Control device

The control device uses machine learning to estimate tension values, eliminating the need for tension sensors, thereby reducing costs and sensor-related failures in industrial machines.

JP7587019B2Active Publication Date: 2024-11-19FANUC LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
JP2023509136
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2021-03-24
Filing Date
2022-03-18
Publication Date
2024-11-19
Estimated Expiration
2042-03-18

AI Technical Summary

Technical Problem

Industrial machines require multiple sensors to detect the driving status, leading to increased costs and potential malfunctions due to sensor failures, necessitating a method to adjust tension without a tension sensor during operation.

Method used

A control device uses a machine learning device to estimate tension values by learning from data related to the machine configuration, workpiece, and operating conditions, eliminating the need for a tension sensor.

Benefits of technology

Reduces the cost and maintenance of tension sensors by enabling accurate tension adjustment based on machine learning, reducing sensor-related failures.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007587019000001
    Figure 0007587019000001
  • Figure 0007587019000002
    Figure 0007587019000002
  • Figure 0007587019000003
    Figure 0007587019000003
Patent Text Reader

Abstract

This control device acquires data pertaining to the machine configuration of a work machine, data pertaining to a workpiece, and data pertaining to an operating condition, and creates machine learning data to be used in machine learning processing on the basis of the acquired data. On the basis of the created machine learning data, a command is issued to execute machine learning processing for estimating data pertaining to tension in a conveyance unit of the work machine. On the basis of the command, machine learning processing for estimating data pertaining to tension in the conveyance unit is executed. Through this configuration, the control device can adjust the tension of the conveyance unit in accordance with a specified condition during actual operation, without a tension sensor.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a control device for controlling an industrial machine having a substantially belt-shaped conveying unit that drives rolls to convey a workpiece. [Background technology]

[0002] In manufacturing sites such as factories, a number of industrial machines such as machine tools, robots, and conveying machines are installed. The conveying machines serve to convey workpieces on the factory line. Figure 6 is a diagram illustrating a schematic configuration of a conveyor that drives rolls to convey the workpieces.

[0003] The conveyor 300 includes a plurality of rolls 301 and a conveyor belt 302 stretched around the rolls 301. The conveyor belt 302 is conveyed in the conveying direction by, for example, a part of the rolls 301 being rotated by a motor (not shown). A tension sensor 303 detects the tension generated in the conveyor belt 302 during conveyance. While the conveyor 300 is driven, the tension of the conveyor belt 302 is adjusted by a tension adjustment mechanism 304 such as a tension roller so that slippage does not occur between the rolls 301 and the conveyor belt 302 and so that excessive tension does not occur in the conveyor belt 302. In the conveyor 300 illustrated in FIG. 6, a workpiece 305 is placed on the conveyor belt 302 and conveyed.

[0004] The target tension T aim Once set, the target tension T aim and the actual tension T detected by the tension sensor 303 FBK The tension deviation ΔT, which is the difference between the target tension T aim Based on the tension deviation ΔT, the tension-torque conversion circuit 310 outputs a target torque Q aim On the other hand, the mechanical loss torque Q is used to compensate for mechanical loss caused by aging (wear, etc.) of the machine. L and the acceleration / deceleration torque Q to compensate for the motor output voltage required for acceleration / deceleration. Fare calculated by the mechanical loss torque compensation circuit 320 and the acceleration / deceleration torque compensation circuit 330, respectively.

[0005] In this way, in a transport device such as a conveyor that transports a workpiece, a tension sensor or the like is used to detect the transport state, and the tension of the transport belt or the like is adjusted according to the detected state (for example, Patent Document 1, etc.). [Prior art documents] [Patent documents]

[0006] [Patent Document 1] International Publication No. 2014 / 103886 Summary of the Invention [Problem to be solved by the invention]

[0007] Industrial machinery uses many sensors to detect the operating status of each part. In this case, there is a demand from the field to reduce the overall cost of the industrial machinery by reducing the number of sensors used. In addition, reducing the number of sensors also reduces malfunctions of the industrial machinery caused by sensor failure. Therefore, there is a demand for a technology that can adjust the tension according to specified conditions without a tension sensor during actual operation. [Means for solving the problem]

[0008] The control device according to the present invention solves the above problem by using a machine learning machine instead of a tension sensor to calculate the torque of each axis according to specified conditions. In this specification, a substantially belt-shaped member used to transport workpieces in a transport device such as a belt conveyor is called a transport section. When the machine learning machine is learning, data related to the mechanical configuration of the transport machine, data related to the transported workpieces, data related to the operating conditions of the machine, and the like are used as data indicating the operating state of the machine, and the correlation between these data and the tension value of the transport section detected by the tension sensor is learned. Then, during actual operation, the machine learning machine estimates the tension value of the transport section instead of the tension sensor, and the tension of the transport belt in the transport machine is adjusted based on the estimated result.

