post-processing devices 3 on the market, even if they are the same model, there are individual differences (for example, differences in the usage environment, individual differences in the sensors that detect sheet physical property information, etc.). Therefore, even among the same post-processing devices 3 equipped with the same trained model, individual differences in the post-processing devices 3 can cause variations in the estimated value of the binding force output from the trained model when crimping and binding a sheet bundle made up of the same number of sheets P. In this case, variations in the determined control parameters can result in variations in the resulting binding force.

[0064] As in the present embodiment, by aggregating the estimated values ​​of binding strength and information at the time of estimation from a plurality of post-processing devices 3 on the market in the cloud system 600, it becomes possible to perform processing such as analyzing the cause of such variations in the cloud system 600. In this case, correction information for correcting the estimated values ​​of binding strength output from the trained model, or correction information for correcting control parameters determined from the estimated values ​​of binding strength output from the trained model, can be fed back to each post-processing device 3. As a result, it becomes possible to suppress variations in binding strength due to individual differences that occur among a plurality of post-processing devices 3 on the market.

[0065] Furthermore, in this embodiment, the trained model is implemented in the control unit 301 of the post-processing device 3. However, for example, as shown in FIG. 14, the trained model may be implemented in the processing unit 601 of a cloud system 600 with which the post-processing device 3 can communicate via a network. In this case, the trained model transmits sheet physical property information and sheet number information used to estimate the binding strength (binding strength evaluation value) to the cloud system 600, and the post-processing device 3 receives the results estimated by the trained model on the cloud system 600, thereby enabling control of the pressure binding process according to the binding strength estimated by the trained model. In this case, there is no need to implement the trained model in each post-processing device 3, eliminating the need to provide advanced processing functions and reducing the cost of the post-processing device 3.

[0066] Furthermore, in order to collect learning data in the learning phase for generating a trained model, sheet physical property information, sheet number information, and the like contained in the learning data may be collected by a cloud system 600 from a plurality of post-processing devices 3 on the market, as shown in Fig. 15. In particular, if a configuration is adopted in which data, including bonding strength evaluation values ​​serving as training data, can be transmitted from a plurality of post-processing devices 3 to the cloud system, it becomes easier to collect learning data, and it is possible to collect more learning data and obtain a trained model with higher accuracy.

[0067] FIG. 16 is a flowchart showing an example of the flow of the pressure binding process in this embodiment. First, when starting the pressure binding process (when executing the trained model), the control unit 301 of the post-processing device 3 acquires sheet physical property information and sheet number information about the sheets to be processed (sheets P that will be pressure bound by the post-processing device 3) (S11). Environmental information about the surrounding environment is also acquired (S12). The control unit 301 then inputs this information into the trained model and outputs an estimated value of the binding force when pressure binding is performed with the control parameters set to the initial settings (S13).

[0068] If the estimated binding strength output by the trained model is greater than a first specified value (for example, a value greater than a necessary and sufficient binding strength) (Yes in S14), the control parameters of the pressure binding process are changed so that the binding force is reduced from the initial setting control parameters (S15).Then, the pressure binding process is performed in the pressure binding processing unit 25 using the changed control parameters (S18).

[0069] On the other hand, if the estimated binding strength output by the trained model is smaller than a second specified value (for example, a necessary and sufficient binding strength, and therefore a value smaller than the first specified value) (Yes in S16), the control parameters of the pressure binding process are changed so that the binding force is greater than the initial setting control parameters (S17). Then, the pressure binding process is performed in the pressure binding processing unit 25 using the changed control parameters (S18).

[0070] On the other hand, if the estimated binding strength output by the trained model is between the first specified value and the second specified value (No in S14, No in S16), the crimp binding processing is performed in the crimp binding processing unit 25 with the initial setting control parameters (S18).

