WO2020203876A1 - Industrial machine, control device, control compensation device, and control method - Google Patents

Industrial machine, control device, control compensation device, and control method Download PDF

Info

Publication number
WO2020203876A1
WO2020203876A1 PCT/JP2020/014300 JP2020014300W WO2020203876A1 WO 2020203876 A1 WO2020203876 A1 WO 2020203876A1 JP 2020014300 W JP2020014300 W JP 2020014300W WO 2020203876 A1 WO2020203876 A1 WO 2020203876A1
Authority
WO
WIPO (PCT)
Prior art keywords
value
tool
work
reaction
actuator
Prior art date
Application number
PCT/JP2020/014300
Other languages
French (fr)
Japanese (ja)
Inventor
板東　賢一
Original Assignee
株式会社小松製作所
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株式会社小松製作所 filed Critical 株式会社小松製作所
Priority to KR1020217025311A priority Critical patent/KR102598075B1/en
Priority to CN202080017865.2A priority patent/CN113518958A/en
Priority to JP2021512062A priority patent/JPWO2020203876A1/ja
Priority to DE112020000402.3T priority patent/DE112020000402T5/en
Publication of WO2020203876A1 publication Critical patent/WO2020203876A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B19/00—Programme-control systems
    • G05B19/02—Programme-control systems electric
    • G05B19/04—Programme control other than numerical control, i.e. in sequence controllers or logic controllers
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B19/00—Programme-control systems
    • G05B19/02—Programme-control systems electric
    • G05B19/18—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form
    • G05B19/404—Numerical control [NC], i.e. automatically operating machines, in particular machine tools, e.g. in a manufacturing environment, so as to execute positioning, movement or co-ordinated operations by means of programme data in numerical form characterised by control arrangements for compensation, e.g. for backlash, overshoot, tool offset, tool wear, temperature, machine construction errors, load, inertia
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D3/00—Control of position or direction
    • G05D3/12—Control of position or direction using feedback
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23Q—DETAILS, COMPONENTS, OR ACCESSORIES FOR MACHINE TOOLS, e.g. ARRANGEMENTS FOR COPYING OR CONTROLLING; MACHINE TOOLS IN GENERAL CHARACTERISED BY THE CONSTRUCTION OF PARTICULAR DETAILS OR COMPONENTS; COMBINATIONS OR ASSOCIATIONS OF METAL-WORKING MACHINES, NOT DIRECTED TO A PARTICULAR RESULT
    • B23Q15/00—Automatic control or regulation of feed movement, cutting velocity or position of tool or work
    • B23Q15/007—Automatic control or regulation of feed movement, cutting velocity or position of tool or work while the tool acts upon the workpiece
    • B23Q15/18—Compensation of tool-deflection due to temperature or force
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00—Program-control systems
    • G05B2219/30—Nc systems
    • G05B2219/49—Nc machine tool, till multiple
    • G05B2219/49176—Compensation of vibration of machine base due to slide movement

