Data collection plan generation device and data collection plan generation method

Abstract

A data collection plan generation apparatus generates a data collection plan consisting of a combination of motion conditions for acquiring data from operating industrial machinery. The apparatus sets the specifications of the industrial machinery, extracts constraints related to the operation of the industrial machinery, and based on these specifications and constraints, generates multiple motion conditions comprising sets of parameter values ​​related to the set specifications. The apparatus then generates and outputs a data collection plan based on the generated motion conditions.

Description

Technical Field

[0001] The present invention relates to a data collection plan generation apparatus and a data collection plan generation method for generating a data collection plan consisting of a combination of motion conditions when data is acquired from a moving machine. Background Technology

[0002] Industrial machinery such as machine tools and robots are installed in manufacturing sites such as factories. Multiple operators operate these machines to manufacture products. Due to heat generated by the movement of various parts and changes in ambient temperature, the components of the industrial machinery expand or contract. This expansion or contraction affects the positioning of the movable parts. Therefore, in situations requiring precise movements of the industrial machinery, such as precision machining or robot-assembled parts, it is necessary to correct the position of the movable parts based on the thermal state of the machinery—a process known as thermal displacement correction. One method for thermal displacement correction involves using machine learning to estimate the correction amount based on motion information obtained from the industrial machinery and the temperature of various parts detected by sensors (e.g., Patent Document 1). This method of performing thermal displacement correction using machine learning is also well-known.

[0003] Furthermore, if industrial machinery operates continuously on the manufacturing site, abnormalities may sometimes occur due to factors such as the deterioration of the components that make up the machinery over the years, changes in external temperature, vibration, or operator errors in settings. When an abnormality is determined to have occurred in the industrial machinery, the operator stops the operation of the malfunctioning machinery, removes the cause of the abnormality, and then restarts the machinery to continue work.

[0004] The normal / abnormal operating status of industrial machinery is monitored, for example, by a management device. The device managing the operating status of industrial machinery monitors, for example, time-series data such as the position, speed, and torque of motors detected by each piece of industrial machinery via a network; data representing changes in signals at predetermined times; and time-series data such as vibrations, sounds, and dynamic images detected by sensors installed on the industrial machinery. The operating status of the industrial machinery is managed by detecting changes in each data point relative to the passage of time (e.g., Patent Documents 2 and 3). Furthermore, as a method for determining the operating status of industrial machinery, there is a method based on physical quantities detected from the industrial machinery by sensors, etc. Methods using machine learning as a determination method are known.

[0005] In this way, machine learning can be applied in various situations on the manufacturing floor.

[0006] Existing technical documents

[0007] Patent documents

[0008] Patent Document 1: Japanese Patent Application Publication No. 2019-166603

[0009] Patent Document 2: Japanese Patent No. 6453504

[0010] Patent Document 3: Japanese Patent Application Publication No. 2019-012473 Summary of the Invention

[0011] The problem that the invention aims to solve

[0012] To utilize machine learning for tasks such as thermal displacement correction and operational status determination in industrial machinery, data collected from the machinery is required. To generate a good learning model—one capable of estimating appropriate thermal displacement correction amounts and determining the machinery's operational status—data that uniformly covers the machinery's specifications is needed. Therefore, data collection must cover the machinery's specifications under various operating conditions. However, arbitrarily increasing the data volume to uniformly cover the specifications significantly increases development time. This is especially true when a single data collection session is lengthy. In other words, data collection must be planned to ensure that the collected data does not excessively or insufficiently cover the machinery's specifications. This challenge is not limited to machine learning but also arises in operational experiments using other industrial machinery or systems.

[0013] Therefore, there is a need for a structure that can generate appropriate data collection plans based on general settings such as the function and specifications of the machinery.

[0014] Methods for solving problems

[0015] One aspect of the present invention is a data collection plan generation apparatus that generates a data collection plan consisting of a combination of motion conditions when data is acquired from a moving machine. The data collection plan generation apparatus comprises: a specification setting unit that sets the specifications of the machine; a constraint condition extraction unit that extracts constraint conditions related to the operation of the machine; a motion condition generation unit that generates multiple motion conditions, including a set of values ​​of parameters related to the specifications set by the specification setting unit, based on the specifications and the constraint conditions; a plan generation unit that generates a data collection plan based on the motion conditions generated by the motion condition generation unit; and an output unit that outputs the data collection plan generated by the plan generation unit.