[0071] The above description is merely an example, and each of the following aspects provides unique effects. [First aspect] The first aspect is a sheet binding device (e.g., post-processing device 3) including a binding means (e.g., a pressure binding processing unit 25) that binds a processing target sheet bundle Pb including a plurality of processing target sheets P by a binding force obtained by applying pressure between the plurality of processing target sheets P, and a control means (e.g., a control unit 301) that controls the binding process (e.g., the pressure binding process) of the binding means, and an information acquisition means (e.g., a sheet physical property value sensor 31) that acquires physical property information of processing target sheets included in the processing target sheet bundle bound by the binding means and information on the number of sheets in the processing target sheet bundle before the binding process is completed, and sheet physical property information of learning sheets, learning sheets including the learning sheets, and the like. The apparatus is characterized by having an estimation means (e.g., control unit 301) that estimates a binding strength evaluation value (e.g., binding strength) after the binding process for the target sheet bundle based on the sheet physical property information and sheet number information acquired by the information acquisition means by having a computer execute a trained model that has been trained using a plurality of training data including sheet number information of the target sheet bundle and a binding strength evaluation value after the binding process for the learning sheet bundle, and a determination means (e.g., control unit 301) that determines the control content of the control means in the binding process for the target sheet bundle based on the binding strength evaluation value estimated by the estimation means. In a sheet binding device that binds a stack of sheets without using staples or other fastening components, the binding process required to achieve the desired binding strength varies depending on the physical properties of the sheets and the number of sheets to be bound. For example, in a binding process in which binding strength is achieved by compressing a sheet stack and entangling loosened sheet fibers, the resulting binding strength varies depending on the physical properties of the sheets to be bound, even if the binding process is the same (sheet pressure, pressure time, number of pressurizations, etc.). Furthermore, even if the binding process is the same, the resulting binding strength varies because the pressure applied between the sheets varies depending on the number of sheets to be bound. Furthermore, the resulting binding strength does not simply increase by increasing the sheet pressure, pressure time, number of pressurizations, etc.; excessive increases in the sheet pressure, pressure time, number of pressurizations, etc. can damage the sheets and actually reduce the binding strength. Generally, the conditions for the binding process, such as the physical properties of the sheets to be bound and the number of sheets, vary widely depending on the user's usage situation, resulting in an enormous number of combinations of conditions. Furthermore, there are also users who use special types of sheets that are not widely available on the market. Therefore, it is practically difficult to determine the appropriate binding process by conducting tests in advance for every possible combination. Therefore, in the past, it was difficult to appropriately select and execute each binding process and achieve the desired binding strength under each user's various usage situations. In this aspect, physical property information of the sheets included in the target sheet bundle and information on the number of sheets in the target sheet bundle are obtained before the binding process is completed, and a binding strength evaluation value after the binding process for the target sheet bundle is estimated in advance based on this information. Then, based on the previously estimated binding strength evaluation value, control details for the control means for the binding process for the target sheet bundle are determined, and the binding means for the target sheet bundle are controlled using the determined control details. According to this aspect, because a binding strength evaluation value for the target sheet bundle obtained by the binding process is estimated in advance, it is possible to appropriately adjust the binding process details for the target sheet bundle in accordance with the binding strength evaluation value. Furthermore, in this embodiment, the prior estimation of the binding strength evaluation value uses a trained model trained using multiple training data sets including sheet physical property information of the learning sheets, sheet count information of the learning sheet bundle including the learning sheets, and a binding strength evaluation value of the learning sheet bundle after binding processing. To obtain this trained model, binding processes are performed in advance for various learning sheets and various numbers of sheets, and binding strength evaluation values ​​related to the respective binding strengths are obtained. The trained model is then obtained by performing machine learning or the like using a training dataset including multiple pieces of training data including the respective sheet physical property information, sheet count information, and binding strength evaluation values. Because the configuration is such that the binding strength evaluation value is estimated using the trained model, even if the type or number of sheets does not match the sheet physical property information and sheet count information used as the training data when actually binding the target sheet bundle, it is possible to accurately estimate the binding strength evaluation value related to the binding strength after binding processing. Therefore, even if the conditions for the binding process, such as the physical properties of the sheets to be bound and the number of sheets, vary widely depending on the usage situation of each user, it is possible to grasp in advance the binding strength after the binding process under each user's usage situation and appropriately adjust the binding process content. Therefore, it is possible to appropriately select the binding process content for the sheet stack under each user's usage situation and obtain the desired binding strength.