Definitions

  • the present invention relates to industrial machines, control devices, control compensators, and control methods.
  • the present application claims priority with respect to Japanese Patent Application No. 2019-068965 filed in Japan on March 29, 2019, the contents of which are incorporated herein by reference.
  • Patent Document 1 discloses a technique for performing feedforward control of a controlled object by an inverse kinematics model. Compared with feedback control, feedforward control can be expected to improve control performance because there is no delay in loop time.
  • An object of the present invention is to provide an industrial machine, a control device, a control correction device, and a control method capable of reducing the influence on the control performance due to the speeding up of processing or the miniaturization of the machine.
  • the industrial machine includes a tool for processing a work, an actuator for relatively moving the tool and the work, and a control device for controlling the behavior of the actuator.
  • the control device includes a reaction calculation unit that calculates a value related to a reaction to the action of the actuator based on a target position related to a relative position between the tool and the work, a value related to the reaction, and the target.
  • a command output unit that outputs a current command to the actuator based on the position is provided.
  • FIG. 1 is a top view showing the configuration of an industrial machine according to the first embodiment.
  • the industrial machine 100 according to the first embodiment is a grinding machine.
  • the industrial machine 100 includes a base 110, a support device 120, a tool base 130, and a control device 140.
  • the base 110 is installed on the floor of the factory.
  • the support device 120 and the tool base 130 are provided on the upper surface of the base 110.
  • the support device 120 supports both ends of the work W and rotates the work W around the main axis.
  • the tool base 130 supports the tool 131 for machining the work W supported by the support device 120.
  • the direction orthogonal to the main axis on the upper surface of the base 110 is referred to as the X direction
  • the direction in which the main axis extends is referred to as the Y direction
  • the direction orthogonal to the upper surface of the base 110 is referred to as the Z direction. That is, in the following description, the positional relationship of the industrial machine 100 will be described with reference to the three-dimensional Cartesian coordinate system including the X-axis, the Y-axis, and the Z-axis.
  • the base 110 includes a Y-axis guide portion 111 that slidably supports the tool base 130 in the Y-axis direction, and a Y-axis actuator 112 that moves the tool base 130 in the Y-axis direction along the Y-axis guide portion 111.
  • the Y-axis actuator 112 may be configured by a linear motor or a combination of a ball screw and a rotary motor.
  • the support device 120 includes a headstock 121 that supports one end of a substantially cylindrical work W, and a tailstock 122 that supports the other end.
  • the headstock 121 is provided with a rotary motor 123 that rotates the work W about an axis.
  • the tool base 130 moves the tool 131 in the X-axis direction along the tool 131, the X-axis guide portion 132 that slidably supports the tool 131 in the X-axis direction with respect to the tool base 130, and the X-axis guide portion 132. It includes an X-axis actuator 133 for rotating the tool 131 and a rotary motor 134 for rotating the tool 131.
  • the X-axis actuator 133 may be configured by a linear motor, or may be configured by a combination of a ball screw and a rotary motor.
  • the tool 131 according to the first embodiment is a grindstone.
  • the work W is supported between the headstock 121 and the tailstock 122 of the support device 120, and the outer peripheral surface of the work W is ground by the tool 131.
  • the work W include a cam and an eccentric pin.
  • FIG. 2 is a schematic block diagram showing a configuration of a control device according to the first embodiment.
  • the control device 140 controls the Y-axis actuator 112, the rotary motor 123, the X-axis actuator 133, and the rotary motor 134.
  • the control device 140 includes a processor 141, a main memory 143, a storage 145, and an interface 147.
  • the processor 141 reads a program from the storage 145, expands it into the main memory 143, and executes the above processing according to the program.
  • the program may be for realizing a part of the functions exerted by the control device 140. For example, the program may exert its function in combination with another program already stored in the storage 145 or in combination with another program mounted on another device.
  • control device 140 may include a custom LSI (Large Scale Integrated Circuit) such as a PLD (Programmable Logic Device) in addition to or in place of the above configuration.
  • PLDs include PAL (Programmable Array Logic), GAL (Generic Array Logic), CPLD (Complex Programmable Logic Device), and FPGA (Field Programmable Gate Array).
  • PLDs include PAL (Programmable Array Logic), GAL (Generic Array Logic), CPLD (Complex Programmable Logic Device), and FPGA (Field Programmable Gate Array).
  • PLDs Programmable Logic Device
  • PAL Programmable Array Logic
  • GAL Generic Array Logic
  • CPLD Complex Programmable Logic Device
  • FPGA Field Programmable Gate Array
  • Examples of storage 145 are HDD (Hard Disk Drive), SSD (Solid State Drive), magnetic disk, magneto-optical disk, CD-ROM (Compact Disc Read Only Memory), DVD-ROM (Digital Versatile Disc Read Only Memory). , Semiconductor memory and the like.
  • the storage 145 may be internal media directly connected to the bus of the control device 140, or external media connected to the control device 140 via the interface 147 or a communication line. When this program is distributed to the control device 140 by a communication line, the distributed control device 140 may expand the program to the main memory 143 and execute the above processing.
  • the storage 145 is a non-temporary tangible storage medium.
  • FIG. 3 is a diagram showing an example of modeling an industrial machine in which the relative positions of a tool and a work are displaced. For example, FIG.
  • the mass component M B of the base 110, the spring component represents the vibration characteristics between the base 110 and the floor K B and the damper component C B
  • the external force F to the X-axis actuator 133 is generated, as represented by the displacement X B of the displacement X L, as well as the supporting device 120 of the tool 131.
  • F L in FIG. 3 is a non-linear frictional force acting between the tool post 130 and the X-axis guide portion 132.
  • FIG. 3 is represented by the following equation (1).
  • B (s) is an image function obtained by Laplace transform of the characteristics of the base 110.
  • L (s) is an image function obtained by the Laplace transform of the characteristics of the tool table 130.
  • G ⁇ (s) is an image function obtained by Laplace transform of the vibration characteristic of the base 110, and corresponds to the dynamic characteristic model M1 of the base. ⁇ 1 to ⁇ 4 and w 0 to w 7 are constants.
  • the image function G ⁇ (s) of the dynamic characteristic model M1 of the base is experimentally calculated by system identification.
  • the work model M2 includes the target shape data of the work W in machining.
  • the target shape data of the work W may be represented by CAD data, or may be represented by shape data of a cross section orthogonal to the rotation axis of the work W.
  • the work model M2 includes a dynamic characteristic model of the work W. The work model M2 calculates the variation in the amount of deflection when an external force acts on the work W.
  • the reverse motion characteristic model M3 inputs the target value of the state quantity of the industrial machine 100, and when the X-axis actuator 133 is driven according to the target value, the tool 131 and the work W generated by the vibration of the base 110 generated by the reaction. This is a model for calculating the current compensation value for compensating for the amount of deviation of the relative position of.
  • the reverse motion characteristic model M3 according to the first embodiment is configured by the neural network shown in FIG.
  • FIG. 4 is a diagram showing an example of the configuration of the neural network.
  • the reverse motion characteristic model M3 is realized by, for example, a trained model of DNN (Deep Neural Network).
  • the trained model is composed of a combination of the trained model and the trained parameters. As shown in FIG.
  • the reverse motion characteristic model M3 includes an input layer M31, one or more intermediate layers M32 (hidden layers), and an output layer M33.
  • Each layer M31, M32, M33 comprises one or more neurons.
  • the number of neurons in the middle layer M32 can be set as appropriate.
  • the number of neurons in the output layer M33 is one.
  • Neurons in layers adjacent to each other are connected to each other, and a weight (connection load) is set for each connection.
  • the number of connections of neurons may be set as appropriate.
  • a threshold value is set for each neuron, and the output value of each neuron is determined by whether or not the sum of the products of the input value and the weight for each neuron exceeds the threshold value.
  • the processor 141 By executing the program, the processor 141 includes a measurement value acquisition unit 411, a target position calculation unit 412, a target state quantity calculation unit 413, a command value calculation unit 414, a reaction calculation unit 415, a compensation value calculation unit 416, and a command output unit 417. It functions as a state quantity calculation unit 418, a processing state identification unit 419, a feedback unit 420, and a learning unit 421.
  • the measured value acquisition unit 411 acquires measured values from a plurality of sensors provided in the industrial machine 100. Specifically, the measured value acquisition unit 411 acquires the torque and rotation angle of the rotary motor 123, the torque and rotation speed of the rotary motor 134, the thrust of the X-axis actuator 133, and the position of the tool 131 in the X-axis direction. The position of the tool 131 in the X-axis direction is the relative position between the tool 131 and the work W.
  • the target position calculation unit 412 calculates the position of the contour of the target shape of the work W facing the tool 131 as the target position of the tool 131 from the target shape data of the work model M2.
  • the target state amount calculation unit 413 calculates the target value of the state amount related to the displacement of the tool 131 based on the target position calculated by the target position calculation unit 412. Specifically, the target state quantity calculation unit 413 calculates the target speed, target acceleration, and target jerk value of the tool 131 in the X-axis direction. That is, the target state quantity calculation unit 413 calculates s, s 2 , and s 3 in the above equation (1).
  • the command value calculation unit 414 calculates the current command value of the X-axis actuator 133 based on the target value of the state quantity of the tool 131 and the feedback signal by the feedback unit 420. Specifically, the command value calculation unit 414 converts the target value of the state quantity of the tool 131 into a current value for achieving the target value, and adds the value of the feedback signal to the target value to obtain the current command value. Is calculated.
  • the reaction calculation unit 415 is based on the target value of the state quantity calculated by the target state quantity calculation unit 413 and the dynamic characteristic model M1 of the base, and the base 110 when the X-axis actuator 133 is driven according to the target value. Calculate the values related to the displacement, velocity, acceleration, and Laplace transform of jerk. That is, the reaction calculation unit 415 converts 1 / G ⁇ (s), s / G ⁇ (s), s 2 / G ⁇ (s), and s 3 / G ⁇ (s) in the above equation (1). calculate.
  • the displacement, velocity, acceleration, and jerk of the base 110 are state quantities related to the reaction to the action of the X-axis actuator 133.
  • the compensation value calculation unit 416 inputs the target value of the state quantity calculated by the target state quantity calculation unit 413 and the value of the state quantity related to the reaction calculated by the reaction calculation unit 415 into the reverse motion characteristic model M3, thereby causing a current. Calculate the compensation value.
  • the current compensation value is a current value required to drive the X-axis actuator 133 by the amount of deviation between the relative positions of the tool 131 and the work W caused by the vibration of the base 110.
  • the command output unit 417 outputs a current command indicating the sum of the current command value calculated by the command value calculation unit 414 and the current compensation value calculated by the compensation value calculation unit 416 to the X-axis actuator 133.
  • the state quantity calculation unit 418 calculates the deflection amount fluctuation data of the work W based on the measurement value acquired by the measurement value acquisition unit 411 and the work model M2.
  • the machining state specifying unit 419 calculates the shape error fluctuation of the work W based on the measured value acquired by the measured value acquisition unit 411 and the deflection amount fluctuation data of the work W calculated by the state quantity calculation unit 418.
  • the feedback unit 420 is based on the target value of the state quantity related to the displacement of the tool 131 calculated by the target state quantity calculation unit 413 and the measured value of the position of the tool 131 in the X-axis direction acquired by the measurement value acquisition unit 411.
  • a feedback signal related to driving the X-axis actuator 133 is output.
  • the shape error variation specified by the machining state specifying unit 419 may be superimposed on the measured value of the position of the tool 131 in the X-axis direction acquired by the measured value acquiring unit 411. In this case, it is possible to output a feedback signal related to the drive of the X-axis actuator 133 in consideration of the fluctuation of the deflection amount of the work W.
  • the feedback unit 420 generates a feedback signal by sliding mode control.
  • the sliding mode control is a control method for switching the feedback signal based on the hyperplane defined by the state quantity of the industrial machine 100 or the work W.
  • the value of the feedback signal related to the sliding mode control is non-linear with respect to the control deviation and the shape error fluctuation.
  • the learning unit 421 receives the target value of the state amount related to the displacement of the tool 131 calculated by the target state amount calculation unit 413, the value of the state amount related to the reaction calculated by the reaction calculation unit 415, and the feedback signal by the feedback unit 420.
  • the feedback characteristic model M3 is trained as a training data set. Specifically, the target value of the state amount calculated by the target state amount calculation unit 413 and the value of the state amount related to the reaction calculated by the reaction calculation unit 415 are input to the input layer M31 of the reverse motion characteristic model M3. .. The value of the feedback signal generated by the feedback unit 420 is input to the output layer M33.
  • the trained parameters of the reverse motion characteristic model M3 obtained by learning are stored in the storage 145.
  • the learned parameters include, for example, the number of layers of the reverse motion characteristic model M3, the number of neurons in each layer, the connection relationship between neurons, the weight of connection between each neuron, and the threshold value of each neuron.
  • FIG. 5 is a block diagram showing the operation of the control device according to the first embodiment.
  • FIG. 6 is a flowchart showing the operation of the control device according to the first embodiment.
  • the target position calculation unit 412 calculates the target position R of the tool 131 based on the work model M2 (step S2).
  • Target state quantity calculation unit 413 calculates a target value of the state quantity relating to the displacement X L of the tool 131 (velocity, acceleration, jerk) based on the target position R calculated in step S2 (step S3).
  • the command value calculation unit 414 converts the target value of the state quantity of the tool 131 into a current value (torque value) for achieving the target value (step S4).
  • the command value calculation unit 414 calculates the current command value V by adding the value of the feedback signal UFB output from the feedback unit 420 to the current value converted in step S4 (step S5).
  • the reaction calculation unit 415 determines the displacement, velocity, acceleration, and displacement of the base 110 caused by the reaction of the drive of the X-axis actuator 133 based on the target value of the state quantity calculated in step S3 and the dynamic characteristic model M1 of the base. A value related to the Laplace transform of jerk is calculated (step S6). Compensation value calculating unit 416, the value of the state quantity relating to the reaction was calculated by the target value and the step S6 of the calculated state quantity at step S3, by entering the reversing characteristic model M3, calculates a current compensation value U NN (Step S7).
  • the command output unit 417 outputs a current command U (thrust command) indicating the sum of the current command value V calculated in step S5 and the current compensation value UNN calculated in step S7 to the X-axis actuator 133 (thrust command). Step S8).
  • the state quantity calculation unit 418 calculates the deflection amount fluctuation data of the work W based on the measured value Y acquired in step S1 and the work model M2 (step S9).
  • the machining state specifying unit 419 calculates the shape error fluctuation of the work W based on the measured value acquired in step S1 and the deflection amount fluctuation data of the work W calculated in step S9 (step S10).
  • the feedback unit 420 transmits the feedback signal UFB related to the drive of the X-axis actuator 133 based on the target value of the state quantity calculated in step S3 and the measured value Y of the position of the tool 131 in the X-axis direction acquired in step S1.
  • Output (step S11) The shape error variation specified in step S10 may be superimposed on the measured value of the position of the tool 131 in the X-axis direction acquired in step S1 above.
  • the learning unit 421 uses a combination of the target value of the state quantity calculated in step S3, the value of the state quantity related to the reaction calculated in step S6, and the value of the feedback signal UFB output in step S11 as a learning data set. , The feedback characteristic model M3 is trained and the learned parameters are updated (step S12). Regarding step S12, either online learning or offline learning may be performed.
  • the control device 140 determines whether or not the machining operation of the work W is completed (step S13). If the machining operation is not completed (step S13: NO), the process is returned to step S1 and the machining control is continued. On the other hand, when the machining operation is completed (step S13: YES), the control device 140 ends the machining control.
  • control performance by the control device 140 according to the first embodiment is compared with the control performance by the control device 140 (control device 140 according to the comparative example) when the correction by the current compensation value is not performed.
  • FIG. 7A is a graph showing a time series of feedback signals output by the control device according to the comparative example.
  • FIG. 7B is a graph showing a time series of contour errors in machining control by the control device according to the comparative example.
  • the control device 140 according to the comparative example is the control device 140 according to the first embodiment in a state where the reverse motion characteristic model M3 has not been learned. That is, in the control device 140 according to the comparative example, the current compensation value calculated by the compensation value calculation unit 416 is always zero. Therefore, as shown in FIG. 7A, the amount of deviation between the relative positions of the tool 131 and the work W caused by the vibration of the base 110 is compensated by the feedback signal by the feedback unit 420. On the other hand, since the feedback control is delayed by the loop time, a contour error occurs as shown in FIG. 7B.
  • FIG. 8A is a graph showing a time series of the current compensation value and the feedback signal output by the control device according to the first embodiment.
  • FIG. 8B is a graph showing a time series of contour errors in machining control by the control device according to the first embodiment.
  • the graphs shown in FIGS. 8A and 8B show the control performance of the control device 140 according to the first embodiment in a state where the reverse motion characteristic model M3 is learned by performing the machining operation of the work W three times.
  • the compensation value calculation unit 416 generates the current compensation value.
  • the X-axis actuator 133 is driven in accordance with the vibration of the base 110, so that the amount of deviation between the relative positions of the tool 131 and the work W is reduced. That is, the current compensation value acts as a feedforward signal. As a result, the magnitude of the feedback signal by the feedback unit 420 becomes smaller than that of the comparative example. Further, as shown in FIG. 8B, the feedforward control by the compensation value calculation unit 416 reduces the amount of deviation between the relative positions of the tool 131 and the work W, so that the contour error can be reduced as compared with the comparative example. ..
  • the control device 140 calculates the value related to the reaction to the action of the X-axis actuator 133 based on the target position related to the relative position between the tool 131 and the work W, and the value is concerned.
  • a current command is output to the X-axis actuator 133 based on the value related to the reaction and the target position.
  • the control device 140 can control the relative positions of the tool 131 and the work W so as to follow the reaction of the X-axis actuator 133. Therefore, the control device 140 according to the first embodiment can reduce the influence on the control performance such as the contour control performance due to the speeding up of processing or the miniaturization of the industrial machine.
  • the control device 140 calculates the current compensation value based on the value related to the reaction, and relates to the sum of the current command value calculated by the command value calculation unit 414 and the current compensation value.
  • the current command is output to the X-axis actuator 133. That is, according to the control method according to the first embodiment, the processing speed is increased by adding the functions of the reaction calculation unit 415 and the compensation value calculation unit 416 to the existing control device 140 that does not consider the influence of the reaction. Alternatively, it is possible to realize control that reduces the influence on the contour control performance due to the miniaturization of the industrial machine.
  • the industrial machine 100 includes a control device 140 having a reaction calculation unit 415 and a command value calculation unit 414, but the control device 140 of the industrial machine 100 according to another embodiment is a conventional control device 140. It may be realized by a combination of a control device and a control correction device having a reaction calculation unit 415, a compensation value calculation unit 416, and a command output unit 417. On the other hand, in another embodiment, the control device 140 may directly calculate the current command value including the current compensation value based on the reverse motion characteristic model M3.
  • the control device 140 trains the reverse motion characteristic model M3 using the feedback signal.
  • the learning process can be automatically performed without manually generating teacher data.
  • the feedback signal according to the first embodiment is generated by the sliding mode control. Since the feedback signal generated by the sliding mode control is non-linear and highly responsive, the control device 140 can be learned at high speed and appropriately.
  • the reverse motion characteristic model M3 may be trained based on a linear feedback signal such as PID control. In another embodiment, the reverse motion characteristic model M3 may be trained using the teacher data manually generated by the operator.
  • the industrial machine 100 according to the first embodiment is a grinding machine, but is not limited to this.
  • the industrial machine 100 according to another embodiment may be another machine tool such as a press machine, a milling machine, or a lathe.
  • control device 140 simultaneously performs the machining operation of the work W and the learning of the reverse motion characteristic model M3, but is not limited to this.
  • learning of the reverse motion characteristic model M3 may be performed in advance, and learning by the learning unit 421 may not be performed after a certain level of performance can be expected.
  • the reverse motion characteristic model M3 may be learned by the learning industrial machine 100, and the learned reverse motion characteristic model M3 may be transferred to another industrial machine 100.
  • control device 140 learns the reverse motion characteristic model M3 for the deviation of the relative position between the tool and the work due to the reaction of the drive, but is not limited to this.
  • the reverse motion characteristic model M3 may be learned for vibration, non-linear friction, quadrant protrusion, etc. other than the base due to the action of driving. Good.
  • control device 140 calculates the deflection amount fluctuation data of the work W, but the present invention is not limited to this.
  • the control device according to another embodiment may calculate the chatter amount fluctuation data of the work W in place of or in addition to the deflection amount fluctuation data of the work W.