[0016] Another aspect of the present invention is a data collection plan generation method, wherein the data collection plan generation device generates a data collection plan consisting of a combination of motion conditions when data is acquired from a moving machine, wherein the data collection plan generation method performs the following steps: a specification setting step, setting the specifications of the machine; a constraint extraction step, extracting constraints related to the movement of the machine; a motion condition generation step, generating a plurality of motion conditions, including a set of values ​​of parameters related to the specifications set in the specification setting step, based on the specifications and the constraints; a plan generation step, generating a data collection plan based on the motion conditions generated in the motion condition generation step; and an output step, outputting the data collection plan generated in the plan generation step.

[0017] Invention Effects

[0018] According to one aspect of the present invention, an appropriate data collection plan can be generated for collecting suitable data from general setpoints such as the function and specifications of the machine, which is expected to achieve stable performance improvements and reductions in development time for machine learning functions. Attached Figure Description

[0019] Figure 1 This is a schematic hardware structure diagram of a data collection plan generation device according to one embodiment.

[0020] Figure 2 This is a schematic functional block diagram of the data collection plan generation device according to the first embodiment.

[0021] Figure 3 This is a diagram representing an example of an action condition.

[0022] Figure 4 This is an explanation Figure 2 A diagram illustrating the method for generating action conditions in the action condition generation unit of the data collection plan generation device.

[0023] Figure 5 This diagram illustrates other methods for generating motion conditions in the motion condition generation section.

[0024] Figure 6 This diagram illustrates other methods for generating motion conditions in the motion condition generation section.

[0025] Figure 7 This diagram illustrates other methods for generating motion conditions in the motion condition generation section.

[0026] Figure 8 This is a diagram representing an example of a selection rule. Detailed Implementation

[0027] The embodiments of the present invention will now be described with reference to the accompanying drawings.

[0028] Figure 1 This is a schematic hardware structure diagram showing the main parts of a data collection plan generation apparatus according to an embodiment of the present invention.

[0029] The data collection plan generation device 1 of the present invention can be installed, for example, as a control device for controlling industrial machinery. Alternatively, it can be installed on a personal computer arranged alongside the control device for controlling industrial machinery, a personal computer connected to the control device via a wired / wireless network, a unit computer, a fog computer, or a cloud server. In this embodiment, an example is shown where the data collection plan generation device 1 is installed on a personal computer connected to the control device for controlling industrial machinery via a network.

[0030] The CPU 11 of the data collection plan generation apparatus 1 in this embodiment is a processor that controls the entire data collection plan generation apparatus 1. The CPU 11 reads the system program stored in the ROM 12 via the bus 22 and controls the entire data collection plan generation apparatus 1 according to the system program. Temporary calculation data, display data, and various data input from external sources are temporarily stored in the RAM 13.

[0031] The non-volatile memory 14 is composed of, for example, a memory backed up by a battery (not shown), an SSD (Solid State Drive), etc., and maintains its storage state even when the power supply to the data collection plan generation device 1 is disconnected. Data read from the external device 72 via interface 15, data input via input device 71, and data obtained from the control device 3 via interface 20 are stored in the non-volatile memory 14. The data stored in the non-volatile memory 14 can also be expanded in RAM 13 during execution / use. Furthermore, various system programs, such as known parsing programs, are pre-written into ROM 12.

[0032] Interface 15 is an interface for connecting the CPU 11 of the data collection plan generation device 1 to an external device 72 such as a USB device. Data obtained from other industrial machinery can be read from the external device 72. Furthermore, data processed within the data collection plan generation device 1 can be stored in an external storage unit (not shown) via the external device 72.

[0033] Interface 20 is used to connect the CPU 11 of the data collection plan generation device 1 to a wired or wireless network 5. Control devices 3, fog computers, cloud servers, etc., are connected to the network 5, and exchange data with the data collection plan generation device 1.

[0034] In the display device 70, data obtained as a result of executing various data and programs read into the memory is output and displayed via the interface 17. In addition, the input device 71, which consists of a keyboard, indicator devices, etc., delivers instructions and data based on the operator's operation to the CPU 11 via the interface 18.

[0035] Figure 2 This is a schematic block diagram illustrating the functions of the data collection plan generation apparatus 1 according to the first embodiment of the present invention.

