Three-dimensional printer using a robot and a control device for a robot

Abstract

Provided is a 3D printer capable of molding an article of high strength and high precision using a robot and a control device for the robot. The robot used as the 3D printer has an input-side encoder for acquiring angle information of an input shaft of a joint of the robot that has acted based on a control command and an output-side encoder for acquiring angle information of an output shaft. The control device has an arithmetic unit having a control command storage unit that stores a control command for a motor of each shaft of the robot, a deviation estimation unit that receives detection results of both the input-side encoder and the output-side encoder and estimates a deviation between the control command and an actual trajectory of the robot, and a control command correction unit that corrects the control command using an estimation result of the deviation estimation unit.

Description

TECHNICAL FIELD

[0001] The present application relates to a 3D printer using a robot and a control device for the robot. BACKGROUND

[0002] A layered molding device (so-called 3D printer) that stacks materials such as resin, metal, and the like to mold (print) a three-dimensional object can easily perform molding even for a complex shape, compared to injection molding and the like. As a typical 3D printer, a gantry-type processing machine can be exemplified, but a 3D printer using an industrial robot is also known (for example, refer to Patent Documents 1 and 2).

[0003] On the other hand, in the internal mechanism of the speed reducer of the robot, bending or deformation can occur due to insufficient rigidity or the like, and therefore a robot provided with an output-side encoder for detecting the rotation angle of the output shaft of the speed reducer in addition to an input-side encoder for detecting the rotation angle of the rotation shaft of the motor is known (for example, refer to Patent Documents 3 and 4).

[0004] PRIOR ART DOCUMENTS

[0005] PATENT DOCUMENTS

[0006] Patent Document 1: Japanese Patent Application Laid-Open No. 2019-537521

[0007] Patent Document 2: Japanese Patent Application Laid-Open No. 2019-098682

[0008] Patent Document 3: Japanese Patent Application Laid-Open No. 2016-027951

[0009] Patent Document 4: Japanese Patent Application Laid-Open No. 2020-121355 SUMMARY

[0010] PROBLEMS TO BE SOLVED BY THE INVENTION

[0011] In a 3D printer using a gantry-type processing machine, since the stacking direction of resin or the like is limited, the strength and rigidity of the molded object are sometimes insufficient. On the other hand, in a 3D printer using a robot, since the stacking direction can be arbitrarily changed, the strength and rigidity of the molded object can be improved, but due to the influence of the backlash of the speed reduction mechanism, the reaction force received by the robot when the melted resin is ejected from the nozzle provided at the front end of the robot arm, or the like, the accuracy of the motion trajectory of the robot sometimes deteriorates. By merely providing an input-side encoder at the motor that drives each axis of the robot as a unit for preventing this accuracy deterioration, the influence of the backlash of the speed reducer or the like cannot be correctly measured or estimated, and therefore it is difficult to appropriately correct the motion trajectory.

[0012] SOLUTION TO THE PROBLEM

[0013] One embodiment of the present disclosure is a robot control device that controls a multi-joint robot that has a nozzle for ejecting a material of a molded object, the robot control device including: a control command storage unit that stores a control command including a predetermined print path that the nozzle should follow when moving while applying the material to the molded object; a deviation estimation unit that estimates a deviation between a print completion path that is an actual movement trajectory of the nozzle and the predetermined print path, based on angle information on an input side and angle information on an output side of a joint of the multi-joint robot that has performed a movement according to the control command; and a control command correction unit that corrects the control command to reduce or eliminate the estimated deviation.

[0014] Another embodiment of the present disclosure is a three-dimensional printer including: a multi-joint robot that has a nozzle for applying a material of a molded object and layering the material; and the robot control device according to the one embodiment, which controls the multi-joint robot.

[0015] Effects of Invention

[0016] According to the present disclosure, in a 3D printer using a robot, not only an angle information acquisition unit is provided on an input shaft side of the robot, but also an angle information acquisition unit is provided on an output shaft side of the robot, and correction of a movement trajectory of the robot is performed also taking into account an influence of an application reaction force, whereby a molded object can be formed with extremely high precision. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 is a schematic view of a robot and a molded object used in a 3D printer according to one embodiment.