[0009] One aspect of the present invention is a control device that controls an industrial machine having a substantially belt-like conveying section that drives a roll to convey a workpiece, and the control device includes: a data acquisition unit that acquires data related to a mechanical configuration relating to the operating state of the industrial machine, data related to a workpiece, and data related to operating conditions; an acquired data storage unit that stores the data related to the mechanical configuration, data related to the workpiece, and data related to operating conditions relating to the operating state of the industrial machine acquired by the data acquisition unit; a machine learning data creation unit that creates machine learning data to be used for machine learning processing based on the data stored in the acquired data storage unit; a machine learning processing command unit that commands the execution of machine learning processing for estimating data related to tension in the conveying section based on the data created by the machine learning data creation unit; and a machine learning unit that executes machine learning processing for estimating data related to tension in the conveying section based on a command from the machine learning processing command unit. Effect of the Invention

[0010] One aspect of the present invention makes it possible to output a more appropriate inter-axis tension without a tension sensor, thereby reducing the cost (including maintenance cost) of the tension sensor that was necessary in conventional tension control. [Brief description of the drawings]

[0011] [Figure 1] FIG. 2 is a schematic hardware configuration diagram of a control device according to an embodiment. [Diagram 2] FIG. 2 is a schematic block diagram showing the functions of a control device according to the first embodiment. [Diagram 3] FIG. 11 is a schematic block diagram showing the functions of a control device according to a second embodiment. [Figure 4] FIG. 11 is a schematic block diagram showing the functions of a control device according to a third embodiment. [Diagram 5] FIG. 1 is a diagram showing an example of an industrial machine that transports a workpiece on a thin film. [Figure 6] FIG. 1 shows an example of a conveying machine according to the prior art. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. 1 is a schematic hardware configuration diagram showing the main parts of a control device according to one embodiment of the present invention. The control device 1 of the present invention has a function of controlling an industrial machine 3, such as a conveying machine. In this embodiment, the configuration of the control device 1 and the industrial machine 3 at the stage of performing machine learning in the machine learning machine 2 is shown.

[0013] The CPU 11 provided in the control device 1 according to this embodiment is a processor that controls the entire control device 1. The CPU 11 reads out a system program stored in the ROM 12 via the bus 22, and controls the entire control device 1 in accordance with the system program. The RAM 13 temporarily stores temporary calculation data, display data, various data input from outside, and the like.

[0014] The non-volatile memory 14 is composed of, for example, a memory backed up by a battery (not shown) or an SSD (Solid State Drive), and the memory state is maintained even when the power supply of the control device 1 is turned off. The non-volatile memory 14 stores control programs and data read from the external device 72 via the interface 15, control programs and data input from the input device 71 via the interface 18, and control programs and data acquired from other devices such as the fog computer 6 and the cloud server 7 via the network 5. The data stored in the non-volatile memory 14 includes, for example, data relating to the mechanical configuration of the industrial machine 3 and data relating to the workpieces to be transported. 、 This may include data related to the operating conditions of the machine, tension value data of a conveyor belt or the like detected by the tension sensor 4, and data related to various physical quantities detected by sensors (not shown) attached to other industrial machines 3. The control programs and data stored in the non-volatile memory 14 may be expanded in the RAM 13 when executed / used. In addition, various system programs such as known analysis programs are written in the ROM 12 in advance.

[0015] One or more tension sensors 4 attached to the industrial machine 3 detect the tension of the conveyor belt of the industrial machine 3. The tension sensor 4 may be a force sensor that detects the reaction force when a force is applied to the conveyor belt, or may be a sensor that detects the tension of the conveyor belt in a non-contact manner using sound waves. The tension sensor 4 is necessary when the control device 1 operates in the learning mode, but is not necessary when the control device 1 operates in the actual operation mode.