Abstract

According to the present invention, a counteraction calculation unit calculates, on the basis of target positions for the relative positions between a tool and a workpiece, a value relating to a counteraction of an actuator that displaces the relative positions between the tool and the workpiece. A command output unit outputs an electric current command to the actuator on the basis of the counteraction value and the target positions.

Description

産業機械、制御装置、制御補正装置、および制御方法Industrial machinery, control devices, control compensators, and control methods
　本発明は、産業機械、制御装置、制御補正装置、および制御方法に関する。
　本願は、２０１９年３月２９日に日本に出願された特願２０１９－０６８９６５号について優先権を主張し、その内容をここに援用する。 The present invention relates to industrial machines, control devices, control compensators, and control methods.
The present application claims priority with respect to Japanese Patent Application No. 2019-068965 filed in Japan on March 29, 2019, the contents of which are incorporated herein by reference.
　特許文献１には、逆運動学モデルによって制御対象のフィードフォワード制御を行う技術が開示されている。フィードフォワード制御は、フィードバック制御と比較して、ループ時間の遅れがない分、制御性能を高めることが期待できる。 Patent Document 1 discloses a technique for performing feedforward control of a controlled object by an inverse kinematics model. Compared with feedback control, feedforward control can be expected to improve control performance because there is no delay in loop time.
特開平６－３２０４５１号公報Japanese Patent Application Laid-Open No. 6-320451
　ところで、工具を用いてワークを加工する産業機械においては、加工の高速化および機械の小型化が望まれている。加工を高速化するためには、工具およびワークの駆動を高速化することが必要となる。他方、機械を小型化すると産業機械自体の重量が軽くなり、また産業機械の基台と床面との接地面積が小さくなる。産業機械を駆動させると、駆動の反作用によって産業機械の基台の振動が生じる。基台の振動は工具およびワークの駆動が速いほど大きくなる。この振動によって、工具とワークとの相対位置にずれが生じ、輪郭制御性能等の制御性能が低下する。加工の高速化および機械の小型化を図ると、制御性能が低下してしまう。 By the way, in an industrial machine that processes a workpiece using a tool, it is desired to increase the processing speed and reduce the size of the machine. In order to speed up machining, it is necessary to speed up the driving of tools and workpieces. On the other hand, when the machine is miniaturized, the weight of the industrial machine itself becomes lighter, and the contact area between the base of the industrial machine and the floor surface becomes smaller. When an industrial machine is driven, the reaction of the drive causes vibration of the base of the industrial machine. The vibration of the base increases as the tool and workpiece are driven faster. Due to this vibration, the relative position between the tool and the work is displaced, and the control performance such as the contour control performance is deteriorated. If the processing speed is increased and the machine is downsized, the control performance will deteriorate.
　本発明の目的は、加工の高速化または機械の小型化による制御性能への影響を低減することができる産業機械、制御装置、制御補正装置、および制御方法を提供することにある。 An object of the present invention is to provide an industrial machine, a control device, a control correction device, and a control method capable of reducing the influence on the control performance due to the speeding up of processing or the miniaturization of the machine.
　本発明の一態様によれば、産業機械は、ワークを加工するための工具と、前記工具と前記ワークを相対的に移動させるアクチュエータと、前記アクチュエータの挙動を制御する制御装置とを備える産業機械であって、前記制御装置は、前記工具と前記ワークの相対位置に係る目標位置に基づいて、前記アクチュエータの作用に対する反作用に係る値を算出する反作用算出部と、前記反作用に係る値と前記目標位置とに基づいて、前記アクチュエータに電流指令を出力する指令出力部とを備える。 According to one aspect of the present invention, the industrial machine includes a tool for processing a work, an actuator for relatively moving the tool and the work, and a control device for controlling the behavior of the actuator. The control device includes a reaction calculation unit that calculates a value related to a reaction to the action of the actuator based on a target position related to a relative position between the tool and the work, a value related to the reaction, and the target. A command output unit that outputs a current command to the actuator based on the position is provided.
　上記態様によれば、加工の高速化または産業機械の小型化による制御性能への影響を低減することができる。 According to the above aspect, it is possible to reduce the influence on the control performance due to the speeding up of processing or the miniaturization of industrial machines.
第１の実施形態に係る産業機械の構成を示す上面図である。It is a top view which shows the structure of the industrial machine which concerns on 1st Embodiment. 第１の実施形態に係る制御装置の構成を示す概略ブロック図である。It is a schematic block diagram which shows the structure of the control device which concerns on 1st Embodiment. 産業機械のモデリングの例を示す図である。It is a figure which shows the example of modeling of an industrial machine. ニューラルネットワークの構成の例を示す図である。It is a figure which shows the example of the structure of the neural network. 第１の実施形態に係る制御装置の動作を示すブロック線図である。It is a block diagram which shows the operation of the control device which concerns on 1st Embodiment. 第１の実施形態に係る制御装置の動作を示すフローチャートである。It is a flowchart which shows the operation of the control device which concerns on 1st Embodiment. 比較例に係る制御装置が出力するフィードバック信号の時系列を示すグラフである。It is a graph which shows the time series of the feedback signal output by the control device which concerns on a comparative example. 比較例に係る制御装置による加工制御における輪郭誤差の時系列を示すグラフである。It is a graph which shows the time series of the contour error in the processing control by the control device which concerns on a comparative example. 第１の実施形態に係る制御装置が出力するフィードバック信号の時系列を示すグラフである。It is a graph which shows the time series of the feedback signal output by the control device which concerns on 1st Embodiment. 第１の実施形態に係る制御装置による加工制御における輪郭誤差の時系列を示すグラフである。It is a graph which shows the time series of the contour error in the processing control by the control device which concerns on 1st Embodiment.
〈第１の実施形態〉
　以下、図面を参照しながら実施形態について詳しく説明する。
　図１は、第１の実施形態に係る産業機械の構成を示す上面図である。第１の実施形態に係る産業機械１００は、研削盤である。 <First Embodiment>
Hereinafter, embodiments will be described in detail with reference to the drawings.
FIG. 1 is a top view showing the configuration of an industrial machine according to the first embodiment. The industrial machine 100 according to the first embodiment is a grinding machine.
《産業機械の構成》
　産業機械１００は、基台１１０、支持装置１２０、工具台１３０、制御装置１４０を備える。基台１１０は、工場の床面に設置される。支持装置１２０および工具台１３０は、基台１１０の上面に設けられる。支持装置１２０は、ワークＷの両端を支持し、ワークＷを主軸回りに回転させる。工具台１３０は、支持装置１２０に支持されたワークＷを加工するための工具１３１を支持する。
　以下、基台１１０の上面において主軸と直交する方向をＸ方向とよび、主軸の伸びる方向をＹ方向とよび、基台１１０の上面に直交する方向をＺ方向と呼ぶ。すなわち、以下の説明においては、Ｘ軸、Ｙ軸、およびＺ軸からなる三次元直交座標系を参照しながら産業機械１００の位置関係を説明する。 << Composition of industrial machinery >>
The industrial machine 100 includes a base 110, a support device 120, a tool base 130, and a control device 140. The base 110 is installed on the floor of the factory. The support device 120 and the tool base 130 are provided on the upper surface of the base 110. The support device 120 supports both ends of the work W and rotates the work W around the main axis. The tool base 130 supports the tool 131 for machining the work W supported by the support device 120.
Hereinafter, the direction orthogonal to the main axis on the upper surface of the base 110 is referred to as the X direction, the direction in which the main axis extends is referred to as the Y direction, and the direction orthogonal to the upper surface of the base 110 is referred to as the Z direction. That is, in the following description, the positional relationship of the industrial machine 100 will be described with reference to the three-dimensional Cartesian coordinate system including the X-axis, the Y-axis, and the Z-axis.
　基台１１０には、工具台１３０をＹ軸方向にスライド可能に支持するＹ軸ガイド部１１１と、Ｙ軸ガイド部１１１に沿って工具台１３０をＹ軸方向に移動させるＹ軸アクチュエータ１１２とを備える。Ｙ軸アクチュエータ１１２は、直動モータによって構成されてもよいし、ボールねじと回転モータとの組み合わせによって構成されてもよい。 The base 110 includes a Y-axis guide portion 111 that slidably supports the tool base 130 in the Y-axis direction, and a Y-axis actuator 112 that moves the tool base 130 in the Y-axis direction along the Y-axis guide portion 111. Be prepared. The Y-axis actuator 112 may be configured by a linear motor or a combination of a ball screw and a rotary motor.
　支持装置１２０は、略円筒状のワークＷの一端を支持する主軸台１２１と、他端を支持する芯押し台１２２とを備える。主軸台１２１には、ワークＷを軸回りに回転させる回転モータ１２３を備える。 The support device 120 includes a headstock 121 that supports one end of a substantially cylindrical work W, and a tailstock 122 that supports the other end. The headstock 121 is provided with a rotary motor 123 that rotates the work W about an axis.
　工具台１３０は、工具１３１と、工具台１３０に対して工具１３１をＸ軸方向にスライド可能に支持するＸ軸ガイド部１３２と、Ｘ軸ガイド部１３２に沿って工具１３１をＸ軸方向に移動させるＸ軸アクチュエータ１３３と、工具１３１を回転させる回転モータ１３４とを備える。Ｘ軸アクチュエータ１３３は、直動モータによって構成されてもよいし、ボールねじと回転モータとの組み合わせによって構成されてもよい。第１の実施形態に係る工具１３１は、砥石である。 The tool base 130 moves the tool 131 in the X-axis direction along the tool 131, the X-axis guide portion 132 that slidably supports the tool 131 in the X-axis direction with respect to the tool base 130, and the X-axis guide portion 132. It includes an X-axis actuator 133 for rotating the tool 131 and a rotary motor 134 for rotating the tool 131. The X-axis actuator 133 may be configured by a linear motor, or may be configured by a combination of a ball screw and a rotary motor. The tool 131 according to the first embodiment is a grindstone.
　すなわち、第１の実施形態に係る産業機械１００では、支持装置１２０の主軸台１２１および芯押し台１２２の間にワークＷを支持し、工具１３１によってワークＷの外周面を研削加工する。ワークＷの例としては、カムおよび偏心ピンなどが挙げられる。 That is, in the industrial machine 100 according to the first embodiment, the work W is supported between the headstock 121 and the tailstock 122 of the support device 120, and the outer peripheral surface of the work W is ground by the tool 131. Examples of the work W include a cam and an eccentric pin.
《制御装置の構成》
　図２は、第１の実施形態に係る制御装置の構成を示す概略ブロック図である。
　制御装置１４０は、Ｙ軸アクチュエータ１１２、回転モータ１２３、Ｘ軸アクチュエータ１３３、および回転モータ１３４を制御する。制御装置１４０は、プロセッサ１４１、メインメモリ１４３、ストレージ１４５、インタフェース１４７を備える。プロセッサ１４１は、プログラムをストレージ１４５から読み出してメインメモリ１４３に展開し、当該プログラムに従って上記処理を実行する。
　プログラムは、制御装置１４０に発揮させる機能の一部を実現するためのものであってもよい。例えば、プログラムは、ストレージ１４５に既に記憶されている他のプログラムとの組み合わせ、または他の装置に実装された他のプログラムとの組み合わせによって機能を発揮させるものであってもよい。なお、他の実施形態においては、制御装置１４０は、上記構成に加えて、または上記構成に代えてＰＬＤ（Programmable Logic Device）などのカスタムＬＳＩ（Large Scale Integrated Circuit）を備えてもよい。ＰＬＤの例としては、ＰＡＬ(Programmable Array Logic)、ＧＡＬ(Generic Array Logic)、ＣＰＬＤ(Complex Programmable Logic Device)、ＦＰＧＡ（Field Programmable Gate Array）が挙げられる。この場合、プロセッサ１４１によって実現される機能の一部または全部が当該集積回路によって実現されてよい。 << Configuration of control device >>
FIG. 2 is a schematic block diagram showing a configuration of a control device according to the first embodiment.
The control device 140 controls the Y-axis actuator 112, the rotary motor 123, the X-axis actuator 133, and the rotary motor 134. The control device 140 includes a processor 141, a main memory 143, a storage 145, and an interface 147. The processor 141 reads a program from the storage 145, expands it into the main memory 143, and executes the above processing according to the program.
The program may be for realizing a part of the functions exerted by the control device 140. For example, the program may exert its function in combination with another program already stored in the storage 145 or in combination with another program mounted on another device. In another embodiment, the control device 140 may include a custom LSI (Large Scale Integrated Circuit) such as a PLD (Programmable Logic Device) in addition to or in place of the above configuration. Examples of PLDs include PAL (Programmable Array Logic), GAL (Generic Array Logic), CPLD (Complex Programmable Logic Device), and FPGA (Field Programmable Gate Array). In this case, some or all of the functions realized by the processor 141 may be realized by the integrated circuit.