[0036] The data collection plan generation device 1 of this embodiment possesses various functions through Figure 1 The CPU 11 of the data collection plan generation device 1 shown executes a system program to control the operation of each part of the data collection plan generation device 1. The data collection plan generation device 1 of this embodiment has the function of generating an appropriate data collection plan for collecting appropriate data from general set values ​​such as the function and specifications of the machine being operated. Furthermore, in this embodiment, the generation of a data collection plan for machine tool thermal displacement correction using machine learning has been shown as an example; however, the data collection plan generation device 1 of this embodiment can also be applied to situations such as determining the operating state of a machine using machine learning.

[0037] The data collection plan generation device 1 of this embodiment includes a specification setting unit 100, a constraint condition extraction unit 110, a parameter selection unit 120, an operation condition generation unit 130, a data collection plan generation unit 135, and an output unit 140. Furthermore, the RAM 13 or non-volatile memory 14 of the data collection plan generation device 1 is pre-prepared with: a specification storage unit 200, which serves as a region for storing specification-related data acquired and set from the input device 71, external device 72, control device 3, etc.; a data collection plan storage unit 210, which serves as a region for storing data related to the generated machine's data collection plan; and a selection rule storage unit 220, which serves as a region pre-stored with rules for selecting parameters as operation conditions.

[0038] Specification setting unit 100 passed Figure 1The CPU 11 of the data collection plan generation device 1 shown executes the system program read from ROM 12. This is mainly achieved by the CPU 11 performing computational processing using RAM 13 and non-volatile memory 14, and input control processing based on interfaces 15, 18, or 20. The specification setting unit 100 acquires and sets specification-related information associated with the operation of the industrial machinery 4 controlled by the control device 3. This specification-related information may include, for example, information specifying predetermined ranges for parameters related to the machinery's operation, such as the travel range of each axis, maximum travel speed of each axis, maximum travel acceleration of each axis, maximum torque of each axis, maximum load capacity, maximum spindle rotation speed, maximum spindle torque, maximum coolant flow rate, maximum coolant temperature, and minimum coolant temperature; and information specifying predetermined ranges for data related to the machinery's operating environment, such as maximum ambient temperature and minimum ambient temperature. This specification-related information can be obtained from information stored in the control device 3, a fog computer (not shown), a cloud server, etc., or read from external devices 72. Alternatively, it can be set by the operator from the input device 71 as needed. The data obtained by the specification setting unit 100 is stored in the specification storage unit 200.

[0039] Constraint Extraction Unit 110 Figure 1 The CPU 11 of the data collection plan generation device 1 shown executes the system program read from the ROM 12, and the main implementation is achieved by the CPU 11 performing arithmetic processing using the RAM 13 and the non-volatile memory 14. The constraint condition extraction unit 110 extracts constraints related to the machine's operating conditions from the specifications set by the specification setting unit 100 and other information input by the operator from the input device 71. For example, the constraint condition extraction unit 110 may also extract the maximum and minimum values ​​of parameters related to the operation of each machine as constraints based on information related to the specifications set by the specification setting unit 100. For example, the constraint condition extraction unit 110 may also extract the maximum and minimum values ​​of data related to the operating environment of each machine as constraints based on information related to the specifications set by the specification setting unit 100. In addition, the constraint conditions may also be a list of predetermined parameters, values ​​that can be obtained in the operating environment, information representing distribution, or conditional expressions of relationships between predetermined parameters. For example, the constraint condition extraction unit 110 may also prompt the operator to input restrictions related to the size of the accompanying machine, restrictions on continuous operating time, restrictions on the number of operations, etc., and extract the constraint conditions based on the information input from the input device 71. The constraint extraction unit 110 can also extract typical spindle rotation speed, spindle rotation time, and other information as constraints from the machining program executed by the control device 3. The constraints extracted by the constraint extraction unit 110 are stored together with specification-related information in the specification storage unit 200.

[0040] Parameter selection unit 120 Figure 1 The CPU 11 of the data collection plan generation device 1 shown executes the system program read from the ROM 12, and is mainly implemented by the CPU 11 performing arithmetic processing using RAM 13 and non-volatile memory 14. The parameter selection unit 120 selects parameters that are the objects to be generated as operating conditions of the machine. For example, the parameter selection unit 120 may also display the parameters that can be selected as operating conditions of the industrial machine 4 on the display device 70 via the output unit 140, based on the specifications set by the specification setting unit 100, so that the operator can select the parameters as operating conditions of the industrial machine 4 from them. Alternatively, the parameter selection unit 120 may also allow the operator to select parameters as operating conditions of the industrial machine 4 according to the selection rules stored in the selection rule storage unit 220, based on the purpose related to the operating conditions of the industrial machine 4. In such a case, as... Figure 8 As illustrated, selection rules are established in advance in the selection rule storage unit 220, which stores a list of parameters to be selected for the purpose of data collection. The parameter selection unit 120 displays the purpose of data collection included in the selection rules to the display device 70, allowing the operator to select that purpose, thereby making it easy to select appropriate parameters. Furthermore, regarding parameters whose specifications are set in the specification setting unit 100, the parameter selection unit 120 is not a necessary structure when all operating conditions are selected. In such cases, even without the parameter selection unit 120, the effects of this invention can be fully enjoyed.