[0016] The interface 15 is an interface for connecting the CPU 11 of the control device 1 to an external device 72 such as an external storage medium. For example, control programs and setting data used to control the industrial machine 3 are read from the external device 72. The control programs and setting data edited in the control device 1 can be stored in an external storage medium (not shown) such as a CF card or a USB memory via the external device 72. The programmable logic controller (PLC) 16 executes a ladder program and outputs signals to the industrial machine 3 and peripheral devices of the industrial machine 3 (for example, a tool changer, an actuator such as a robot, a tension sensor 4 attached to the industrial machine 3, a temperature sensor, a humidity sensor, and other sensors) via an I / O unit 19 to control them. The PLC 16 also receives signals from various switches and peripheral devices on an operation panel provided on the main body of the industrial machine 3, performs necessary signal processing on the signals, and then passes them to the CPU 11.

[0017] The interface 20 is an interface for connecting the CPU 11 of the control device 1 to a wired or wireless network 5. The network 5 may communicate using technologies such as serial communication such as RS-485, Ethernet (registered trademark), optical communication, wireless LAN, Wi-Fi (registered trademark), Bluetooth (registered trademark), etc. The network 5 is connected to control devices that control other industrial machines, and higher-level management devices such as a fog computer 6 and a cloud server 7, and exchanges data with the control device 1.

[0018] The display device 70 displays various data loaded into the memory, data obtained as a result of executing programs, etc., outputted via the interface 17. An input device 71 consisting of a keyboard, pointing device, etc., transfers commands, data, etc. based on operations by an operator to the CPU 11 via the interface 18.

[0019] An axis control circuit 30 for driving the drive units of the industrial machine 3 receives a movement command amount from the CPU 11 and outputs a movement command to each servo amplifier 40. The servo amplifier 40 receives this command and drives each servo motor 50 of the industrial machine 3. The servo motor 50 has a built-in position and speed detector, and feeds back position and speed feedback signals from this position and speed detector to the axis control circuit 30, respectively, to perform feedback control of the position and speed. Note that while the hardware configuration diagram of FIG. 1 shows only one each of the axis control circuit 30, servo amplifier 40, and servo motor 50, in reality, there are provided as many as the number of drive units of the industrial machine 3 to be controlled.

[0020] The interface 21 is an interface for connecting the CPU 11 and the machine learning machine 2. The machine learning machine 2 includes a processor 201 that controls the entire machine learning machine 2, a ROM 202 that stores a system program, etc., a RAM 203 that temporarily stores data in each process related to machine learning, and a non-volatile memory 204 that is used to store a learning model, etc. The machine learning machine 2 can observe data that can be acquired by the control device 1 through the interface 21 (for example, data related to the machine configuration, data related to the workpiece to be conveyed, data related to the operating conditions of the machine, tension value data of the conveyor belt, etc. detected by the tension sensor 4, etc.). In addition, the control device 1 acquires the processing results output from the machine learning machine 2 through the interface 21, and stores or displays the acquired results, or transmits them to other devices via the network 5, etc. Note that, although the machine learning machine 2 is built into the control device 1 in FIG. 1, it may be externally connected to the control device 1 through a predetermined interface.

[0021] 2 is a schematic block diagram showing functions of the control device 1 according to the first embodiment of the present invention. Each function of the control device 1 according to this embodiment is realized by the CPU 11 of the control device 1 and the processor 201 of the machine learning machine 2 shown in FIG. 1 executing a system program and controlling the operation of each part of the control device 1 and the machine learning machine 2.

[0022] The control device 1 of this embodiment includes a control unit 100, a data acquisition unit 110, a machine learning data creation unit 120, a machine learning processing command unit 130, and a tension adjustment unit 140. Furthermore, a machine learning unit 205 constituting the machine learning device 2 includes a learning unit 206 and an estimation unit 208. Furthermore, the RAM 13 to the nonvolatile memory 14 of the control device 1 include an acquired data storage unit 190 which is an area for storing data acquired from the industrial machine 3, and a model storage unit 209 is prepared in advance on the RAM 203 to the nonvolatile memory 204 of the machine learning device 2 as an area for storing the learning model created by the learning unit 206.