　ストレージ１４５の例としては、ＨＤＤ（Hard Disk Drive）、ＳＳＤ（Solid State Drive）、磁気ディスク、光磁気ディスク、ＣＤ－ＲＯＭ（Compact Disc Read Only Memory）、ＤＶＤ－ＲＯＭ（Digital Versatile Disc Read Only Memory）、半導体メモリ等が挙げられる。ストレージ１４５は、制御装置１４０のバスに直接接続された内部メディアであってもよいし、インタフェース１４７または通信回線を介して制御装置１４０に接続される外部メディアであってもよい。また、このプログラムが通信回線によって制御装置１４０に配信される場合、配信を受けた制御装置１４０が当該プログラムをメインメモリ１４３に展開し、上記処理を実行してもよい。少なくとも１つの実施形態において、ストレージ１４５は、一時的でない有形の記憶媒体である。 Examples of storage 145 are HDD (Hard Disk Drive), SSD (Solid State Drive), magnetic disk, magneto-optical disk, CD-ROM (Compact Disc Read Only Memory), DVD-ROM (Digital Versatile Disc Read Only Memory). , Semiconductor memory and the like. The storage 145 may be internal media directly connected to the bus of the control device 140, or external media connected to the control device 140 via the interface 147 or a communication line. When this program is distributed to the control device 140 by a communication line, the distributed control device 140 may expand the program to the main memory 143 and execute the above processing. In at least one embodiment, the storage 145 is a non-temporary tangible storage medium.
　ストレージ１４５には、基台の動特性モデルＭ１、ワークモデルＭ２、および逆動特性モデルＭ３が記憶される。
　産業機械を駆動させると、駆動の反作用によって産業機械の基台の振動が生じる。この振動によって、工具とワークとの相対位置にずれが生じ、輪郭制御性能が低下する。図３は、工具とワークとの相対位置にずれが生じる産業機械のモデリングの例を示す図である。例えば、図３は、基台１１０の質量成分ＭＢ、基台１１０と床面との間の振動特性を表すばね成分ＫＢおよびダンパ成分ＣＢ、工具台１３０の質量成分ＭＬ、工具台１３０と基台１１０との間のダンパ成分ＣＬ、Ｘ軸アクチュエータ１３３が発生する外力Ｆ、工具１３１の変位ＸＬ、ならびに支持装置１２０の変位ＸＢによって表される。図３のＦＬは、工具台１３０とＸ軸ガイド部１３２の間に働く非線形摩擦力である。具体的には、図３は、以下の式（１）によって表される。 The base dynamic characteristic model M1, the work model M2, and the reverse dynamic characteristic model M3 are stored in the storage 145.
When an industrial machine is driven, the reaction of the drive causes vibration of the base of the industrial machine. Due to this vibration, the relative position between the tool and the work is displaced, and the contour control performance is deteriorated. FIG. 3 is a diagram showing an example of modeling an industrial machine in which the relative positions of a tool and a work are displaced. For example, FIG. 3, the mass component M B of the base 110, the spring component represents the vibration characteristics between the base 110 and the floor K B and the damper component C B, the mass component of the tool post 130 M L, tool post 130 and damper components C L between the base 110, the external force F to the X-axis actuator 133 is generated, as represented by the displacement X B of the displacement X L, as well as the supporting device 120 of the tool 131. F L in FIG. 3 is a non-linear frictional force acting between the tool post 130 and the X-axis guide portion 132. Specifically, FIG. 3 is represented by the following equation (1).
Figure JPOXMLDOC01-appb-M000001
Figure JPOXMLDOC01-appb-M000001
　なお、Ｂ（ｓ）は、基台１１０の特性のラプラス変換によって得られる像関数である。Ｌ（ｓ）は、工具台１３０の特性のラプラス変換によって得られる像関数である。ＧΩ（ｓ）は、基台１１０の振動特性のラプラス変換によって得られる像関数であり、基台の動特性モデルＭ１に相当する。α１～α４、ｗ０～ｗ７は、定数である。基台の動特性モデルＭ１の像関数ＧΩ（ｓ）は、システム同定によって実験的に算出される。 B (s) is an image function obtained by Laplace transform of the characteristics of the base 110. L (s) is an image function obtained by the Laplace transform of the characteristics of the tool table 130. G Ω (s) is an image function obtained by Laplace transform of the vibration characteristic of the base 110, and corresponds to the dynamic characteristic model M1 of the base. α 1 to α 4 and w 0 to w 7 are constants. The image function G Ω (s) of the dynamic characteristic model M1 of the base is experimentally calculated by system identification.
　ワークモデルＭ２は、加工におけるワークＷの目標形状データを含む。ワークＷの目標形状データは、ＣＡＤデータによって表されてもよいし、ワークＷの回転軸に直交する断面の形状データによって表されてもよい。また、ワークモデルＭ２は、ワークＷの動特性モデルを含む。ワークモデルＭ２によって、ワークＷに外力が作用したときの撓み量変動が算出される。 The work model M2 includes the target shape data of the work W in machining. The target shape data of the work W may be represented by CAD data, or may be represented by shape data of a cross section orthogonal to the rotation axis of the work W. Further, the work model M2 includes a dynamic characteristic model of the work W. The work model M2 calculates the variation in the amount of deflection when an external force acts on the work W.
　逆動特性モデルＭ３は、産業機械１００の状態量の目標値を入力し、当該目標値に従ってＸ軸アクチュエータ１３３が駆動したときに、その反作用によって生じる基台１１０の振動によって生じる工具１３１とワークＷの相対位置のずれ量を補償するための電流補償値を算出するモデルである。第１の実施形態に係る逆動特性モデルＭ３は、図４に示すニューラルネットワークによって構成される。図４は、ニューラルネットワークの構成の例を示す図である。逆動特性モデルＭ３は、例えば、ＤＮＮ（Deep Neural Network）の学習済みモデルによって実現される。学習済みモデルは、学習モデルと学習済みパラメータの組み合わせによって構成される。
　図４に示すように、逆動特性モデルＭ３は、入力層Ｍ３１、１つまたは複数の中間層Ｍ３２（隠れ層）、及び出力層Ｍ３３を含む。各層Ｍ３１、Ｍ３２、Ｍ３３は、１又は複数のニューロンを備えている。中間層Ｍ３２のニューロンの数は、適宜設定することができる。出力層Ｍ３３のニューロンの数は、１つである。 The reverse motion characteristic model M3 inputs the target value of the state quantity of the industrial machine 100, and when the X-axis actuator 133 is driven according to the target value, the tool 131 and the work W generated by the vibration of the base 110 generated by the reaction. This is a model for calculating the current compensation value for compensating for the amount of deviation of the relative position of. The reverse motion characteristic model M3 according to the first embodiment is configured by the neural network shown in FIG. FIG. 4 is a diagram showing an example of the configuration of the neural network. The reverse motion characteristic model M3 is realized by, for example, a trained model of DNN (Deep Neural Network). The trained model is composed of a combination of the trained model and the trained parameters.
As shown in FIG. 4, the reverse motion characteristic model M3 includes an input layer M31, one or more intermediate layers M32 (hidden layers), and an output layer M33. Each layer M31, M32, M33 comprises one or more neurons. The number of neurons in the middle layer M32 can be set as appropriate. The number of neurons in the output layer M33 is one.
　互いに隣接する層のニューロン同士は結合されており、各結合には重み（結合荷重）が設定されている。ニューロンの結合数は、適宜設定されてよい。各ニューロンには閾値が設定されており、各ニューロンへの入力値と重みとの積の和が閾値を超えているか否かによって各ニューロンの出力値が決定される。 Neurons in layers adjacent to each other are connected to each other, and a weight (connection load) is set for each connection. The number of connections of neurons may be set as appropriate. A threshold value is set for each neuron, and the output value of each neuron is determined by whether or not the sum of the products of the input value and the weight for each neuron exceeds the threshold value.
　プロセッサ１４１は、プログラムの実行により、計測値取得部４１１、目標位置算出部４１２、目標状態量算出部４１３、指令値算出部４１４、反作用算出部４１５、補償値算出部４１６、指令出力部４１７、状態量算出部４１８、加工状態特定部４１９、フィードバック部４２０、学習部４２１として機能する。 By executing the program, the processor 141 includes a measurement value acquisition unit 411, a target position calculation unit 412, a target state quantity calculation unit 413, a command value calculation unit 414, a reaction calculation unit 415, a compensation value calculation unit 416, and a command output unit 417. It functions as a state quantity calculation unit 418, a processing state identification unit 419, a feedback unit 420, and a learning unit 421.
　計測値取得部４１１は産業機械１００に設けられた複数のセンサから計測値を取得する。具体的には、計測値取得部４１１は、回転モータ１２３のトルクおよび回転角、回転モータ１３４のトルクおよび回転数、Ｘ軸アクチュエータ１３３の推力、工具１３１のＸ軸方向における位置を取得する。工具１３１のＸ軸方向における位置は、工具１３１とワークＷの相対位置である。 The measured value acquisition unit 411 acquires measured values from a plurality of sensors provided in the industrial machine 100. Specifically, the measured value acquisition unit 411 acquires the torque and rotation angle of the rotary motor 123, the torque and rotation speed of the rotary motor 134, the thrust of the X-axis actuator 133, and the position of the tool 131 in the X-axis direction. The position of the tool 131 in the X-axis direction is the relative position between the tool 131 and the work W.
　目標位置算出部４１２は、ワークモデルＭ２の目標形状データから、工具１３１と対向するワークＷの目標形状の輪郭の位置を、工具１３１の目標位置として算出する。 The target position calculation unit 412 calculates the position of the contour of the target shape of the work W facing the tool 131 as the target position of the tool 131 from the target shape data of the work model M2.
　目標状態量算出部４１３は、目標位置算出部４１２が算出した目標位置に基づいて工具１３１の変位に係る状態量の目標値を算出する。具体的には、目標状態量算出部４１３は、工具１３１のＸ軸方向の目標速度、目標加速度、および目標ジャークの値を算出する。すなわち、目標状態量算出部４１３は、上記の式（１）におけるｓ、ｓ２、およびｓ３を算出する。 The target state amount calculation unit 413 calculates the target value of the state amount related to the displacement of the tool 131 based on the target position calculated by the target position calculation unit 412. Specifically, the target state quantity calculation unit 413 calculates the target speed, target acceleration, and target jerk value of the tool 131 in the X-axis direction. That is, the target state quantity calculation unit 413 calculates s, s 2 , and s 3 in the above equation (1).
　指令値算出部４１４は、工具１３１の状態量の目標値とフィードバック部４２０によるフィードバック信号とに基づいてＸ軸アクチュエータ１３３の電流指令値を算出する。具体的には、指令値算出部４１４は、工具１３１の状態量の目標値を当該目標値を達成するための電流値に変換し、これにフィードバック信号の値を加算することで、電流指令値を算出する。 The command value calculation unit 414 calculates the current command value of the X-axis actuator 133 based on the target value of the state quantity of the tool 131 and the feedback signal by the feedback unit 420. Specifically, the command value calculation unit 414 converts the target value of the state quantity of the tool 131 into a current value for achieving the target value, and adds the value of the feedback signal to the target value to obtain the current command value. Is calculated.
　反作用算出部４１５は、目標状態量算出部４１３が算出した状態量の目標値と、基台の動特性モデルＭ１とに基づいて、目標値に従ってＸ軸アクチュエータ１３３を駆動させたときの基台１１０の変位、速度、加速度、およびジャークのラプラス変換に係る値を算出する。すなわち、反作用算出部４１５は、上記の式（１）における１／ＧΩ（ｓ）、ｓ／ＧΩ（ｓ）、ｓ２／ＧΩ（ｓ）、およびｓ３／ＧΩ（ｓ）を算出する。基台１１０の変位、速度、加速度、およびジャークは、Ｘ軸アクチュエータ１３３の作用に対する反作用に係る状態量である。 The reaction calculation unit 415 is based on the target value of the state quantity calculated by the target state quantity calculation unit 413 and the dynamic characteristic model M1 of the base, and the base 110 when the X-axis actuator 133 is driven according to the target value. Calculate the values related to the displacement, velocity, acceleration, and Laplace transform of jerk. That is, the reaction calculation unit 415 converts 1 / G Ω (s), s / G Ω (s), s 2 / G Ω (s), and s 3 / G Ω (s) in the above equation (1). calculate. The displacement, velocity, acceleration, and jerk of the base 110 are state quantities related to the reaction to the action of the X-axis actuator 133.
　補償値算出部４１６は、目標状態量算出部４１３が算出した状態量の目標値および反作用算出部４１５が算出した反作用に係る状態量の値を、逆動特性モデルＭ３に入力することで、電流補償値を算出する。電流補償値は、基台１１０の振動によって生じる工具１３１とワークＷの相対位置のずれ量だけＸ軸アクチュエータ１３３を駆動させるために必要な電流値である。 The compensation value calculation unit 416 inputs the target value of the state quantity calculated by the target state quantity calculation unit 413 and the value of the state quantity related to the reaction calculated by the reaction calculation unit 415 into the reverse motion characteristic model M3, thereby causing a current. Calculate the compensation value. The current compensation value is a current value required to drive the X-axis actuator 133 by the amount of deviation between the relative positions of the tool 131 and the work W caused by the vibration of the base 110.