[0018] Figure 2 is a schematic view of a nozzle unit of the robot mounted to Figure 1

[0019] Figure 3 is a schematic view of a door-type 3D printer and a molded object according to a comparative example.

[0020] Figure 4 is a view schematically showing main parts of the robot of Figure 1

[0021] Figure 5 is a functional block diagram of a robot control device according to a first embodiment.

[0022] Figure 6 is a schematic view showing a state in which a resin filament is applied to a molded object. ​​

[0023] Figure 7 is a schematic view showing a state in which resin filaments are further applied in the state of Figure 6

[0024] Figure 8 is a view showing an example of the density of printing.

[0025] Figure 9 is a view showing another example of the density of printing.

[0026] Figure 10 is a view showing still another example of the density of printing.

[0027] Figure 11 is a functional block diagram of a robot control device according to the second embodiment.

[0028] Figure 12 is a schematic view showing a state in which resin filaments are further applied in the state of Figure 6

[0029] Figure 13 is a view showing an example of the process of correcting the trajectory of the robot in stages.

[0030] Figure 14 is a schematic view showing a state in which resin filaments are applied to a coated surface having unevenness as a comparative example.

[0031] Figure 15 is a view schematically showing main parts of the robot of Figure 1 and additionally showing a force sensor.

[0032] Figure 16 is a functional block diagram of a robot control device according to the third embodiment.

[0033] Figure 17 is a graph showing an example of the relationship between the application direction and the application reaction force.

[0034] Figure 18 is a schematic view showing a state in which resin filaments are applied to a coated surface having unevenness in the third embodiment.

[0035] Figure 19 is a schematic view showing a state in which resin filaments are layered on a coated surface in the third embodiment.

[0036] Figure 20 is a functional block diagram of a robot control device according to the fourth embodiment.

[0037] Figure 21 is a schematic view showing a state in which resin filaments are layered on a coated surface in the fourth embodiment.​​ DETAILED DESCRIPTION

[0038] Figure 1 is a schematic diagram of a robot 10 that is used as a preferred embodiment of a 3D printer and a molded object 12 that is molded by the 3D printer. The robot 10 is, for example, a six-axis industrial vertical multi-joint robot that has a base 14 provided on the ground or the like, a swing cylinder 16 connected to the base 14 so as to be rotatable about a substantially vertical axis, an upper arm 18 connected to the swing cylinder 16 so as to be rotatable, a forearm 20 connected to the upper arm 18 so as to be rotatable, a wrist 22 connected to the forearm 20 so as to be rotatable, and an end effector 24 connected to the wrist 22 so as to be rotatable. In the present embodiment, the end effector 24 is a nozzle unit for ejecting and applying a resin filament that is a material constituting the molded object 12 to the molded object 12.

[0039] Figure 2 A schematic structure of the nozzle unit 24 is shown. The nozzle unit 24 has a nozzle 27 that applies a filament 26 that is a resin material in a filament shape after being melted to the molded object 12, a heater (not shown) that melts the filament 26, and a roller 28 that is used to feed out the filament 26, and these structures can be the same as known structures. The temperature of the heater and the amount of rotation of the roller 28 can be controlled using at least a control command from a robot control device described later. In addition, data indicating a correlation between the temperature of the heater and the amount of rotation of the roller and the amount of feed of the filament can be acquired and stored in advance, and the amount of feed of the filament can be controlled based on the data.

[0040] Figure 3 is a schematic diagram of a gantry-type 3D printer 11 that is involved in a comparative example and a molded object 13 that is molded by the gantry-type 3D printer 11, where the shape of the molded object 13 is assumed to be the same as the shape of the molded object 12. As shown in the drawing, in the gantry-type 3D printer 11, the direction of application of the resin filament is defined as a substantially horizontal direction in the entire molded object 13, and the filaments are stacked in a substantially vertical direction. If the direction of application and the direction of stacking are the same as in this case, there is a problem that the strength of the molded object against a specific direction is weaker than the strength against other directions, and for example, breakage is likely to occur at a portion where stress is likely to concentrate, such as the leg portion 15.