[0041] Action condition generation unit 130 passes Figure 1 The data collection plan generation device 1 shown has a CPU 11 that executes the system program read from ROM 12. This is mainly achieved by the CPU 11 performing calculations using RAM 13 and non-volatile memory 14. The operation condition generation unit 130 generates the operation conditions for the industrial machine 4. For example... Figure 3 As illustrated, each motion condition generated by the motion condition generation unit 130 includes a set of parameter values ​​selected by the parameter selection unit 120.

[0042] Data Collection Plan Generation Department 135 passed Figure 1 The CPU 11 of the data collection plan generation device 1 shown executes the system program read from the ROM 12, and the main implementation is achieved by the CPU 11 performing arithmetic processing using the RAM 13 and the non-volatile memory 14. The data collection plan generation unit 135 generates an appropriate data collection plan for collecting appropriate data from the industrial machine 4. That is, in a vector space of vectors containing parameters included in the action conditions, a combination of action conditions for data collection is generated so that the values ​​of the parameters included in each action condition are evenly distributed in the vector space.

[0043] In order to distribute the motion conditions uniformly, the data collection plan generation unit 135 may, for example, determine a value for each parameter included in the motion conditions at a predetermined interval between the minimum and maximum values ​​of the parameter, and generate a set of the determined values ​​as a data collection plan.

[0044] Figure 4 This example illustrates how a machine learning device, designed to predict thermal displacement correction values, collects appropriate learning data, selecting ambient temperature and spindle rotation speed as parameters to generate a data collection plan. Figure 4 In the example, for each of the ambient temperature and spindle rotation speed, a value is determined between the minimum and maximum values ​​at a predetermined interval, and a set of these determined values ​​is generated as a data collection plan.

[0045] In order to distribute the motion conditions evenly, the data collection plan generation unit 135 may, for example, determine a value for each parameter included in the motion conditions at a predetermined interval between the minimum and maximum values ​​of the physical quantity associated with the parameter, and generate a set of the determined values ​​as a data collection plan.

[0046] Figure 5 Indicates and Figure 4 Similarly, in order to collect appropriate learning data for a machine learning device that predicts the value of thermal displacement correction, an example is given of generating a data collection plan by selecting ambient temperature and spindle rotation speed as parameters. Figure 5 In the example, regarding ambient temperature, the value is determined at a predetermined interval between the minimum and maximum values, and regarding spindle rotation speed, the value of spindle rotation speed is determined by the heat generated by the spindle at predetermined intervals, and a set of these determined values ​​is generated as a data collection plan.

[0047] The data collection plan generation unit 135 can also reduce the number of action conditions when the number of action conditions included in the data collection plan to be generated exceeds the limit on the number of actions extracted by the constraint extraction unit 110. For example, such as Figure 6 As illustrated, motion conditions can be reduced uniformly or reduced in a manner that distributes motion conditions evenly according to prescribed rules.

[0048] Conversely to the example above, the data collection plan generation unit 135 can also repeatedly generate action conditions within the limit of the number of actions extracted by the constraint condition extraction unit 110. For example, it can also be as follows: Figure 7As illustrated, after generating action conditions from sets of minimum and maximum values ​​for each parameter, the point with the thinnest density of action conditions is selected in the parameter vector space, and the process of generating sets of values ​​for that point as action conditions is repeated repeatedly. Here, the density of action conditions can also vary according to the axes in the vector space. By thinning the density of the axes of parameters whose values ​​have almost no impact on the machine learning results, it is possible to avoid increasing the number of action conditions. Through this selection, sets of parameter values ​​for any number of actions can be generated, uniformly distributed in the parameter vector space.

[0049] The data collection plan extracted by the data collection plan generation unit 135 is stored in the data collection plan storage unit 210. The data collection plan stored in the data collection plan storage unit 210 can also be displayed on the display device 70 via the output unit 140. In addition, the output unit 140 can also be configured to not only display output to the display device 70, but also perform output processing such as output processing to the external device 72 or output processing via the network 5. In this case, the data collection plan stored in the data collection plan storage unit 210 can also be stored in an external storage device (not shown) via the external device 72 and used, or sent via the network to the control device 3, a fog computer (not shown), or a cloud server for use.