　指令出力部４１７は、指令値算出部４１４が算出した電流指令値と補償値算出部４１６が算出した電流補償値との和の値を示す電流指令を、Ｘ軸アクチュエータ１３３に出力する。 The command output unit 417 outputs a current command indicating the sum of the current command value calculated by the command value calculation unit 414 and the current compensation value calculated by the compensation value calculation unit 416 to the X-axis actuator 133.
　状態量算出部４１８は、計測値取得部４１１が取得した計測値とワークモデルＭ２に基づいて、ワークＷの撓み量変動データを算出する。 The state quantity calculation unit 418 calculates the deflection amount fluctuation data of the work W based on the measurement value acquired by the measurement value acquisition unit 411 and the work model M2.
　加工状態特定部４１９は、計測値取得部４１１が取得した計測値と、状態量算出部４１８が算出したワークＷの撓み量変動データとに基づいて、ワークＷの形状誤差変動を算出する。 The machining state specifying unit 419 calculates the shape error fluctuation of the work W based on the measured value acquired by the measured value acquisition unit 411 and the deflection amount fluctuation data of the work W calculated by the state quantity calculation unit 418.
　フィードバック部４２０は、目標状態量算出部４１３が算出した工具１３１の変位に係る状態量の目標値と、計測値取得部４１１が取得した工具１３１のＸ軸方向における位置の計測値に基づいて、Ｘ軸アクチュエータ１３３の駆動に係るフィードバック信号を出力する。上記の計測値取得部４１１が取得した工具１３１のＸ軸方向における位置の計測値には、加工状態特定部４１９が特定した形状誤差変動を重畳してもよい。この場合、ワークＷの撓み量変動までも考慮したＸ軸アクチュエータ１３３の駆動に係るフィードバック信号を出力することができる。第１の実施形態に係るフィードバック部４２０は、スライディングモード制御によってフィードバック信号を生成する。スライディングモード制御とは、産業機械１００またはワークＷの状態量で定義される超平面に基づいてフィードバック信号を切り換える制御方法である。スライディングモード制御に係るフィードバック信号の値は、制御偏差や形状誤差変動に対して非線形である。スライディングモード制御によるフィードバック制御を行うことで、比例制御等によるフィードバック制御と比較して高い追従性を得ることができる。 The feedback unit 420 is based on the target value of the state quantity related to the displacement of the tool 131 calculated by the target state quantity calculation unit 413 and the measured value of the position of the tool 131 in the X-axis direction acquired by the measurement value acquisition unit 411. A feedback signal related to driving the X-axis actuator 133 is output. The shape error variation specified by the machining state specifying unit 419 may be superimposed on the measured value of the position of the tool 131 in the X-axis direction acquired by the measured value acquiring unit 411. In this case, it is possible to output a feedback signal related to the drive of the X-axis actuator 133 in consideration of the fluctuation of the deflection amount of the work W. The feedback unit 420 according to the first embodiment generates a feedback signal by sliding mode control. The sliding mode control is a control method for switching the feedback signal based on the hyperplane defined by the state quantity of the industrial machine 100 or the work W. The value of the feedback signal related to the sliding mode control is non-linear with respect to the control deviation and the shape error fluctuation. By performing feedback control by sliding mode control, higher followability can be obtained as compared with feedback control by proportional control or the like.
　学習部４２１は、目標状態量算出部４１３が算出した工具１３１の変位に係る状態量の目標値および反作用算出部４１５が算出した反作用に係る状態量の値、並びにフィードバック部４２０によるフィードバック信号を、学習用データセットとして、逆動特性モデルＭ３を訓練する。具体的には、逆動特性モデルＭ３の入力層Ｍ３１には、目標状態量算出部４１３が算出した状態量の目標値および反作用算出部４１５が算出した反作用に係る状態量の値が入力される。出力層Ｍ３３には、フィードバック部４２０が生成したフィードバック信号の値が入力される。学習によって得られた逆動特性モデルＭ３の学習済みパラメータは、ストレージ１４５に記憶される。学習済みパラメータは、例えば、逆動特性モデルＭ３の層数、各層におけるニューロンの個数、ニューロン同士の結合関係、各ニューロン間の結合の重み、及び各ニューロンの閾値を含む。 The learning unit 421 receives the target value of the state amount related to the displacement of the tool 131 calculated by the target state amount calculation unit 413, the value of the state amount related to the reaction calculated by the reaction calculation unit 415, and the feedback signal by the feedback unit 420. The feedback characteristic model M3 is trained as a training data set. Specifically, the target value of the state amount calculated by the target state amount calculation unit 413 and the value of the state amount related to the reaction calculated by the reaction calculation unit 415 are input to the input layer M31 of the reverse motion characteristic model M3. .. The value of the feedback signal generated by the feedback unit 420 is input to the output layer M33. The trained parameters of the reverse motion characteristic model M3 obtained by learning are stored in the storage 145. The learned parameters include, for example, the number of layers of the reverse motion characteristic model M3, the number of neurons in each layer, the connection relationship between neurons, the weight of connection between each neuron, and the threshold value of each neuron.
《制御装置の動作》
　図５は、第１の実施形態に係る制御装置の動作を示すブロック線図である。
　図６は、第１の実施形態に係る制御装置の動作を示すフローチャートである。
　まず、産業機械１００による加工動作が開始されると、制御装置１４０の計測値取得部４１１は産業機械１００に設けられた複数のセンサから、回転モータ１２３のトルクおよび回転角、回転モータ１３４のトルクおよび回転数、Ｘ軸アクチュエータ１３３の推力、工具１３１のＸ軸方向における位置の計測値を取得する（ステップＳ１）。 << Operation of control device >>
FIG. 5 is a block diagram showing the operation of the control device according to the first embodiment.
FIG. 6 is a flowchart showing the operation of the control device according to the first embodiment.
First, when the machining operation by the industrial machine 100 is started, the measured value acquisition unit 411 of the control device 140 receives the torque and rotation angle of the rotary motor 123 and the torque of the rotary motor 134 from a plurality of sensors provided in the industrial machine 100. And the number of rotations, the thrust of the X-axis actuator 133, and the measured values of the position of the tool 131 in the X-axis direction are acquired (step S1).
　次に、目標位置算出部４１２は、ワークモデルＭ２に基づいて、工具１３１の目標位置Ｒを算出する（ステップＳ２）。目標状態量算出部４１３は、ステップＳ２で算出した目標位置Ｒに基づいて工具１３１の変位ＸＬに係る状態量（速度、加速度、ジャーク）の目標値を算出する（ステップＳ３）。 Next, the target position calculation unit 412 calculates the target position R of the tool 131 based on the work model M2 (step S2). Target state quantity calculation unit 413 calculates a target value of the state quantity relating to the displacement X L of the tool 131 (velocity, acceleration, jerk) based on the target position R calculated in step S2 (step S3).
　指令値算出部４１４は、工具１３１の状態量の目標値を、当該目標値を達成するための電流値（トルク値）に変換する（ステップＳ４）。指令値算出部４１４は、ステップＳ４で変換した電流値にフィードバック部４２０から出力されたフィードバック信号ＵＦＢの値を加算することで電流指令値Ｖを算出する（ステップＳ５）。 The command value calculation unit 414 converts the target value of the state quantity of the tool 131 into a current value (torque value) for achieving the target value (step S4). The command value calculation unit 414 calculates the current command value V by adding the value of the feedback signal UFB output from the feedback unit 420 to the current value converted in step S4 (step S5).
　反作用算出部４１５は、ステップＳ３で算出した状態量の目標値と基台の動特性モデルＭ１とに基づいて、Ｘ軸アクチュエータ１３３の駆動の反作用によって生じる基台１１０の変位、速度、加速度、およびジャークのラプラス変換に係る値を算出する（ステップＳ６）。補償値算出部４１６は、ステップＳ３で算出した状態量の目標値およびステップＳ６で算出した反作用に係る状態量の値を、逆動特性モデルＭ３に入力することで、電流補償値ＵＮＮを算出する（ステップＳ７）。指令出力部４１７は、ステップＳ５で算出した電流指令値ＶとステップＳ７で算出した電流補償値ＵＮＮとの和の値を示す電流指令Ｕ（推力指令）を、Ｘ軸アクチュエータ１３３に出力する（ステップＳ８）。 The reaction calculation unit 415 determines the displacement, velocity, acceleration, and displacement of the base 110 caused by the reaction of the drive of the X-axis actuator 133 based on the target value of the state quantity calculated in step S3 and the dynamic characteristic model M1 of the base. A value related to the Laplace transform of jerk is calculated (step S6). Compensation value calculating unit 416, the value of the state quantity relating to the reaction was calculated by the target value and the step S6 of the calculated state quantity at step S3, by entering the reversing characteristic model M3, calculates a current compensation value U NN (Step S7). The command output unit 417 outputs a current command U (thrust command) indicating the sum of the current command value V calculated in step S5 and the current compensation value UNN calculated in step S7 to the X-axis actuator 133 (thrust command). Step S8).
　状態量算出部４１８は、ステップＳ１で取得した計測値Ｙと、ワークモデルＭ２とに基づいて、ワークＷの撓み量変動データを算出する（ステップＳ９）。加工状態特定部４１９は、ステップＳ１で取得した計測値と、ステップＳ９で算出したワークＷの撓み量変動データとに基づいて、ワークＷの形状誤差変動を算出する（ステップＳ１０）。 The state quantity calculation unit 418 calculates the deflection amount fluctuation data of the work W based on the measured value Y acquired in step S1 and the work model M2 (step S9). The machining state specifying unit 419 calculates the shape error fluctuation of the work W based on the measured value acquired in step S1 and the deflection amount fluctuation data of the work W calculated in step S9 (step S10).
　フィードバック部４２０は、ステップＳ３で算出した状態量の目標値とステップＳ１で取得した工具１３１のＸ軸方向における位置の計測値Ｙに基づいて、Ｘ軸アクチュエータ１３３の駆動に係るフィードバック信号ＵＦＢを出力する（ステップＳ１１）。上記のステップＳ１で取得した工具１３１のＸ軸方向における位置の計測値には、ステップＳ１０で特定した形状誤差変動を重畳してもよい。 The feedback unit 420 transmits the feedback signal UFB related to the drive of the X-axis actuator 133 based on the target value of the state quantity calculated in step S3 and the measured value Y of the position of the tool 131 in the X-axis direction acquired in step S1. Output (step S11). The shape error variation specified in step S10 may be superimposed on the measured value of the position of the tool 131 in the X-axis direction acquired in step S1 above.
　学習部４２１は、ステップＳ３で算出した状態量の目標値、ステップＳ６で算出した反作用に係る状態量の値、およびステップＳ１１で出力されたフィードバック信号ＵＦＢの値の組み合わせを学習用データセットとして、逆動特性モデルＭ３を訓練し、学習済みパラメータを更新する（ステップＳ１２）。ステップＳ１２については、オンライン学習またはオフライン学習のどちらでもよい。 The learning unit 421 uses a combination of the target value of the state quantity calculated in step S3, the value of the state quantity related to the reaction calculated in step S6, and the value of the feedback signal UFB output in step S11 as a learning data set. , The feedback characteristic model M3 is trained and the learned parameters are updated (step S12). Regarding step S12, either online learning or offline learning may be performed.
　制御装置１４０は、ワークＷの加工動作が終了したか否かを判定する（ステップＳ１３）。加工動作が終了していない場合（ステップＳ１３：ＮＯ）、ステップＳ１に処理を戻し、加工制御を継続する。他方、加工動作が終了した場合（ステップＳ１３：ＹＥＳ）、制御装置１４０は加工制御を終了する。 The control device 140 determines whether or not the machining operation of the work W is completed (step S13). If the machining operation is not completed (step S13: NO), the process is returned to step S1 and the machining control is continued. On the other hand, when the machining operation is completed (step S13: YES), the control device 140 ends the machining control.
《性能比較》
　ここで、第１の実施形態に係る制御装置１４０による制御性能と、電流補償値による補正を行わない場合の制御装置１４０（比較例に係る制御装置１４０）による制御性能とを比較する。 《Performance comparison》
Here, the control performance by the control device 140 according to the first embodiment is compared with the control performance by the control device 140 (control device 140 according to the comparative example) when the correction by the current compensation value is not performed.
　図７Ａは、比較例に係る制御装置が出力するフィードバック信号の時系列を示すグラフである。図７Ｂは、比較例に係る制御装置による加工制御における輪郭誤差の時系列を示すグラフである。
　比較例に係る制御装置１４０は、逆動特性モデルＭ３が学習されていない状態における第１の実施形態に係る制御装置１４０である。すなわち、比較例に係る制御装置１４０においては、補償値算出部４１６が算出する電流補償値が常にゼロである。そのため、図７Ａに示すように、基台１１０の振動によって生じる工具１３１とワークＷの相対位置のずれ量がフィードバック部４２０によるフィードバック信号によって補償される。他方、フィードバック制御はループ時間だけ遅れが生じるため、図７Ｂに示すように、輪郭誤差が生じてしまう。 FIG. 7A is a graph showing a time series of feedback signals output by the control device according to the comparative example. FIG. 7B is a graph showing a time series of contour errors in machining control by the control device according to the comparative example.