[0041] On the other hand, in a robot 3D printer using the multi-joint robot 10, the position and posture of the nozzle 27 can be arbitrarily changed, determined within the movable range of the robot 10, and the filament can be applied in a manner of drawing an arbitrary trajectory. Generally, the strength of the molded object varies depending on the application direction of the filament, but in the present embodiment, since the filament can be applied in a manner of drawing an arbitrary trajectory, 3D printing can be performed with priority given to the strength of the molded object. For example, as shown in FIG. 13, in the molded object 12, the stacking direction of the resin can be changed for each portion thereof, and a product having a higher strength as a whole than the molded object 13 can be molded and manufactured. In particular, in the case where the filament is made anisotropic by mixing fibers into the resin filament, the trajectory can be determined in a manner of making the anisotropy effectively function, and the strength of the molded object can be made higher. Figure 1

[0042] In a 3D printer like this using the robot 10, a molded product having a high strength can be formed, but on the other hand, there is also a demand to be able to apply the resin filament with at least the same degree of accuracy as the gantry type 3D printer. Hereinafter, a specific scheme for achieving this demand will be described.

[0043] Figure 4 The main part around each axis of the robot 10 is schematically shown together with a robot control device 30 for controlling the robot 10. The robot 10 is provided with an angle information acquisition section that acquires angle information of an input shaft and angle information of an output shaft of a joint of the robot 10 that has been actuated based on a control command, and specifically, the robot 10 has an input side encoder 36 that detects the rotation angle of an input shaft 34 driven by a motor 32, and an output side encoder 42 that detects the rotation angle of an output shaft 40 linked to the input shaft 34 via a reduction mechanism 38 or the like. The output side encoder 42 in the illustrated example has a circular plate or ring-shaped scale member 44 that rotates integrally with the output shaft 40, and a sensor 46 that optically reads a pattern for angle detection provided to the scale member 44, and the same structure can be provided for the input side encoder 36.

[0044] ​By providing both the input-side encoder 36 and the output-side encoder 42 on the rotary shaft, even if bending or deformation occurs due to the wobble, lack of rigidity of the internal mechanism of the speed reducer 38, the positioning accuracy and the trajectory accuracy of each joint shaft of the robot 10 can be improved by using both the rotation angles detected by the input-side encoder 36 and the output-side encoder 42. Further, the structure of such input-side encoder 36 and output-side encoder 42 can be the same as that described in Patent Document 4, for example, and thus detailed description is omitted. Further, as a unit that acquires angle information of the output shaft 40, in addition to the output-side encoder 42, an image processing device (not shown) that provides a camera (not shown) on the robot arm and processes the image of the camera can be used.

[0045] (First Embodiment)

[0046] Figure 5 is a functional block diagram of a first embodiment of the robot control device 30. The control device 30 is provided with, for example, an arithmetic unit 50 having a CPU, a RAM, and the like, a display device 52 such as a liquid crystal display, and a storage device 54 having a non-volatile storage apparatus, a ROM, and the like, wherein the arithmetic unit 50 can be included in the storage device 54. In addition, a portable operation panel 56 that can be carried by an operator is connected to the control device 30 in a manner that enables communication by wire or wirelessly, and transmits, inputs, and the like to the control device 30 the input content of the operator to the operation panel 56.

[0047] The control device 30 is connected to the motor 32 of each shaft of the multi-joint robot 10 and the nozzle unit 24, and controls these structures by a control section 57 such as a CPU. The input-side encoder 36 and the output-side encoder 42 provided on each shaft can detect angle information of the input shaft 34 and the output shaft 40, respectively, and transmit the detected angle information to the control device 30. The nozzle unit 24 can detect the feed amount of the filament, can transmit the detected feed amount to the control device 30, and can receive an instruction related to the feed amount from the control device 30.