[0072] [Second mode] A second aspect is the first aspect, characterized in that the binding strength evaluation value includes a sheet binding strength of the sheet bundle after binding processing. According to this, since a binding strength evaluation value that directly indicates the sheet binding strength is estimated, it is possible to more appropriately determine the control content of the control means for the binding process on the sheet bundle to be processed.

[0073] [Third aspect] A third aspect is characterized in that, in the first or second aspect, the control content determined by the determination means is at least one of the sheet pressure, sheet pressure time, and number of sheet pressure applications of the binding means in the binding process. This allows for appropriate adjustment of the binding force by adjusting the control content.

[0074] [Fourth aspect] A fourth aspect is characterized in that, in any of the first to third aspects, the information acquisition means also acquires environmental information including at least one of temperature, humidity, and air pressure detected by an environmental information detection means (e.g., environmental sensor 32) before the completion of the binding process, the trained model is trained using a plurality of training data including environmental information detected by the environmental information detection means during the binding process for the learning sheet stack, and the estimation means estimates the binding strength evaluation value based also on the environmental information acquired by the information acquisition means by having a computer execute the trained model. A trained model trained using training data that includes environmental information that affects binding strength makes it possible to estimate binding strength evaluation values ​​with high accuracy.

[0075] [Fifth mode] A fifth aspect is characterized in that, in any of the first to fourth aspects, the determination means determines the control content of the control means in the binding process so that the binding strength evaluation value estimated by the estimation means falls within a target range (for example, a second specified value or greater) and the amount of power consumption required for the binding process for the sheet stack to be processed is reduced. This makes it possible to optimize the binding process while also taking into account power consumption.

[0076] [Sixth aspect] The sixth aspect is characterized in that, in any of the first to fifth aspects, the estimation means has a communication means (e.g., a communication device) for transmitting the binding strength evaluation value estimated by the estimation means and information used to estimate the binding strength evaluation value to the outside. According to this, for example, by consolidating the binding force evaluation values ​​from multiple sheet binding devices and the information used to estimate the binding force evaluation values ​​in an external device, it becomes possible to suppress variation in the estimation results of the binding force evaluation values ​​due to the fixation between the sheet binding devices in the external device.

[0077] [Seventh aspect] A seventh aspect is a sheet binding system having a sheet binding device (e.g., a post-processing device 3) and an external device (e.g., a cloud system 600) that is communicatively connected to the sheet binding device, wherein the sheet binding device is a sheet binding device of any of the first to sixth aspects without the estimation means, and has a communication means (e.g., a communication device) for communicating with the external device, and the external device has the estimation means (e.g., a processing unit 601), wherein the communication means of the sheet binding device, when binding the sheet stack to be processed, transmits sheet physical property information and sheet number information acquired by the information acquisition means to the external device, and receives the binding force evaluation value estimated by the sheet physical property information and sheet number information by the estimation means of the external device, and the determination means of the sheet binding device determines the control content of the control means in the binding process for the sheet stack to be processed based on the binding force evaluation value received by the communication means. This eliminates the need to implement a trained model in the post-processing device, eliminating the need to provide advanced processing functions, and making it possible to reduce the cost of the post-processing device.

[0078] [Eighth aspect] An eighth aspect is characterized in that the sheet binding system according to the seventh aspect includes a plurality of the sheet binding devices. This eliminates the need to implement trained models in multiple post-processing devices, making it possible to reduce the cost of the entire system.

[0079] [Ninth aspect] The ninth aspect is an image forming system 1 having an image forming device 2 and a sheet binding device 3 that binds a stack of sheets to be processed, which includes image-formed sheets on which the image forming device has formed images as sheets to be processed, and is characterized in that the sheet binding device is any of the sheet binding devices of the first to sixth aspects. According to this aspect, it is possible to provide an image forming system that can appropriately select the content of the binding process for the sheet bundle under the usage conditions of each user, and can obtain a desired binding strength. [Explanation of symbols]