The control device 140 according to the comparative example is the control device 140 according to the first embodiment in a state where the reverse motion characteristic model M3 has not been learned. That is, in the control device 140 according to the comparative example, the current compensation value calculated by the compensation value calculation unit 416 is always zero. Therefore, as shown in FIG. 7A, the amount of deviation between the relative positions of the tool 131 and the work W caused by the vibration of the base 110 is compensated by the feedback signal by the feedback unit 420. On the other hand, since the feedback control is delayed by the loop time, a contour error occurs as shown in FIG. 7B.
　図８Ａは、第１の実施形態に係る制御装置が出力する電流補償値およびフィードバック信号の時系列を示すグラフである。図８Ｂは、第１の実施形態に係る制御装置による加工制御における輪郭誤差の時系列を示すグラフである。
　図８Ａ、図８Ｂに示すグラフは、３回のワークＷの加工動作を行うことで逆動特性モデルＭ３が学習された状態における第１の実施形態に係る制御装置１４０の制御性能を示す。図８Ａに示すように、第１の実施形態に係る制御装置１４０によれば、補償値算出部４１６が電流補償値を生成する。これにより、基台１１０の振動に合わせてＸ軸アクチュエータ１３３が駆動するため、工具１３１とワークＷの相対位置のずれ量が低減される。すなわち、電流補償値は、フィードフォワード信号として働く。これにより、フィードバック部４２０によるフィードバック信号の大きさは、比較例と比較して小さくなる。また、図８Ｂに示すように、補償値算出部４１６によるフィードフォワード制御によって工具１３１とワークＷの相対位置のずれ量が低減されるため、比較例と比較して輪郭誤差を低減することができる。 FIG. 8A is a graph showing a time series of the current compensation value and the feedback signal output by the control device according to the first embodiment. FIG. 8B is a graph showing a time series of contour errors in machining control by the control device according to the first embodiment.
The graphs shown in FIGS. 8A and 8B show the control performance of the control device 140 according to the first embodiment in a state where the reverse motion characteristic model M3 is learned by performing the machining operation of the work W three times. As shown in FIG. 8A, according to the control device 140 according to the first embodiment, the compensation value calculation unit 416 generates the current compensation value. As a result, the X-axis actuator 133 is driven in accordance with the vibration of the base 110, so that the amount of deviation between the relative positions of the tool 131 and the work W is reduced. That is, the current compensation value acts as a feedforward signal. As a result, the magnitude of the feedback signal by the feedback unit 420 becomes smaller than that of the comparative example. Further, as shown in FIG. 8B, the feedforward control by the compensation value calculation unit 416 reduces the amount of deviation between the relative positions of the tool 131 and the work W, so that the contour error can be reduced as compared with the comparative example. ..
《作用・効果》
　このように、第１の実施形態によれば、制御装置１４０は、工具１３１とワークＷの相対位置に係る目標位置に基づいて、Ｘ軸アクチュエータ１３３の作用に対する反作用に係る値を算出し、当該反作用に係る値と目標位置とに基づいて、Ｘ軸アクチュエータ１３３に電流指令を出力する。これにより、制御装置１４０は、Ｘ軸アクチュエータ１３３の反作用に追従するように工具１３１とワークＷの相対位置を制御することができる。したがって、第１の実施形態に係る制御装置１４０は、加工の高速化または産業機械の小型化による輪郭制御性能等の制御性能への影響を低減することができる。 《Action / Effect》
As described above, according to the first embodiment, the control device 140 calculates the value related to the reaction to the action of the X-axis actuator 133 based on the target position related to the relative position between the tool 131 and the work W, and the value is concerned. A current command is output to the X-axis actuator 133 based on the value related to the reaction and the target position. As a result, the control device 140 can control the relative positions of the tool 131 and the work W so as to follow the reaction of the X-axis actuator 133. Therefore, the control device 140 according to the first embodiment can reduce the influence on the control performance such as the contour control performance due to the speeding up of processing or the miniaturization of the industrial machine.
　また、第１の実施形態によれば、制御装置１４０は、反作用に係る値に基づいて電流補償値を算出し、指令値算出部４１４が算出した電流指令値と当該電流補償値の和に係る電流指令をＸ軸アクチュエータ１３３に出力する。つまり、第１の実施形態に係る制御方法によれば、反作用による影響を鑑みない既存の制御装置１４０に、反作用算出部４１５および補償値算出部４１６の機能を付加することで、加工の高速化または産業機械の小型化による輪郭制御性能への影響を低減する制御を実現させることができる。つまり、第１の実施形態に係る産業機械１００は、反作用算出部４１５および指令値算出部４１４を有する制御装置１４０を備えるが、他の実施形態に係る産業機械１００の制御装置１４０は、従前の制御装置と反作用算出部４１５、補償値算出部４１６、および指令出力部４１７を有する制御補正装置との組み合わせによって実現されてもよい。他方、他の実施形態においては、制御装置１４０は、逆動特性モデルＭ３に基づいて電流補償値を加味した電流指令値を直接算出するものであってもよい。 Further, according to the first embodiment, the control device 140 calculates the current compensation value based on the value related to the reaction, and relates to the sum of the current command value calculated by the command value calculation unit 414 and the current compensation value. The current command is output to the X-axis actuator 133. That is, according to the control method according to the first embodiment, the processing speed is increased by adding the functions of the reaction calculation unit 415 and the compensation value calculation unit 416 to the existing control device 140 that does not consider the influence of the reaction. Alternatively, it is possible to realize control that reduces the influence on the contour control performance due to the miniaturization of the industrial machine. That is, the industrial machine 100 according to the first embodiment includes a control device 140 having a reaction calculation unit 415 and a command value calculation unit 414, but the control device 140 of the industrial machine 100 according to another embodiment is a conventional control device 140. It may be realized by a combination of a control device and a control correction device having a reaction calculation unit 415, a compensation value calculation unit 416, and a command output unit 417. On the other hand, in another embodiment, the control device 140 may directly calculate the current command value including the current compensation value based on the reverse motion characteristic model M3.
　また、第１の実施形態によれば、制御装置１４０は、フィードバック信号を用いて逆動特性モデルＭ３を訓練する。これにより、手作業で教師データを生成することなく、自動的に学習処理を行うことができる。また第１の実施形態に係るフィードバック信号は、スライディングモード制御によって生成される。スライディングモード制御によって生成されるフィードバック信号は、非線形かつ高応答であるため、制御装置１４０を高速にかつ適切に学習させることができる。なお、他の実施形態においては、逆動特性モデルＭ３がＰＩＤ制御などの線形のフィードバック信号に基づいて訓練されてもよい。また他の実施形態においては、作業者が手動で生成した教師データを用いて逆動特性モデルＭ３を訓練してもよい。 Further, according to the first embodiment, the control device 140 trains the reverse motion characteristic model M3 using the feedback signal. As a result, the learning process can be automatically performed without manually generating teacher data. Further, the feedback signal according to the first embodiment is generated by the sliding mode control. Since the feedback signal generated by the sliding mode control is non-linear and highly responsive, the control device 140 can be learned at high speed and appropriately. In other embodiments, the reverse motion characteristic model M3 may be trained based on a linear feedback signal such as PID control. In another embodiment, the reverse motion characteristic model M3 may be trained using the teacher data manually generated by the operator.
〈他の実施形態〉
　第１の実施形態に係る産業機械１００は、研削盤であるが、これに限られない。例えば、他の実施形態に係る産業機械１００は、プレス機、フライス盤、旋盤などの他の工作機械であってもよい。 <Other Embodiments>
The industrial machine 100 according to the first embodiment is a grinding machine, but is not limited to this. For example, the industrial machine 100 according to another embodiment may be another machine tool such as a press machine, a milling machine, or a lathe.
　また、第１の実施形態に係る制御装置１４０は、ワークＷの加工動作と逆動特性モデルＭ３の学習とを同時に行うが、これに限られない。例えば、他の実施形態においては、予め逆動特性モデルＭ３の学習を実施しておき、ある程度の性能が見込めるようになった以降、学習部４２１による学習を行わなくてもよい。この場合、学習用の産業機械１００にて逆動特性モデルＭ３の学習を行っておき、学習済みの逆動特性モデルＭ３を他の産業機械１００に移転させてもよい。 Further, the control device 140 according to the first embodiment simultaneously performs the machining operation of the work W and the learning of the reverse motion characteristic model M3, but is not limited to this. For example, in another embodiment, learning of the reverse motion characteristic model M3 may be performed in advance, and learning by the learning unit 421 may not be performed after a certain level of performance can be expected. In this case, the reverse motion characteristic model M3 may be learned by the learning industrial machine 100, and the learned reverse motion characteristic model M3 may be transferred to another industrial machine 100.
　また、第１の実施形態に係る制御装置１４０は、駆動の反作用による工具とワークとの相対位置のずれを対象として逆動特性モデルＭ３の学習を行うが、これに限られない。例えば、他の実施形態に係る制御装置１４０は、反作用に係る値に加え、駆動の作用による基台以外の振動、非線形摩擦、象限突起などを対象として逆動特性モデルＭ３の学習をおこなってもよい。 Further, the control device 140 according to the first embodiment learns the reverse motion characteristic model M3 for the deviation of the relative position between the tool and the work due to the reaction of the drive, but is not limited to this. For example, in the control device 140 according to another embodiment, in addition to the value related to the reaction, the reverse motion characteristic model M3 may be learned for vibration, non-linear friction, quadrant protrusion, etc. other than the base due to the action of driving. Good.
　また、第１の実施形態に係る制御装置１４０は、ワークＷの撓み量変動データを算出するが、これに限られない。例えば、他の実施形態に係る制御装置は、ワークＷの撓み量変動データに代えて、または加えて、ワークＷのびびり量変動データを算出してもよい。 Further, the control device 140 according to the first embodiment calculates the deflection amount fluctuation data of the work W, but the present invention is not limited to this. For example, the control device according to another embodiment may calculate the chatter amount fluctuation data of the work W in place of or in addition to the deflection amount fluctuation data of the work W.
　上記産業機械によれば、加工の高速化または産業機械の小型化による制御性能への影響を低減することができる。 According to the above-mentioned industrial machine, it is possible to reduce the influence on the control performance due to the speeding up of processing or the miniaturization of the industrial machine.
　１００…産業機械　１１０…基台　１１１…Ｙ軸ガイド部　１１２…Ｙ軸アクチュエータ　１２０…支持装置　１２１…主軸台　１２２…芯押し台　１２３…回転モータ　１３０…工具台　１３１…工具　１３２…Ｘ軸ガイド部　１３３…Ｘ軸アクチュエータ　１３４…回転モータ　１４０…制御装置　１４１…プロセッサ　１４３…メインメモリ　１４５…ストレージ　１４７…インタフェース　４１１…計測値取得部　４１２…目標位置算出部　４１３…目標状態量算出部　４１４…指令値算出部　４１５…反作用算出部　４１６…補償値算出部　４１７…指令出力部　４１８…状態量算出部　４１９…加工状態特定部　４２０…フィードバック部　４２１…学習部　Ｍ１…動特性モデル　Ｍ２…ワークモデル　Ｍ３…逆動特性モデル　Ｍ３１…入力層　Ｍ３２…中間層　Ｍ３３…出力層　Ｗ…ワーク 100 ... Industrial machinery 110 ... Base 111 ... Y-axis guide part 112 ... Y-axis actuator 120 ... Support device 121 ... Headstock 122 ... tailstock 123 ... Rotating motor 130 ... Tool stand 131 ... Tool 132 ... X-axis guide part 133 ... X-axis actuator 134 ... Rotating motor 140 ... Control device 141 ... Processor 143 ... Main memory 145 ... Storage 147 ... Interface 411 ... Measurement value acquisition unit 412 ... Target position calculation unit 413 ... Target state amount calculation unit 414 ... Command value calculation unit 415 ... Reaction calculation unit 416 ... Compensation value calculation unit 417 ... Command output unit 418 ... State amount calculation unit 419 ... Processing state specification unit 420 ... Feedback unit 421 ... Learning unit M1 ... Dynamic characteristic model M2 ... Work model M3 ... Reverse motion characteristic Model M31 ... Input layer M32 ... Intermediate layer M33 ... Output layer W ... Work

