Machining device

Abstract

[Problem] To bring a machined result closer to an assumed machined result even if the shape of a machining part changes with machining. [Solution] This machining device comprises: a machining part which performs machining by contacting a part to be machined and is composed of a material of which the shape changes with the machining; an adjustment part which adjusts an interval between the part to be machined and the machining part; a measurement part which measures the shape of the machining part; and a control part which controls the adjustment part on the basis of the shape of the machining part obtained in advance and the shape of the measured machining part so that the interval between the part to be machined and the machining part approaches an interval determined according to the shape of the machining part obtained in advance.

Description

processing equipment

[0001] The technology of the present disclosure relates to a processing device.

[0002] Patent Literature 1 discloses a deburring device that creates a robot program in advance to control the robot's operation based on shape data of the deburring portion of the object extracted from three-dimensional data of the object and the posture of the deburring tool. This deburring device uses a visual sensor to detect the actual position of the deburring portion of the object, and updates the robot program to match the actual position of the object based on the detected actual position of the deburring portion. The deburring device also controls the robot's operation based on the force acting on the deburring tool detected by a force sensor and a predetermined target value for the force.

[0003] JP 2015-134407

[0004] Conventional processing devices, including the above-described deburring device, are based on the assumption that the shape of the processing part, such as a deburring tool, does not change during processing. Therefore, even if the shape of the processing part changes during processing, conventional processing devices process the part to be processed using the processing part, assuming the shape before the change. Therefore, conventional processing devices cannot obtain the expected processing results when the shape of the processing part changes during processing, specifically when the size of the processing part becomes smaller.

[0005] The technology disclosed herein aims to provide a processing device that can bring processing results closer to expected processing results even if the shape of the processing portion changes during processing.

[0006] In order to achieve the above-mentioned object, a processing device according to a first aspect of the technology disclosed herein comprises a processing unit made of a material that processes a part to be processed by contacting the part and whose shape changes as a result of the processing, an adjustment unit that adjusts the distance between the part to be processed and the processing unit, a measurement unit that measures the shape of the processing unit, and a control unit that controls the adjustment unit based on a predetermined shape of the processing unit and the measured shape of the processing unit so that the distance between the part to be processed and the processing unit approaches a distance determined in accordance with the predetermined shape of the processing unit.

[0007] The technology disclosed herein can make the processing results closer to the expected processing results even if the shape of the processing portion changes during processing.

[0008] It is a schematic diagram showing an example of a processing device of an embodiment. It is a diagram showing an example of how the shape of a processing part changes with processing. It is a block diagram showing an example of a control system of a processing device of an embodiment. It is a flowchart showing an example of a teaching processing program. It is a flowchart showing an example of a model creation program. It is a flowchart showing an example of a reproduction processing program.

[0009] Hereinafter, embodiments of the technology of the present disclosure will be described with reference to the drawings.

[0010] 1A is a schematic diagram showing an example of a processing device 1 according to an embodiment. As shown in FIG. 1A, the processing device 1 includes an arm 11 having a plurality of joints, a processing unit 2 attached to the tip of the arm 11, which processes a workpiece 13 by contacting the workpiece 13 and is made of a material whose shape changes as the workpiece 13 is processed, an adjustment unit 3 which adjusts the distance between the workpiece 13 and the processing unit 2, and a camera 5 which measures the shape of the processing unit 2.

[0011] 1B is a diagram showing an example of how the shape of the processing portion 2 changes with processing. Fig. 1B shows an initial shape (cross section) 12A, which is when the processing portion 2 has never processed the workpiece 13, and a post-processing shape (cross section) 12B, which is when the processing portion 2 has processed the workpiece 13 for a certain period of time. As can be seen by comparing the initial shape 12A and the post-processing shape 12B in Fig. 1B, when the processing portion 2 processes the workpiece 13 by contacting it, the shape changes with the processing.

[0012] The arm 11 is configured so that an external force can be applied to move the processing unit 2. Specifically, as will be described later, the motors 22 provided at the respective joints of the arm 11 are freely rotatable when not controlled by the processor 32.

[0013] The arm 11 is provided with a rotating part (not shown) that rotates the processing part 2 around an axis along a horizontal plane.