[0080] 1: Image forming system 2: Image forming device 3: Post-processing device 10 to 19: Transport roller pair 20: Switching claw 21: First discharge tray 22: Internal tray 23: End fence 24L, 24R: Side fence 25: Pressure binding processing section 26: Second output tray 27: End fence 28: Saddle stitching processing unit 29: Paper folding blade 30: Third output tray 31: Seat physical property sensor 32: Environmental sensor 101: Lower uneven teeth 102: Upper uneven teeth 103: Lower teeth arm 104: Upper teeth arm 104A: First arm 104B: Second arm part 104Ba~104Bc: Split arm section 104a: Arm rotation center 104b: Arm movable axis 105: Gear 106: Cam drive gear 107, 107a, 107b, 107c: Cam 108: Return spring 109: Pedestal 110: Motor 111: Arm fixing pin 155: Staple binding processing section 200: Image forming device 201:Display section 202:Operation unit 203:Paper feed section 204: Imaging section 205: Fixing section 206: Control unit 301: Control section 600: Cloud system 601: Processing section P: Seat Pb: Sheet stack Phase 1: First transport route Ph2:Second conveyance path Phase 3: Third transport route [Prior art documents] [Patent documents]

[0081] [Patent Document 1] Japanese Patent Application Laid-Open No. 2014-168890

Claims

1. a binding means for binding a bundle of sheets to be processed, the bundle including the plurality of sheets to be processed, by a binding force obtained by applying pressure between the plurality of sheets to be processed; A sheet binding device including: a control unit that controls the binding process of the binding unit; an information acquiring means for acquiring, before the completion of the binding process, physical property information of the target sheets included in the target sheet bundle to be bound by the binding means and information on the number of sheets in the target sheet bundle; an estimation means for estimating a binding strength evaluation value after the binding process for the target sheet bundle based on the sheet physical property information and sheet number information acquired by the information acquisition means by having a computer execute a trained model trained using a plurality of training data including sheet physical property information of the learning sheet, sheet number information of the learning sheet bundle including the learning sheet, and a binding strength evaluation value after the binding process for the learning sheet bundle; a determining means for determining a control content of the control means in the binding process for the target sheet bundle based on the binding strength evaluation value estimated by the estimating means.

2. The sheet binding device according to claim 1, The sheet binding device, wherein the binding strength evaluation value includes a sheet binding strength of the sheet stack after binding processing.

3. The sheet binding device according to claim 1 or 2, The sheet binding device, wherein the control content determined by the determination means is at least one of a sheet pressure, a sheet pressure time, and a number of sheet pressure applications of the binding means in the binding process.

4. The sheet binding device according to claim 1 or 2, The information acquiring means also acquires environmental information including at least one of temperature, humidity, and atmospheric pressure detected by an environmental information detecting means before the binding process is completed; the trained model is trained using a plurality of training data including environmental information detected by an environmental information detection means during a binding process on the training sheet bundle, A sheet binding device characterized in that the estimation means estimates the binding force evaluation value based on the environmental information acquired by the information acquisition means by having a computer execute the learned model.

5. The sheet binding device according to claim 1 or 2, A sheet binding device characterized in that the determination means determines the control content of the control means in the binding process so that the binding force evaluation value estimated by the estimation means falls within a target range and the amount of power consumption required for the binding process for the sheet stack to be processed is reduced.

6. The sheet binding device according to claim 1 or 2, A sheet binding device comprising: a communication means for transmitting the binding strength evaluation value estimated by the estimation means and information used to estimate the binding strength evaluation value to an external device.

7. a sheet binding device; A sheet binding system having an external device communicably connected to the sheet binding device, The sheet binding device is the sheet binding device according to claim 1 or 2, except that the estimation means is removed, and the sheet binding device has a communication means for communicating with the external device, the external device has the estimation means, The communication means of the sheet binding device, when binding the target sheet bundle, transmits the sheet physical property information and sheet number information acquired by the information acquisition means to the external device, and receives the binding strength evaluation value estimated by the estimation means of the external device based on the sheet physical property information and the sheet number information; A sheet binding system characterized in that the determination means of the sheet binding device determines the control content of the control means in the binding process for the sheet stack to be processed based on the binding strength evaluation value received by the communication means.

8. The sheet binding system according to claim 7, A sheet binding system comprising a plurality of the sheet binding devices.

9. an image forming apparatus; and a sheet binding device that binds a sheet bundle to be processed, the sheet bundle including image-formed sheets on which the image forming device has formed an image, as sheets to be processed, 3. An image forming system, comprising: the sheet binding device according to claim 1 or 2;

Images