Claims (8)

  1. 　ワークを加工するための工具と、
    　前記工具と前記ワークを相対的に移動させるアクチュエータと、
    　前記アクチュエータの挙動を制御する制御装置とを備える産業機械であって、
    　前記制御装置は、
    　前記工具と前記ワークの相対位置に係る目標位置に基づいて、前記工具の変位に係る状態量の目標値を算出する目標状態量算出部と、
    　前記状態量の目標値に基づいて、前記アクチュエータの作用に対する反作用に係る値を算出する反作用算出部と、
    　前記反作用に係る値と前記状態量の目標値とに基づいて、前記アクチュエータに電流指令を出力する指令出力部と
    　を備える産業機械。     Tools for machining workpieces and
    An actuator that relatively moves the tool and the work,
    An industrial machine including a control device for controlling the behavior of the actuator.
    The control device
    A target state quantity calculation unit that calculates a target value of a state quantity related to displacement of the tool based on a target position related to a relative position between the tool and the work.
    A reaction calculation unit that calculates a value related to the reaction to the action of the actuator based on the target value of the state quantity.
    An industrial machine including a command output unit that outputs a current command to the actuator based on a value related to the reaction and a target value of the state quantity.
  2. 　前記制御装置は、
    　前記状態量の目標値に基づいて、前記アクチュエータの電流指令値を算出する指令値算出部と、
    　前記反作用に係る値に基づいて、前記反作用によって生じる前記工具と前記ワークの相対位置のずれ量だけ前記アクチュエータを駆動させるための電流補償値を算出する補償値算出部と、
    　を備え、
    　前記指令出力部は、前記電流指令値と前記電流補償値の和に係る前記電流指令を前記アクチュエータに出力する
    　請求項１に記載の産業機械。     The control device
    A command value calculation unit that calculates the current command value of the actuator based on the target value of the state quantity, and
    A compensation value calculation unit that calculates a current compensation value for driving the actuator by the amount of deviation between the relative positions of the tool and the work caused by the reaction based on the value related to the reaction.
    With
    The industrial machine according to claim 1, wherein the command output unit outputs the current command related to the sum of the current command value and the current compensation value to the actuator.
  3. 　前記電流指令による前記アクチュエータの駆動後の前記工具と前記ワークの相対位置と、前記状態量の目標値とに基づくフィードバック信号を生成するフィードバック部を備え、
    　前記指令値算出部は、前記状態量の目標値と前記フィードバック信号とに基づいて前記電流指令値を算出し、
    　前記補償値算出部は、前記反作用に係る値と前記フィードバック信号とを含む学習用データセットを用いて、前記反作用に係る値が入力されると前記フィードバック信号に相当する電流補償値を出力するように訓練された学習済みモデルに基づいて、前記電流補償値を算出する
    　請求項２に記載の産業機械。     A feedback unit that generates a feedback signal based on the relative position of the tool and the work after the actuator is driven by the current command and the target value of the state quantity is provided.
    The command value calculation unit calculates the current command value based on the target value of the state quantity and the feedback signal.
    The compensation value calculation unit uses a learning data set including the value related to the reaction and the feedback signal, and outputs a current compensation value corresponding to the feedback signal when the value related to the reaction is input. The industrial machine according to claim 2, wherein the current compensation value is calculated based on the trained model trained in.
  4. 　前記フィードバック部は、スライディングモード制御によって前記フィードバック信号を生成する
    　請求項３に記載の産業機械。     The industrial machine according to claim 3, wherein the feedback unit generates the feedback signal by sliding mode control.
  5. 　前記ワークおよび前記工具を支持する基台を備え、
    　前記反作用に係る値は、前記アクチュエータの作用によって生じる前記基台の振動に係る値である
    　請求項１から請求項４の何れか１項に記載の産業機械。     With a base for supporting the work and the tool,
    The industrial machine according to any one of claims 1 to 4, wherein the value related to the reaction is a value related to the vibration of the base caused by the action of the actuator.
  6. 　ワークを加工するための工具と前記ワークを相対的に移動させるアクチュエータの挙動を制御する制御装置であって、
    　前記工具と前記ワークの相対位置に係る目標位置に基づいて、前記工具の変位に係る状態量の目標値を算出する目標状態量算出部と、
    　前記状態量の目標値に基づいて、前記アクチュエータの作用に対する反作用に係る値を算出する反作用算出部と、
    　前記反作用に係る値と前記状態量の目標値とに基づいて、前記アクチュエータに電流指令を出力する指令出力部と
    　を備える制御装置。     A control device that controls the behavior of a tool for machining a work and an actuator that moves the work relative to each other.
    A target state quantity calculation unit that calculates a target value of a state quantity related to displacement of the tool based on a target position related to a relative position between the tool and the work.
    A reaction calculation unit that calculates a value related to the reaction to the action of the actuator based on the target value of the state quantity.
    A control device including a command output unit that outputs a current command to the actuator based on a value related to the reaction and a target value of the state quantity.
  7. 　ワークを加工するための工具と前記ワークの相対位置に係る状態量の目標値に基づいて前記工具と前記ワークを相対的に移動させるアクチュエータの電流指令を出力する制御装置の制御補正装置であって、
    　前記工具と前記ワークの相対位置に係る状態量の目標値に基づいて、前記アクチュエータの作用に対する反作用に係る値を算出する反作用算出部と、
    　前記反作用に係る値に基づいて、前記反作用によって生じる前記工具と前記ワークの相対位置のずれ量だけ前記アクチュエータを駆動させるための電流補償値を算出する補償値算出部と、
    　前記制御装置が生成した電流指令が示す電流指令値に、前記電流補償値を加算した補正電流指令を前記アクチュエータに出力する指令出力部とを備える
    　制御補正装置。     It is a control correction device of a control device that outputs a current command of an actuator that relatively moves the tool and the work based on a target value of a state quantity related to a relative position between the tool for machining the work and the work. ,
    A reaction calculation unit that calculates a value related to the reaction to the action of the actuator based on the target value of the state quantity related to the relative position of the tool and the work.
    A compensation value calculation unit that calculates a current compensation value for driving the actuator by the amount of deviation between the relative positions of the tool and the work caused by the reaction based on the value related to the reaction.
    A control correction device including a command output unit that outputs a correction current command to the actuator by adding the current compensation value to the current command value indicated by the current command generated by the control device.
  8. 　ワークを加工するための工具と前記ワークを相対的に移動させるアクチュエータの制御方法であって、
    　前記工具と前記ワークの相対位置に係る目標位置に基づいて、前記工具の変位に係る状態量の目標値を算出するステップと、
    　前記状態量の目標値に基づいて、前記アクチュエータの作用に対する反作用に係る値を算出するステップと、
    　前記反作用に係る値と前記状態量の目標値とに基づいて、前記アクチュエータに電流指令を出力するステップと
    　を備える制御方法。     It is a control method of a tool for machining a work and an actuator for moving the work relatively.
    A step of calculating a target value of a state quantity related to displacement of the tool based on a target position related to a relative position between the tool and the work, and
    A step of calculating a value related to a reaction to the action of the actuator based on the target value of the state quantity, and
    A control method including a step of outputting a current command to the actuator based on a value related to the reaction and a target value of the state quantity.
PCT/JP2020/014300 2019-03-29 2020-03-27 Industrial machine, control device, control compensation device, and control method WO2020203876A1 (en)