[0014] The material of the processing portion 2 is, for example, but not limited to, a nylon nonwoven fabric. The processed portion 13 is, for example, but not limited to, an aircraft turbine blade.

[0015] The adjustment unit 3 is configured so that the arm 11 can move horizontally (X-axis direction and Y-axis direction) and vertically (Z-axis direction), as well as rotate circumferentially around an axis (not shown) along the Z-axis.

[0016] In addition, adjustment of the distance between the workpiece 13 and the processing unit 2 is not limited to moving or rotating the arm 11 itself, but may also be performed by rotating the joint of the arm 11 or by using a moving device attached to the arm 11 that moves the processing unit 2.

[0017] The camera 5 is an example of a "measurement unit" of the technology of the present disclosure. The measurement unit is not limited to a camera, and may be, for example, a line sensor.

[0018] Fig. 2 is a block diagram showing an example of a control system of the processing apparatus 1 according to the embodiment. As shown in Fig. 2, the control system of the processing apparatus 1 includes a computer 30, a motor 22, an encoder 24, and a torque sensor 26 provided at each of the joints of the arm 11, an adjustment unit 3, and a camera 5. Note that instead of the torque sensor 26, torque may be calculated from a command value sent from the computer 30 to the encoder 24.

[0019] The computer 30 includes a processor 32, a non-volatile memory (NVM) 34, a random access memory (RAM) 36, and an input / output (I / O) port 38. The processor 32, the NVM 34, the RAM 36, and the input / output (I / O) port 38 are interconnected by a bus 40.

[0020] The processor 32 is an example of a "controller" of the technology of the present disclosure. The NVM 34 is an example of a "storage unit" of the technology of the present disclosure.

[0021] The processor 32 is a processing device including a DSP (Digital Signal Processor), a CPU (Central Processing Unit), and a GPU (Graphics Processing Unit), and the DSP and GPU operate under the control of the CPU and are responsible for executing the processes described below. Here, a processing device including a DSP, a CPU, and a GPU is given as an example of the processor 32, but this is merely an example, and the processor 32 may be one or more CPUs and DSPs with integrated GPU functionality, one or more CPUs and DSPs without integrated GPU functionality, or may be equipped with a TPU (Tensor Processing Unit).

[0022] The NVM 34 is a non-volatile storage device that stores programs, various parameters, etc. The NVM 34 may be, for example, a flash memory (e.g., an EEPROM (Electrically Erasable and Programmable Read Only Memory)). The NVM 34 stores a teaching processing program 34P1, a model creation program 34P2, and a reproduction processing program 34P3. The NVM 34 also stores a trained model 34M, which will be described in detail later.

[0023] The RAM 36 is a memory that temporarily stores information and is used as a work memory by the processor 32. The RAM 36 may be, for example, a dynamic random access memory (DRAM) or a static random access memory (SRAM).

[0024] When the teaching processing program 34P1, model creation program 34P2, and reproduction processing program 34P3 are read into the RAM 36 and executed by the processor 32, the processor 32 functions as a teaching processing unit 32A, a model creation unit 32B, and a reproduction processing unit 32C.

[0025] (Function) The processor 32 controls the adjustment unit 3 based on the predetermined shape of the processed portion 2 and the measured shape of the processed portion 2 so that the distance between the processed portion 13 and the processed portion 2 approaches the distance determined according to the predetermined shape of the processed portion 2.

[0026] FIG. 3 is a flowchart showing an example of the teaching processing program 34P1.

[0027] In step 52, the teaching processing unit 32A photographs the processing unit 2 via the camera 5 and measures the shape of the processing unit 2 during the teaching phase (when the teaching processing program 34P1 is executed) from the image data of the processing unit 2 obtained by the photograph.