[0048] As shown in Figure 2 , in a case where the robot 10 is used as a 3D printer, since the nozzle 27 is ejected with the molten filament 26 having viscosity while maintaining a state close to the coating surface 60 to perform coating, the nozzle 27 is subjected to a coating reaction force. More specifically, a reaction force 58 in the axial direction of the nozzle 27 and a reaction force 62 in the coating surface 60 of the coating object 12 act on the nozzle 27. Due to the influence of these reaction forces and the friction, backlash, and the like in the internal mechanism of the robot 10, a deviation can occur between the control command and the actual robot trajectory. Therefore, in feedback using only the input-side encoder 36, for example, as shown in Figure 6The possibility of generating a coating deviation due to a deviation of the actual robot trajectory from the control command is high as with the filament 26b opposite the filament 26a. In addition, since it is difficult to perform appropriate feedback after a coating deviation like the filament 26b has occurred only by the input side encoder 36, if it is desired to further coat the filament from the state Figure 6 the nozzle 27 cannot obtain an appropriate coating reaction force by contacting the filament 26b and the like, as a result, as Figure 7 illustrated, there is a possibility that the filament is deformed like the filament 26c or protrudes from the coating surface 60, resulting in so-called coating failure.

[0049] Therefore, as Figure 5 illustrated, the arithmetic unit 50 has a control command storage unit 66 that stores a control command for the motor 32 of each axis of the robot 10 that is designated or made in advance, a deviation estimation unit 68 that receives detection results of both the input side encoder 36 and the output side encoder 42 and estimates a deviation between the control command and the actual trajectory of the robot 10, and a control command correction unit 70 that corrects the control command using the estimation result of the deviation estimation unit 68.

[0050] The basic operation of the robot 3D printer configured as such will be described. In the storage device 54 of the control device 30, 3D shape data of the molded object 12, a print path generation algorithm required to form the molded object by the 3D printer in accordance with the 3D shape data, a predetermined print path, a print completion path, a path correction algorithm, and a predetermined print correction path are stored. Hereinafter, each term will be described.

[0051] The 3D shape data is data indicating the shape of the molded object to be finally formed by the 3D printer, and also contains information on the strength required for the molded object (each portion thereof), the print direction (the moving direction of the nozzle when the filament is ejected to the molded object), and the density of the print, and the like. As a specific example of the density, any one of hollow portions 72a to 72c having different cross-sectional areas formed in the inside of the substantially cylindrical molded object 12 as illustrated in Figure 8 to Figure 10 is formed, and the substantial density of the object 12 is made to be a desired value. In the illustrated example, the cross-sectional shape of the hollow portion is a hexagon, but is not limited thereto, and can be any shape. Such a density of the print can also be achieved in the conventional gantry type 3D printer, but in the case of using the robot, since the filament can be stacked in an arbitrary direction, a hollow portion of a more complex and preferable shape can be formed.

[0052] The predetermined printing path is a path to be followed by the nozzle 27 when moving while applying the resin filament to the molding object 12, and is generated by reading the 3D shape data of the molding object 12 into a path generation algorithm, and the robot 10 starts and performs the action based on the predetermined printing path included in the control command. The path generation algorithm includes a mechanism interference analysis algorithm for a multi-joint robot, for generating a path as a trajectory in which the robot can perform the action. However, the predetermined printing path can also be fine-tuned by the operator's operation, not completely relying on the mechanical path generation algorithm. Furthermore, in the present embodiment, the correction of the path is accumulated for each printing to achieve high precision, and the printing is also referred to as a printing job.

[0053] The printing completed path is an actual action trajectory of the nozzle 27 reproduced based on the detection results of the input-side encoder 36 and the output-side encoder 42, the filament feed amount, and feedback from a force sensor described later, and is also referred to as a printing history for each printing job.

[0054] The predetermined printing correction path is a path obtained by at least using the above-described 3D shape data, the executed predetermined printing path, and the printing completed path corresponding thereto, to smoothly connect the executed predetermined printing path and the printing completed path corresponding thereto during the printing, and to correct the remaining (unexecuted) predetermined printing path by applying a solid deformation thereto, so that the final shape of the molding object coincides with the shape represented by the 3D shape data. The predetermined printing path, the printing completed path, and the predetermined printing correction path can also be configured to be able to be confirmed in a state that can be visually understood from the external user interface using the display device 52 or the operation panel 56 or the like.