[0023] The control unit 100 controls the operation of the industrial machine 3 based on preset operating conditions and a control program. The control unit 100 has general functions required for controlling the operation of the industrial machine 3, for example, as illustrated in Fig. 6. The control unit 100 controls the operation of the industrial machine 3 in accordance with the feed speed, acceleration, target tension, and the like specified by the operating conditions, the control program, and the like, while referring to the measured speed, measured tension, and the like fed back from the industrial machine 3.

[0024] The data acquisition unit 110 acquires data I relating to the mechanical configuration of the industrial machine 3. s and data related to the work to be transported I w , Data related to machine operating conditions I c , tension value data O of a conveyor belt or the like detected by a tension sensor 4 attached to an industrial machine 3 t from the control unit 100, for example. The data acquisition unit 110 may also acquire data related to a warning issued in the industrial machine 3 and input value data input by an operator via the input device 71. The data acquisition unit 110 may also acquire data related to the operation of the industrial machine 3 acquired and stored by the external device 72, the fog computer 6, the cloud server 7, etc. The data acquired by the data acquisition unit 110 is at least data related to the operating state of the industrial machine 3 during operation.

[0025] Data I relating to the machine configuration acquired by the data acquisition unit 110 s For example, the number of axes of the industrial machine 3 (the number of rotors to be driven, etc.), the distance between each axis, whether it is a gravity axis or not, the direction of movement, the diameter of the roll, the type and material of the conveyor belt, etc. are included in the data I w This includes the workpiece material, thickness, weight, shape, etc. In addition, data related to operating conditions I c Includes feed speed, acceleration, total operation time, total operation distance, etc. Tension value data t are tension values ​​detected by the tension sensor 4 attached to the industrial machine 3 when data related to each operating state is acquired. t may include a plurality of tension values ​​measured by a plurality of tension sensors 4. Each of these data is stored in the acquired data storage unit 190 after being organized according to the timing of acquisition or detection.

[0026] The machine learning data creation unit 120 creates learning data to be used in machine learning processing related to tension value adjustment by the machine learning unit 205, based on the data stored in the acquired data storage unit 190. More specifically, when the control device 1 operates in the learning mode, the machine learning data creation unit 120 creates learning data to be used in learning processing by the machine learning unit 205, based on the data stored in the acquired data storage unit 190. The learning data created by the machine learning data creation unit 120 includes at least data I related to the machine configuration. s , Data related to the work I w and data related to operating conditions I c On the other hand, tension value data O of a conveyor belt or the like detected by a tension sensor 4 t The learning data created by the machine learning data creation unit 120 may further include additional data depending on the machine learning method used by the machine learning unit 205. On the other hand, when the control device 1 is operating in the actual operation mode, the machine learning data creation unit 120 creates estimation data to be used in the estimation process by the estimation unit 208 based on the data stored in the acquired data storage unit 190. The estimation data created by the machine learning data creation unit 120 includes at least data I relating to the machine configuration. s , Data related to the work I w and data related to operating conditions I c The estimation data created by the machine learning data creation unit 120 may further include additional data depending on the machine learning method used by the machine learning unit 205.

[0027] The machine learning process command unit 130 commands the machine learning unit 205 to perform machine learning processing related to the estimation of data related to tension, based on the data created by the machine learning data creation unit 120. More specifically, when the control device 1 is operating in the learning mode, the machine learning process command unit 130 commands the machine learning unit 205 to perform machine learning processing related to the estimation of data related to tension, based on the learning data created by the machine learning data creation unit 120, for example. s , Data related to the work I w and data related to operating conditions I c and tension value data O of a conveyor belt or the like detected by the tension sensor 4. t The machine learning unit 205 is instructed to learn the correlation between the adjustment behavior data and the tension value. On the other hand, when the control device 1 is operating in the actual operation mode, the machine learning processing command unit 130 generates, for example, data I relating to the machine configuration based on the estimation data created by the machine learning data creation unit 120. s , Data related to the work I w and data related to operating conditions I c The machine learning unit 205 is instructed to estimate the tension value and the tension value adjustment action using the above.

[0028] The tension adjustment unit 140 issues an action command to the control unit 100 to adjust the tension value based on data relating to the tension estimated by the machine learning unit 205 (such as the tension value itself or an adjustment action for the tension value). For example, when a tension value is estimated by the machine learning unit 205, the tension adjustment unit 140 issues an action command to the control unit 100 to adjust the estimated tension value to a target tension value. Also, for example, when an adjustment action for the tension value is estimated by the machine learning unit 205, the tension adjustment unit 140 issues an estimated adjustment action command to the control unit 100. Note that the tension adjustment unit 140 is not necessarily a necessary component when the control device 1 is operating in the learning mode.