[0028] In step 54, the teaching processing unit 32A performs teaching processing. As described above, the arm 11 is configured to be able to move the processing unit 2 when an external force is applied. Specifically, the teaching processing is processing in which the operator directly moves the arm 11 to teach the movement of the arm 11. The movement of the arm 11 is determined by data on each position of the processing unit 2 or data on the magnitude and direction of each force acting from the processing unit 2 to the workpiece 13 during teaching. The position of each joint is determined by output from an encoder 24 provided at each joint of the arm 11, and the position of each joint determines the position of the processing unit 2 attached to the tip of the arm 11. The magnitude of the torque of each joint is determined by output from a torque sensor 26 provided at each joint of the arm 11, and the magnitude and direction of the force acting from the processing unit 2 attached to the tip of the arm 11 to the workpiece 13 are determined by the torque of each joint.

[0029] In step 56, the teaching processing unit 32A determines whether the actual machining result, which is the result of the machining unit 2 contacting the workpiece 13 during the teaching process, is within an acceptable range, for example, whether it is good. Specifically, it determines whether the difference between the actual shape of the workpiece 13, which is the result of the machining unit 2 contacting the workpiece 13 during the teaching process, and the expected shape is within a predetermined acceptable range. In this embodiment, it determines whether the actual shape is the expected shape of the workpiece 13 (stored in the NVM 34). If the actual shape of the workpiece 13 is the expected shape, the machining result is determined to be good. If the actual shape of the workpiece 13 is not the expected shape, the machining result is not determined to be good. If it is determined that the machining result is not good, the teaching processing returns to step 54. If it is determined that the machining result is good, the teaching processing proceeds to step 58.

[0030] In step 58, the teaching processing unit 32A stores the tool shape and the operation data in association with each other. Specifically, the teaching processing unit 32A stores in the NVM 34 the shape of the machining unit 2 during the teaching phase measured in step 52 and the operation data of the arm 11 obtained in step 54 (the data on each position or the data on the magnitude and direction of each force).

[0031] FIG. 4 is a flowchart showing an example of the model creation program 34P2.

[0032] In step 62, the model creation unit 32B photographs the processing unit 2 via the camera 5 and measures the shape of the processing unit 2 during the model creation phase (when the model creation program 34P2 is executed) from the image data of the processing unit 2 obtained by the photograph.

[0033] In step 64, the model creation unit 32B reads the shape of the tool during the teaching phase (the shape stored in the NVM 34 in step 58 in FIG. 3).

[0034] In step 66, the model creation unit 32B calculates the difference between the shape in the teaching phase and the shape in the model creation phase.

[0035] In step 68, model creation unit 32B extracts the force direction in the motion data (the force direction stored in NVM 34 in step 58 of FIG. 3).

[0036] In step 70, the model creation unit 32B corrects each position of the machining unit 2 in the operation data using the learned model based on the difference in the shape of the tool and the direction of the force in the operation data.

[0037] If there is no difference between the shape in the teaching phase and the shape in the model creation phase, the distance between the workpiece 13 and the machining part 2 coincides with the distance determined according to the shape in the teaching phase. There is no need to correct the positions of the machining part 2 in the operation data.

[0038] However, if there is a difference between the shape during the teaching phase and the shape during the model creation phase, the greater the difference, the greater the distance between the workpiece 13 and the machining part 2 will be, which is greater than the distance determined according to the shape during the teaching phase. Therefore, each position of the machining part 2 in the operation data (i.e., each position during the teaching phase) needs to be corrected so that it approaches the distance determined according to the shape during the teaching phase.

[0039] The trained model is a trained model trained using the following two types of multiple data pairs as training data: The first training data is multiple shape differences of the machining portion 2 during the teaching phase and at a time different from the teaching phase; The second training data is a correction amount for correcting the data at each position during the teaching phase so that the distance between the workpiece 13 and the machining portion 2 approaches the distance determined according to the shape during the teaching phase, for each of the multiple shape differences in the first data. Therefore, when the difference in the shape of the tool is input to the trained model, the trained model outputs a correction amount for correcting the data at each position during the teaching phase so that the distance between the workpiece 13 and the machining portion 2 approaches the distance determined according to the shape during the teaching phase.

[0040] Therefore, in this step 70, the model creation unit 32B inputs the difference in tool shape calculated in step 66 into the trained model, and corrects the data for each position during the teaching phase using the correction amount output from the trained model so that the distance between the workpiece 13 and the machining portion 2 approaches the distance determined according to the shape during the teaching phase.