[0055] The path correction algorithm is an algorithm for generating the predetermined printing correction path. It is possible to additionally specify to what extent the original predetermined printing path is faithfully reproduced, the timing of reflecting the correction, and the timing of updating.

[0056] Next, a specific example of the process in the control device 30 related to the first embodiment Figure 5 will be described. First, the control command based on the above-described predetermined printing path is stored in the control command storage section 66 of the arithmetic unit 50. When the printing of the molding object 12 is started, the control command is transmitted to the motors 32 of each axis of the robot 10 and the nozzle unit 24, so that the nozzle 27 moves and sprays the resin filament 26 according to the predetermined printing path of the molding object 12. For example, the robot 10 sprays the resin filament 26a to the molding object 12 as shown in FIG. 6 according to the transmitted control command. Figure 6

[0057] ​Here, the printing completion path is generated in the calculation unit 50 based on the detection results of the input-side encoder 36 and the output-side encoder 42. Next, in the deviation estimation unit 68, the deviation between the control command and the actual robot motion trajectory is calculated and estimated based on the difference between the predetermined printing path and the printing completion path. Then, the control command correction unit 70 corrects the control command to reduce or eliminate the estimated deviation. For example, in... Figure 6 In the example, whenever a straight filament is coated, or whenever segments obtained by dividing a filament into pre-specified lengths are coated, a printed path is created, and the control command is corrected based on its deviation from the predetermined printed path. This allows for the detection of significant deviations of the filament 26b from the predetermined path, and the correction of the filament 26b's coating path to reduce or eliminate deviations. Therefore, it is possible to significantly reduce deviations such as... Figure 7 The possibility of further increasing the problem, as seen with filament 26c. As for the completed printing path in the first embodiment, it is sufficient to have information about the filaments or segments near the filaments or segments to be coated.

[0058] In the first embodiment, information from the output-side encoder 42 is read in before printing begins, and the read information is used to correct control commands, thereby achieving high precision in the motion trajectory of the robot 10 (nozzle 27). That is, the output-side encoder 42 can accurately detect errors in the nozzle trajectory, including those caused by the coating reaction force, a phenomenon unique to 3D printers, thus solving problems specific to 3D printers and achieving extremely high-precision 3D printing. For example, producing... Figure 6 The probability of coating deviations like those in filament 26b is significantly reduced. Even if coating deviations like those in filament 26b occur, the coating deviations are as follows: Figure 7 The probability of further expansion of the filament, such as 26c, is also greatly reduced.

[0059] (Second Embodiment)

[0060] Figure 11 This is a functional block diagram of a second embodiment of the robot control device 30. Furthermore, in the second embodiment, only the points that differ from the first embodiment are described; for points that are the same as in the first embodiment, their description is omitted.

[0061] In addition to a control command storage unit 66, a deviation estimation unit 68, and a control command correction unit 70, the calculation unit 50 also includes a recalculation unit 74. This recalculation unit 74 accumulates historical deviation correction data and recalculates the trajectory that the nozzle 27 should follow. The control command correction unit 70 then adjusts the deviation correction amount based on the recalculated trajectory. In other words, after a coating deviation occurs in the width direction of the coating, the recalculation unit 74 generates one or more predetermined printing correction paths to gradually approximate the target trajectory. The control command correction unit 70 then performs phased corrections to the control commands based on these predetermined printing correction paths. For example, after a coating deviation occurs in the width direction of the coating, the recalculation unit 74 generates one or more predetermined printing correction paths to gradually approximate the target trajectory. Figure 6 After the coating deviation is achieved as shown by filament 26b, coating is performed in stages to gradually reduce the coating deviation, as shown by filament 26d, so that the final coating is consistent with the target trajectory as shown by filament 26e. In the second embodiment, the printing completion path preferably includes all past printing history.