[0029] The learning unit 206 included in the machine learning unit 205 creates a model that learns data related to tension of a conveyor belt or the like in relation to the operating state of the industrial machine 3 based on the learning data included in the command received from the machine learning process command unit 130, and stores the created model in the model storage unit 209 based on the learning data received from the machine learning process command unit 130. More specifically, when performing supervised learning, for example, the learning unit 206 stores data I related to the machine configuration that indicates the operating state of the industrial machine 3. s , Data related to the work I w and data related to operating conditions I c The tension value data O of the conveyor belt detected by the tension sensor 4 for the data such as t Alternatively, when performing reinforcement learning, the learning unit 206 creates a model that learns the correlation between the data I s , Data related to the work I w and data related to operating conditions I c A model is created that learns the correlation between data such as the above and adjustment actions for tension values ​​of conveyor belts, etc. detected by tension sensor 4. The machine learning performed by the learning unit 206 may be known supervised learning or reinforcement learning. The model created by the learning unit 206 is generated by performing machine learning on the data I related to the machine configuration. s , Data related to the work I w and data related to operating conditions I cFor example, the learning unit 206 can estimate data related to the tension of the conveyor belt, etc. (the tension value itself, or the adjustment action of the tension value). Examples of the model created by the learning unit 206 include a regression learner and a multi-layer neural network.

[0030] The estimation unit 208 included in the machine learning unit 205 executes estimation processing of data related to tension (the tension value itself, or the adjustment action of the tension value) using a model stored in the model storage unit 209, based on the estimation data received from the machine learning processing command unit 130. The estimation processing performed by the estimation unit 208 may be estimation processing by known supervised learning or reinforcement learning. Note that the estimation unit 208 is not necessarily a necessary component when the control device 1 is operating in the learning mode.

[0031] The control device 1 according to the present embodiment having the above configuration stores data I relating to the machine configuration in the learning mode. s , Data related to the work I w and data related to operating conditions I c For example, tension data for conveyor belts, etc. t The model created by the learning unit 206 can be used to estimate the tension value of a conveyor belt or the like in response to the operating state of the industrial machine 3, instead of the tension sensor 4, during actual operation of the industrial machine 3. In addition, the control device 1, in the actual operation mode, uses the data I relating to the machine configuration s , Data related to the work I w and data related to operating conditions I c Based on this, the tension value of a conveyor belt of the industrial machine 3 or the like is estimated. Therefore, the industrial machine 3 does not need a tension sensor 4, and the manufacturing and operating costs of the industrial machine 3 can be reduced.

[0032] 3 is a schematic block diagram showing functions of a control device 1 according to a second embodiment of the present invention. In this embodiment, the machine learning unit 205 performs supervised learning.

[0033] The machine learning data creation unit 120 according to this embodiment includes a state observation unit 122 and a label creation unit 124. The state observation unit 122 creates state data S indicating the operating state of the industrial machine 3 during operation based on the data stored in the acquired data storage unit 190. The state data S indicating the operating state of the industrial machine 3 during operation includes at least data I relating to the machine configuration. s , Data related to the work I w and data related to operating conditions I c Includes.

[0034] The label creation unit 124 creates label data L for the data created by the state observation unit 122 based on the data stored in the acquired data storage unit 190. The label data L includes tension value data O of the conveyor belt and the like during the operation of the industrial machine 3 when the state data S indicating the operating state of the above-mentioned industrial machine 3 is observed. t At least

[0035] When the control device 1 operates in the learning mode, the machine learning data creation unit 120 creates teacher data T to be used in machine learning processing, based on the state data S created by the state observation unit 122 and indicating the operating state of the industrial machine 3 during operation, and the label data L created by the label creation unit 124. The teacher data T is output to the machine learning process command unit 130. Meanwhile, when the control device 1 operates in the actual operation mode, the machine learning data creation unit 120 outputs the state data S created by the state observation unit 122 and indicating the operating state of the industrial machine 3 during operation to the machine learning process command unit 130 as estimation data.