[0041] Note that the method is not limited to correcting the data for each position using a trained model. For example, a correction amount may be searched for using a data table that stores the first data and the second data in correspondence with each other, and the data for each position during the teaching phase may be corrected using the searched correction amount.

[0042] In step 72, the model creating unit 32B moves the arm 11 via the adjusting unit 3 so that the position of the processing unit 2 is positioned at the corrected position.

[0043] In step 74, the model creating unit 32B controls the motors of the joints of the arm 11 based on the operation data with the corrected positions, thereby causing the processing unit 2 to reproduce the processing of the workpiece 13.

[0044] In step 76, the model creation unit 32B determines whether the machining results are good. Note that the processing in step 76 is similar to that in step 56, and therefore its description will be omitted. If it is determined that the machining results are good, the model creation processing ends. If it is not determined that the machining results are good, the model creation processing proceeds to step 78.

[0045] In step 78, the model creation unit 32B corrects the parameters of the trained model based on the difference between the shape of the workpiece 13 after processing by the reproduction processing and the expected shape.

[0046] After processing step 78, the model creation process returns to step 70.

[0047] 5 is a flowchart showing an example of the reproduction processing program 34P3. The reproduction processing unit 32C executes the processes of steps 82 to 98. The processes of steps 82 to 98 are substantially the same as the processes of steps 62 to 78 in FIG. 4, except that the model creation time in the processes of steps 62 to 78 in FIG. 4 is replaced by this reproduction processing time, and therefore a description thereof will be omitted.

[0048] As described above, in this embodiment, the adjustment unit 3 is controlled based on the shape of the machining portion during the teaching phase and the shape of the machining portion measured during the model creation phase or the reproduction phase so that the distance between the machined portion 13 and the machining portion 2 approaches the distance determined according to the shape of the machining portion 2 during the teaching phase. Therefore, this embodiment has the effect of making the machining result closer to the expected machining result even if the shape of the machining portion 2 changes with machining.

[0049] As described above, in this embodiment, in steps 70 and 90, the positions of the machining unit 2 in the operation data are corrected using a trained model based on the difference in the shape of the tool and the direction of the force in the operation data. The technology of the present disclosure is not limited to this. Based on the difference in the shape of the tool and the direction of the force in the operation data, positions in the operation data that bring the distance between the machining unit 2 and the workpiece 13 closer to the distance determined based on the shape of the machining unit during the teaching phase may be geometrically calculated, and the positions in the operation data may be corrected (changed) to the calculated positions.

[0050] REFERENCE SIGNS LIST 1 Processing device 11 Arm 13 Processing part 2 Processing part 3 Adjustment part 5 Camera 32 Processor 34 NVM

Claims

1. A processing device comprising: a processing unit made of a material that processes a part to be processed by contacting the part and whose shape changes as a result of the processing; an adjustment unit that adjusts the distance between the part to be processed and the processing unit; a measurement unit that measures the shape of the processing unit; and a control unit that controls the adjustment unit based on a predetermined shape of the processing unit and the measured shape of the processing unit so that the distance between the part to be processed and the processing unit approaches a distance determined in accordance with the predetermined shape of the processing unit.

2. The processing device of claim 1, wherein the adjustment unit is configured to be able to move the processing unit by the application of an external force, and further comprises a memory unit, and the control unit calculates operation data of the processing unit that has been operated by the application of the external force to the adjustment unit, and when the processing result of the operation of the processing unit on the processed part is within an acceptable range, controls the memory unit so that the shape of the processing part measured when calculating the operation data and the calculated operation data are stored in correspondence with each other.

3. The processing apparatus of claim 2, further comprising an acquisition unit that acquires the direction of the force applied from the outside to the adjustment unit, wherein the control unit corrects the operation data based on the difference between the shape of the processing part measured when calculating the operation data and the shape of the processing part measured after the operation data is stored, and the acquired direction of the force, and controls the adjustment unit so that the processing part applies processing to the processed part based on the corrected operation data.

4. The processing device according to claim 3, wherein the control unit corrects the operation data when a processing result of the workpiece processed based on the corrected operation data is outside an allowable range.

Images