[0062] Figure 13 Used to illustrate implementation Figure 12 A specific example of such coating process. Here, the path of the coated filament is represented by dot group data. In filament 26a, the predetermined printing path, indicated by the circle mark 76, is approximately the same as the printed completion path, indicated by the triangle mark 78. However, it is assumed that due to uneven coating reaction force, the printed completion path 78 deviates from the predetermined printing path 76, as shown in filament 26b.

[0063] In the subsequent coating of filament 26d, the path correction is not performed by an amount equivalent to the deviation produced in filament 26b, but rather multiple times along the predetermined print correction path indicated by square mark 80 (in Figure 12 In the example (4 times), a smaller amount of correction is performed than the aforementioned deviation, thus ultimately transforming the filament 26e into a predetermined printing correction path 80 that is approximately consistent with the predetermined printing path 76. This allows for smooth correction of deviations to obtain a molded object of the desired shape. Various methods can be used to harmonize the coordinates between such point groups; for example, methods such as simply calculating the average with the surrounding area or using a numerical analysis thermal balance model can be employed.

[0064] In the second embodiment, similar to the first embodiment, after detecting and estimating the deviation between the predetermined printing path and the printed path, a path correction algorithm is executed. Based on the 3D shape data or the estimated deviation, the predetermined printing correction path is generated according to the pre-specified correction intensity, and the 3D printing path (control command) is updated at a specified time interval.

[0065] Here, the correction intensity refers to the frequency at which the trajectory deviation is corrected during the generation of the predetermined printing correction path, as preset and modified by the operator; the stage at which the path gradually approaches the predetermined printing path; and the range of other trajectories within a certain distance from the target trajectory used for deviation correction. Furthermore, the generation of the predetermined printing correction path can be performed simultaneously with the actual 3D printing. However, when the correction intensity is set to reflect a large range of deviations in the correction, sometimes the generation of the predetermined printing correction path is delayed until the entire printing path within that range has been generated.

[0066] In this second embodiment, regarding filament coating in the width direction of the coating (the direction parallel to the coating surface), a predetermined printing correction path is generated that gradually reduces the deviation between the predetermined printing path and the finished printing path. Using the predetermined printing correction path and the deviation between the predetermined printing correction path and the finished printing path, the control command is corrected in a way that progressively becomes the target coating state. Therefore, coating deviation can be smoothly eliminated, and the appearance of the molded object is the desired appearance.

[0067] (Third Embodiment)

[0068] Figure 14 This example illustrates a coating state where, as a comparative example, an appropriate reaction force is not obtained during filament coating. Here, the coated surface 61 of the molded object 12 is uneven and has irregularities. For example, regarding... Figure 2 As explained, coating reaction forces 58 and 62 act on nozzle 27 when coating filament 26, but especially in cases where the coating surface 61 has unevenness, the coating reaction force can sometimes vary significantly. If the coating reaction force is greater than the expected value, problems such as filament 26 crushing may occur; conversely, if the coating reaction force is less than the expected value, the filament 26 may not conform to the coating surface 61 and may... Figure 14 As shown, there are problems such as peeling from the coating surface 61. In addition, besides the unevenness of the coating surface, the coating reaction force can sometimes vary depending on the temperature and humidity of the coating surface or surrounding environment, the vibration of the robot's trajectory, the state of the nozzle, etc., which can have an adverse effect on print quality.

[0069] Therefore, as Figure 15As shown, in the third embodiment, a force sensor 82 for measuring the reaction force acting on the nozzle 27 is disposed within or near the nozzle unit 24. A specific example of the force sensor 82 is a six-axis force sensor capable of detecting forces in three orthogonal directions and torques in three rotational directions. Using such a force sensor 82, the magnitude and direction of the resultant force of the reaction forces 58 and 62 can be detected, and the magnitudes of the reaction forces 58 and 62 can be determined through simple calculations using this resultant force. However, the unit for estimating the coating reaction force is not limited to a six-axis force sensor. Examples include methods such as providing a slider and spring at the front end of the nozzle unit 24 and determining the force based on the slider's insertion amount and the spring constant; or using an electrostatic capacitive sensor or laser displacement meter to determine the distance between the coating surface and the nozzle and determining the force based on the physical properties of the filament. Furthermore, these methods can also be implemented as auxiliary functions within the nozzle unit.