[0036] Then, when the control device 1 operates in the learning mode, the machine learning process command unit 130 commands the machine learning unit 205 to perform supervised learning processing based on the teacher data T. On the other hand, when the control device 1 operates in the actual operation mode, the machine learning process command unit 130 commands the machine learning unit 205 to estimate a tension value based on the estimation data.

[0037] 4 is a schematic block diagram showing functions of a control device 1 according to a third embodiment of the present invention. In this embodiment, the machine learning unit 205 performs reinforcement learning.

[0038] The machine learning data creation unit 120 according to this embodiment includes a state observation unit 122 and a judgment data creation unit 126 . The state observation unit 122 creates state data S indicating the operating state of the industrial machine 3 during operation based on the data stored in the acquired data storage unit 190. The state data S indicating the operating state of the industrial machine 3 during operation includes at least data I relating to the machine configuration. s , Data related to the work I w and data related to operating conditions I c Includes.

[0039] The judgment data creation unit 126 creates judgment data D for the data created by the state observation unit 122, based on the data stored in the acquired data storage unit 190. The judgment data D is obtained by observing state data S indicating the operating state of the above-mentioned industrial machine 3 during operation, and obtaining tension value data O of the conveyor belt or the like during operation of the industrial machine 3 after adjustment when a predetermined tension value adjustment action a is taken. t and the target tension value.

[0040] When the control device 1 operates in the learning mode, the machine learning data creation unit 120 creates learning data to be used for machine learning processing, based on the state data S created by the state observation unit 122 and indicating the operating state of the industrial machine 3 during operation, and the adjustment action a of the predetermined tension value and the judgment data D created by the judgment data creation unit 126. The created learning data is output to the machine learning process command unit 130. Meanwhile, when the control device 1 operates in the actual operation mode, the machine learning data creation unit 120 outputs the state data S created by the state observation unit 122 and indicating the operating state of the industrial machine 3 during operation to the machine learning process command unit 130 as estimation data.

[0041] Then, when the control device 1 is operating in the learning mode, the machine learning processing command unit 130 instructs the machine learning unit 205 to perform reinforcement learning processing based on the learning data, while when the control device 1 is operating in the actual operation mode, it instructs the machine learning unit 205 to estimate adjustment actions for the tension value based on the estimation data.

[0042] The machine learning unit 205 according to this embodiment includes a learning unit 206, an estimation unit 208, and a reward calculation unit 207. The learning unit 206 executes reinforcement learning to learn whether a predetermined tension value adjustment action a is taken for the operating state of the industrial machine 3, based on the learning data received from the machine learning process command unit 130. Reinforcement learning learns whether the result of a predetermined action being taken for the current state of the learning subject is good or bad, in the form of a reward indicating the value of that action. By repeating the learning cycle of reinforcement learning by trial and error, a measure (adjustment action of tension value for the operating state of the industrial machine 3) that maximizes the total reward is learned as the optimal solution. Examples of reinforcement learning methods include Q-learning. In this case, the learning unit 206 uses, for example, a regression learner or a neural network as a value in reinforcement learning. Seki It may be created as a number Q (model).

[0043] The reward calculation unit 207 is realized by the processor 201 included in the machine learning device 2 shown in FIG. 1 executing a system program read from the ROM 202, and the processor 201 mainly performing arithmetic processing using the RAM 203 and non-volatile memory 204. The reward calculation unit 207 calculates a predetermined reward R indicating the value of a predetermined action for the judgment data D included in the learning data received from the machine learning processing command unit 130, and outputs the calculation result to the learning unit 206. In, for example, Q-learning by the learning unit 206, the reward R is calculated when the judgment data D is 0 (tension value data O of a conveyor belt or the like during operation of the industrial machine 3) t The closer the actual tension value is to 0 (i.e., the same as the target tension value), the more positive the reward R will be, and the further away from 0 it will be, the more negative the reward R will be.

[0044] The estimation unit 208 executes estimation of an action command for adjusting the tension value using a model stored in the model storage unit 209, based on the estimation data received from the machine learning processing command unit 130. The estimation unit 208 executes estimation of an action command for adjusting the tension value using a model generated by the reinforcement learning by the learning unit 206, in a state in which state data S relating to the operating state of the industrial machine 3 included in the estimation data is observed, to estimate a plurality of adjustment actions a for the tension value that can be currently taken. i A reward is calculated for each of (i is 1 to n), and the tension value adjustment action a for which the largest reward is calculated is estimated as the optimal solution. The tension value adjustment action command estimated by the estimation unit 208 is output to the tension adjustment unit 140.