Priority Applications (4)

Application Number Priority Date Filing Date Title
KR1020217025311A KR102598075B1 (en) 2019-03-29 2020-03-27 Industrial machines, control devices, control compensation devices, and control methods
CN202080017865.2A CN113518958A (en) 2019-03-29 2020-03-27 Industrial machine, control device, control correction device, and control method
JP2021512062A JPWO2020203876A1 (en) 2019-03-29 2020-03-27
DE112020000402.3T DE112020000402T5 (en) 2019-03-29 2020-03-27 Industrial machine, control device, control compensator and control method

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2019-068965 2019-03-29
JP2019068965 2019-03-29

Publications (1)

Publication Number Publication Date
WO2020203876A1 true WO2020203876A1 (en) 2020-10-08

Family

ID=72668370

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2020/014300 WO2020203876A1 (en) 2019-03-29 2020-03-27 Industrial machine, control device, control compensation device, and control method

Country Status (5)

Country Link
JP (1) JPWO2020203876A1 (en)
KR (1) KR102598075B1 (en)
CN (1) CN113518958A (en)
DE (1) DE112020000402T5 (en)
WO (1) WO2020203876A1 (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0628006A (en) * 1992-04-06 1994-02-04 Mitsubishi Electric Corp Two-freedom control device and servo control device for motor
JPH06320451A (en) * 1993-04-26 1994-11-22 Sanyo Electric Co Ltd Manipulator control method
JP2008310603A (en) * 2007-06-14 2008-12-25 Nakamura Sangyo Gakuen Motor position control device
JP2010108399A (en) * 2008-10-31 2010-05-13 Okuma Corp Position control apparatus for numerically controlled machine
JP2013123757A (en) * 2011-12-13 2013-06-24 Mori Seiki Co Ltd Machine tool
CN106338970A (en) * 2016-11-17 2017-01-18 沈阳工业大学 Servo system control method for five-shaft linked numerically controlled machine tool

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR100458920B1 (en) * 2002-03-19 2004-12-03 고종선 precision speed and position control method of actuator system
JP4598617B2 (en) * 2005-07-14 2010-12-15 オークマ株式会社 Position control device for numerical control machine
JP4944806B2 (en) * 2007-10-09 2012-06-06 オークマ株式会社 Position control device
JP4442694B2 (en) * 2008-02-29 2010-03-31 ソニー株式会社 Component mounting apparatus, vibration control apparatus, and vibration control method
JP5656193B2 (en) * 2011-04-20 2015-01-21 株式会社ハーモニック・ドライブ・システムズ Actuator positioning device with wave gear device
JP6741525B2 (en) * 2015-08-31 2020-08-19 キヤノン株式会社 Driving device, positioning device, lithographic apparatus, and article manufacturing method
JP6490131B2 (en) * 2017-03-31 2019-03-27 ファナック株式会社 Machine learning device, servo control device, servo control system, and machine learning method
JP2019068965A (en) 2017-10-06 2019-05-09 株式会社大都技研 Game machine

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH0628006A (en) * 1992-04-06 1994-02-04 Mitsubishi Electric Corp Two-freedom control device and servo control device for motor
JPH06320451A (en) * 1993-04-26 1994-11-22 Sanyo Electric Co Ltd Manipulator control method
JP2008310603A (en) * 2007-06-14 2008-12-25 Nakamura Sangyo Gakuen Motor position control device
JP2010108399A (en) * 2008-10-31 2010-05-13 Okuma Corp Position control apparatus for numerically controlled machine
JP2013123757A (en) * 2011-12-13 2013-06-24 Mori Seiki Co Ltd Machine tool
CN106338970A (en) * 2016-11-17 2017-01-18 沈阳工业大学 Servo system control method for five-shaft linked numerically controlled machine tool

Also Published As

Publication number Publication date
KR102598075B1 (en) 2023-11-02
KR20210113331A (en) 2021-09-15
CN113518958A (en) 2021-10-19
JPWO2020203876A1 (en) 2020-10-08
DE112020000402T5 (en) 2021-10-28

Similar Documents

Publication Publication Date Title
JP5566469B2 (en) Numerical control method
Chalhoub et al. Control of a flexible robot arm: experimental and theoretical results
CN101727088B (en) Position control apparatus for numerically controlled machine
JP5935958B2 (en) Trajectory control device
JP3313643B2 (en) Servo control method in orbit machining with bite tool and servo controller for orbit machining
WO2014136686A1 (en) Working machine feed shaft control method and feed shaft control device
Wells et al. Feedforward and feedback control of a flexible robotic arm
JP5943650B2 (en) Servo control device and servo control method
JP4879091B2 (en) Method for controlling numerically controlled machine tool and numerically controlled machine tool
CN103853099A (en) SERVO CONTROL device WITH POSITION COMPENSATION FUNCTION FOR DRIVEN MEMBER
CN112601640A (en) Robot adjustment
Escorcia-Hernández et al. A new solution for machining with RA-PKMs: Modelling, control and experiments
JP2007072943A (en) Position controller
WO2020203876A1 (en) Industrial machine, control device, control compensation device, and control method
Saied et al. A novel model-based robust super-twisting sliding mode control of pkms: Design and real-time experiments
KR20240000643A (en) Manipulator for finishng operation and operating method of the same
JP6068779B2 (en) Control devices for linear and rotary robots
JPS5981705A (en) Compensating method of nc machine tool
JP4576530B2 (en) Servo gain calculation method, servo gain calculation program, and servo gain calculation device
JPH1158285A (en) Force control system of hand mechanism
JP5334932B2 (en) Parameter setting method and parameter setting device
US11338406B2 (en) Estimating device and estimating system
JPH1063339A (en) Controller of numerical value control machine tool
JP2021002194A (en) Numerical control apparatus, cnc machine tool, numerical control method, and numerical control program
CN114281018A (en) Offline prediction method for reverse jump error peak value of numerical control machine tool inclined feeding system

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20782587

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2021512062

Country of ref document: JP

Kind code of ref document: A

ENP Entry into the national phase

Ref document number: 20217025311

Country of ref document: KR

Kind code of ref document: A

122 Ep: pct application non-entry in european phase

Ref document number: 20782587

Country of ref document: EP

Kind code of ref document: A1