[0070] like Figure 16 As shown, the computation unit 50 of the robot control device 30, in addition to the control command storage unit 66, the deviation estimation unit 68, and the control command correction unit 70, also has a coating state estimation unit 84. The coating state estimation unit 84 uses the detection result of the force sensor 82 when the resin filament is ejected to estimate the coating state of the filament, which is strongly correlated with the coating reaction force. The control command correction unit 70 uses the estimation result of the deviation estimation unit 68 and the estimated coating state to correct the control command.

[0071] Specifically, for example, pre-prepared Figure 17 The relationship between the coating direction and the ideal coating reaction force, as shown, is set as target data and stored in the storage device 54 of the control device 30. The ideal reaction force is used to obtain a good coating state without crushing or peeling of the filament. This ideal reaction force varies depending on the filament material and diameter, printing speed, printing quality, and coating direction, but this information can also be provided to the storage device 54 in advance. In addition, the relationship between the coating reaction force, the filament feed rate, and the deformation mode during filament coating can also be acquired and stored in advance as data.

[0072] Furthermore, feedback control can be implemented to ensure that the actual coating reaction force detected by the force sensor 82 is within a predetermined target range that achieves a good coating state. In this way, by placing the force sensor 82 between the robotic arm 18 or 20 and the nozzle unit 24, the coating reaction force can be acquired by the control device 30. This allows the nozzle 27 to be pressed against the coating surface 61 using historical records of the coating reaction force, achieving a specified target coating reaction force. For example, [further details are needed]. Figure 18 The printing shown conforms to the bends, bumps, etc. of the coating surface 61.

[0073] (Fourth Embodiment)

[0074] Next, the fourth embodiment of the robot control device 30 will be described. Further, the fourth embodiment is roughly equivalent to an embodiment obtained by combining the second embodiment and the third embodiment, and only the points different from the second embodiment and the third embodiment will be described herein, and the description will be omitted for the points that can be the same as the second embodiment or the third embodiment.

[0075] In the third embodiment, even if the unintended unevenness is present on the coating surface 61, the filament coating that conforms to the unevenness can be performed, but in the case where such coating is repeatedly performed in the thickness direction (the direction substantially perpendicular to the coating surface 61), the outer shape of the final molded object can sometimes reflect the unevenness of the coating surface 61 as shown in Figure 19 . Specifically, the filament 26f that conforms to the unevenness of the coating surface 61 is coated on the coating surface 61, and the filament 26g that is the same shape as the filament 26f is stacked on the filament 26f.

[0076] Therefore, as shown in Figure 20 , the operation section 50 of the robot control device 30 has, in addition to the control command storage section 66, the deviation estimation section 68, the control command correction section 70, and the coating state estimation section 84, a recalculation section 74 that accumulates the deviation correction and recalculates the trajectory, and the control command correction section 70, like the second embodiment, changes the correction amount of the deviation based on the recalculated trajectory. In other words, after the coating deviation occurs in the thickness direction of the coating, the control command is corrected in stages in such a way that the actual trajectory gradually becomes the target trajectory. For example, after the filament that conforms to the unevenness of the coating surface is coated like the filament 26f, the coating that makes the unevenness decrease in stages is performed as shown by the filament 26h, and finally the coating that coincides with the target trajectory is performed as shown by the filament 26i. Figure 19 Figure 21

[0077] Further, in the fourth embodiment, like the second embodiment described using Figure 13 , the predetermined print correction path 80 is generated based on the predetermined print path 76 included in the designated control command, but a new predetermined print path can also be generated based on the previously generated printed path. For example, in the case where an unintended large recess is present on the coating surface 61, the deviation between the printed path and the original predetermined print path is large, and thus it can be difficult to perform appropriate coating by gradual correction, the new predetermined print path that includes a path that first only fills the recess can be generated based on the difference between the printed path and the shape of the molded object. Such processing can also be performed in the second embodiment including the recalculation section 74.​​

[0078] The functions of the respective sections of the control device 30 described above can also be provided by a computer program. For example, the storage device 54 stores various data used in each process executed by the arithmetic section 50, various data generated in the course of each process. Each constituent element of the arithmetic section 50 (the control command correction section 70 and the like) can be a functional module realized by a computer program executed by a processor. Alternatively, such a computer program can be provided in the form of a non-transitory recording medium that is readable by a computer, such as a semiconductor memory, a magnetic recording medium, or an optical recording medium.