[0045] Although one embodiment of the present invention has been described above, the present invention is not limited to the above-described embodiment, and can be embodied in various forms by making appropriate modifications. For example, in the above embodiment, the data related to the operating state of the industrial machine 3 includes data I related to the machine configuration. s , Data related to the work I w and data related to operating conditions I c In this example, the data I related to the environment in which the industrial machine 3 operates is used. e The data related to the environment I may be used as the data related to the operating state of the industrial machine 3. e Examples of environmental factors include environmental temperature and environmental humidity. Changes in environmental temperature affect the hardness, tenacity, and rigidity of the materials used in the transport section. Changes in environmental humidity also affect the friction between the transport section and other parts. Therefore, data related to these environments I e By using this, it is possible to adjust the tension with greater precision.

[0046] In the above embodiment, a conveying machine such as a belt conveyor is exemplified as the industrial machine 3, but the present invention can also be applied to a machine 400 that sends out a thin-film workpiece 402, as exemplified in FIG. 5. In the machine 400 as exemplified in FIG. 5, the workpiece 402 itself plays the role of a conveying unit. In general, a thin-film workpiece 402 such as a film is sent out from a work roll 401 through a plurality of rolls 403. Then, while detecting the tension of the workpiece 402 with a tension sensor 404, a tension adjustment mechanism 405 adjusts the tension so that no slip occurs between the workpiece 402 and the roll 403 and no excessive load is applied to the workpiece 402. By applying the control device 1 according to the present invention to such a machine, it becomes possible to perform tension adjustment without the need for a tension sensor 404. [Explanation of symbols]

[0047] 1 Control device 3. Industrial machinery 4 Tension Sensor 5. Network 6. Fog Computer 7. Cloud Server 11 CPU 12 ROM 13 RAM 14 Non-volatile memory 15,17,18,20,21 Interface 16 PLC 19 I / O Unit 22 Bus 30-axis control circuit 40 Servo amplifier 50 Servo motor 70 Display device 71 Input Devices 72 External equipment 100 Control section 110 Data Acquisition Section 120 Machine Learning Data Creation Department 130 Machine Learning Processing Command Center 140 Tension adjustment section 2 Machine learning 201 Processor 202 ROM 203 RAM 204 Non-volatile memory 205 Machine Learning Department 206 Learning Department 207 Remuneration Calculation Department 208 Estimation Department 209 Model Memory Section

Claims

1. A control device for controlling an industrial machine having a substantially belt-shaped conveying unit that drives a roll to convey a workpiece, a data acquisition unit that acquires data on a machine configuration related to an operating state of the industrial machine, data on a workpiece, and data on operating conditions; an acquired data storage unit that stores data on a machine configuration related to an operating state of the industrial machine acquired by the data acquisition unit, data on a workpiece, and data on operating conditions; a machine learning data creation unit that creates machine learning data to be used in machine learning processing based on the data stored in the acquired data storage unit; a machine learning processing command unit that commands execution of a machine learning process for estimating data related to the tension in the conveying unit based on the data created by the machine learning data creation unit; a machine learning unit that executes a machine learning process for estimating data related to tension in the conveying unit based on a command from the machine learning process command unit; A control device comprising:

2. the data acquisition unit acquires data relating to a machine configuration, data relating to a workpiece, and data relating to operating conditions as data relating to an operating state of the industrial machine, and also acquires tension value data in the conveying unit; The acquired data storage unit stores data on a machine configuration, data on a workpiece, and data on operating conditions of the industrial machine, which are acquired by the data acquisition unit; The machine learning data creation unit includes: a status observation unit that creates status data including data related to a machine configuration, data related to a workpiece, and data related to operating conditions based on the data stored in the acquired data storage unit; a label creating unit that creates label data including tension value data in the conveying unit based on the data stored in the acquired data storage unit, the machine learning process command unit commands the machine learning unit to execute a learning process for creating a model for estimating data related to tension in the conveying unit, as the machine learning process, based on the state data observed by the state observation unit and the label data created by the label creation unit; The machine learning unit is a learning unit that creates a supervised learning model that learns a correlation between data related to an operational state of the industrial machine and a tension value of the conveying unit based on the state data and the label data; A model storage unit that stores the model created by the learning unit. The control device according to claim 1 .