[0079] Explanation of Reference Numerals

[0080] 10: robot; 12: molded object; 24: nozzle unit; 26: filament; 27: nozzle; 30: robot control device; 32: motor; 36: input side encoder; 42: output side encoder; 50: arithmetic section; 52: display device; 54: storage device; 56: operation panel; 57: control section; 58, 62: coating reaction force; 60, 61: coating surface; 66: control command storage section; 68: deviation estimation section; 70: control command correction section; 74: recalculation section; 82: force sensor; 84: coating state estimation section.

Claims

1. A robot control device that controls a multi-joint robot that is provided with a nozzle for ejecting a material of a molded object, the robot control device having: a control command storage section that stores a control command including a predetermined print path that the nozzle should follow when moving while coating the material to the molded object; a deviation estimation section that, based on angle information on an input side and angle information on an output side of a joint of the multi-joint robot that has acted in accordance with the control command and a feed amount of the material, calculates a print completed path that is an actual action locus of the nozzle, and estimates a deviation between the print completed path and the predetermined print path; and a control command correction section that corrects the control command so as to reduce or eliminate the estimated deviation.

2. The robot control device according to claim 1, wherein: further having a recalculation section that generates a predetermined print correction path in such a way that the deviation between the predetermined print path and the print completed path is reduced in stages so as to gradually approach the predetermined print path in a width direction of the coating, the control command correction section corrects the control command based on the predetermined print correction path generated by the recalculation section and a deviation between the predetermined print correction path and the print completed path.

3. The robot control device according to claim 2, wherein: the recalculation section generates a new predetermined print path based on a previously generated print completed path and a shape of the molded object.

4. A robot control device that controls a multi-joint robot that is provided with a nozzle for ejecting a material of a molded object, the robot control device having: a control command storage section that stores a control command including a predetermined print path that the nozzle should follow when moving while coating the material to the molded object; a deviation estimation section that, based on angle information on an input side and angle information on an output side of a joint of the multi-joint robot that has acted in accordance with the control command, calculates a print completed path that is an actual action locus of the nozzle, and estimates a deviation between the print completed path and the predetermined print path; a control command correction section that corrects the control command so as to reduce or eliminate the estimated deviation; and a coating state estimation section that estimates a coating state of the molded object based on a reaction force acting on the nozzle during ejection of the material, wherein the control command correction section corrects the control command in such a way that the reaction force is within a target range that is decided in advance based on the coating state.

5. The robot control device according to claim 4, wherein: further having a recalculation section that generates a plurality of predetermined print correction paths that reduce the deviation between the predetermined print path and the print completed path in stages, the control command correction section corrects the control command based on the plurality of predetermined print correction paths generated by the recalculation section and a deviation between the predetermined print correction path and the print completed path. ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ The control command correction section corrects the control command based on the predetermined printing correction path so that the printing completion path gradually approaches the predetermined printing path in a thickness direction of the coating of the material in stages.

6. The robot control device according to claim 5, wherein The recalculation section generates a new predetermined printing path based on a previously generated printing completion path and the shape of the molded object.

7. A three-dimensional printer comprising: a multi-joint robot including a nozzle for coating a material of a molded object and layering the material; and The robot control device according to any one of claims 1 to 6 controls the multi-joint robot.

8. The three-dimensional printer according to claim 7, wherein The multi-joint robot has an input side encoder for acquiring angle information of the input side and an output side encoder for acquiring angle information of the output side.

9. The three-dimensional printer according to claim 7 or 8, wherein The multi-joint robot has a force sensor for detecting a reaction force acting on the nozzle.

Images