3. The data acquisition unit acquires data related to a machine configuration, data related to a workpiece, and data related to operating conditions as data related to an operating state of the industrial machine, The acquired data storage unit stores data on a machine configuration, data on a workpiece, and data on operating conditions of the industrial machine, which are acquired by the data acquisition unit; The machine learning data creation unit includes a state observation unit that creates state data including data related to a machine configuration, data related to a workpiece, and data related to operating conditions based on the data stored in the acquired data storage unit, the machine learning processing command unit commands the machine learning unit to execute a process of estimating data related to tension in the conveying unit as the machine learning processing, based on the state data observed by the state observation unit; The machine learning unit includes an estimation unit that estimates the tension value of the conveying unit using a supervised learning model that has learned a correlation between data related to the operating state of the industrial machine and the tension value of the conveying unit based on the state data, and The control device further includes a tension adjustment unit that adjusts the tension of the conveying section based on the tension value of the conveying section estimated by the estimation unit. The control device according to claim 1 .

4. the data acquisition unit acquires data relating to a machine configuration, data relating to a workpiece, and data relating to operating conditions as data relating to an operating state of the industrial machine, and also acquires tension value data in the conveying unit; The acquired data storage unit stores data on a machine configuration, data on a workpiece, and data on operating conditions of the industrial machine, which are acquired by the data acquisition unit; The machine learning data creation unit includes: a status observation unit that creates status data including data related to a machine configuration, data related to a workpiece, and data related to operating conditions based on the data stored in the acquired data storage unit; a judgment data creation unit that creates judgment data including a difference between tension value data in the conveying unit after adjustment and a target tension value of the conveying unit when a predetermined tension adjustment action is taken for the conveying unit when the state data is observed, based on the data stored in the acquired data storage unit, the machine learning processing command unit instructs the machine learning unit to execute, as the machine learning processing, a learning process for creating a model for estimating data related to tension in the conveying unit, and a process for estimating an adjustment action for the tension in the conveying unit, based on the state data observed by the state observation unit and the judgment data created by the judgment data creation unit; The machine learning unit is a reward calculation unit that calculates a reward indicating a value of the tension adjustment action of the predetermined conveying unit based on the determination data; a learning unit that creates a reinforcement learning model that learns the value of the adjustment action with respect to an operating state of the industrial machine based on the state data and the reward; a model storage unit that stores the model created by the learning unit; and an estimation unit that estimates an adjustment action of the tension of the conveying unit using a reinforcement learning model stored in the model storage unit based on the state data, and The control device further includes a tension adjustment unit that adjusts the tension of the conveying section based on the tension adjustment action of the conveying section estimated by the estimation unit. The control device according to claim 1 .

5. The data acquisition unit acquires data related to a machine configuration, data related to a workpiece, and data related to operating conditions as data related to an operating state of the industrial machine, The acquired data storage unit stores data on a machine configuration, data on a workpiece, and data on operating conditions of the industrial machine, which are acquired by the data acquisition unit; The machine learning data creation unit includes a state observation unit that creates state data including data related to a machine configuration, data related to a workpiece, and data related to operating conditions based on the data stored in the acquired data storage unit, the machine learning processing command unit commands the machine learning unit to execute a process of estimating an adjustment action for the tension of the transport unit as the machine learning processing, based on the state data observed by the state observation unit; The machine learning unit includes an estimation unit that estimates an adjustment action of the tension of the conveying unit using a reinforcement learning model that has learned the value of the adjustment action of the tension of the conveying unit with respect to the operating state of the industrial machine based on the state data, and The control device further includes a tension adjustment unit that adjusts the tension of the conveying section based on the tension adjustment action of the conveying section estimated by the estimation unit. The control device according to claim 1 .

Citation Information

Patent Citations

  • Conveyor belt device and driving method for waste aggregate transportation

    KR101930066B1

  • Track tensiong monitoring method and system in construction machinery

    KR1020200109459A

  • Semiconductor imprint device and operating method of semiconductor imprint device

    US20190361340A1

  • Conveyor device for conveying food

    WO2014103886A1

  • Mathematical model calculation apparatus for seat material production line, and control apparatus

    WO2018158828A1