[0050] By using the data collection plan generation device 1 of this embodiment, which has the above-described structure, the operator can generate an appropriate data collection plan for collecting appropriate data from general settings such as the function and specifications of the machine.

[0051] The present invention has been described above as an embodiment of the invention, but the present invention is not limited to the example of the above embodiment, and can be implemented in various ways by applying appropriate modifications.

[0052] For example, in the above-described embodiments, the data collection plan generation device 1 generates operating conditions based on the specifications and constraints of the machinery. However, the specifications and constraints processed by the data collection plan generation device 1 are not limited to the specifications and constraints of the machinery. For example, it can also generate operating conditions for a system based on the specifications and constraints of each component of a system composed of multiple machines and devices, such as a production system or an experimental system. Even in such cases, the functions of the data collection plan generation device 1 of this application can be used effectively.

[0053] Symbol Explanation

[0054] 1 Data collection plan generation device

[0055] 3 Control devices

[0056] 4 Industrial Machinery

[0057] 5 Networks

[0058] 6 sensors

[0059] 11 CPU

[0060] 12ROM

[0061] 13 RAM

[0062] 14 Non-volatile memory

[0063] Interfaces 15, 17, 18, and 20

[0064] 22 bus

[0065] 70 display devices

[0066] 71 Input Device

[0067] 72 External Devices

[0068] 100 Specification Setting Department

[0069] 110 Constraint Extraction Department

[0070] 120 parameter selection section

[0071] 130 Action Condition Generation Department

[0072] 135 Data Collection Plan Generation Department

[0073] 140 Output Unit

[0074] 200 specification storage unit

[0075] 210 Data Collection Project Storage Department

[0076] 220 Select rule storage department.

Claims

1. A data collection plan generation device, which generates a data collection plan consisting of a combination of motion conditions when data is acquired from a moving machine, characterized in that, The data collection plan generation device includes: The specification setting unit sets the specifications of the machine. A constraint extraction unit extracts constraints related to the operation of the machine. The action condition generation unit generates multiple action conditions based on the specifications and the constraints, each containing a set of values ​​of parameters related to the specifications set by the specification setting unit. The data collection plan generation unit generates a data collection plan based on the action conditions generated by the action condition generation unit. as well as The output unit outputs the data collection plan generated by the data collection plan generation unit. If the number of action conditions in the data collection plan to be generated is greater than the limit on the number of actions extracted by the constraint extraction unit, the data collection plan generation unit reduces the number of action conditions. Alternatively, if the number of action conditions in the data collection plan to be generated is less than the limit on the number of actions extracted by the constraint extraction unit, the data collection plan generation unit repeatedly generates action conditions within the limit on the number of actions.

2. The data collection plan generation device according to claim 1, characterized in that, The data collection plan generation device further includes a parameter selection unit that selects parameters included in the operating conditions of the machine based on the specifications and the constraints. The action condition generation unit generates multiple action conditions based on the specifications and the constraints, each containing a set of values ​​for parameters selected by the parameter selection unit.

3. A data collection plan generation method for a data collection plan generation device, wherein the data collection plan generation device generates a data collection plan consisting of a combination of motion conditions when data is acquired from a moving machine, characterized in that, The data collection plan generation method performs the following steps: Specification setting steps: Set the specifications of the machine; The constraint extraction step extracts the constraints related to the operation of the machine. The action condition generation step generates multiple action conditions, which include a group of values ​​of parameters related to the specifications set in the specification setting step, based on the specifications and the constraints. The data collection plan generation step generates a data collection plan based on the action conditions generated in the action condition generation step. as well as The output step outputs the data collection plan generated in the data collection plan generation step. In the data collection plan generation step, if the number of action conditions included in the data collection plan to be generated is greater than the limit on the number of actions extracted in the constraint extraction step, the action conditions are reduced; or, if the number of action conditions included in the data collection plan to be generated is less than the limit on the number of actions extracted in the constraint extraction step, the action conditions are repeatedly generated within the limit on the number of actions.

4. The data collection plan generation method according to claim 3, characterized in that, Following the constraint extraction step, a parameter selection step is also performed, in which parameters included in the machine's operating conditions are selected based on the specifications and the constraints. In the action condition generation step, based on the specifications and the constraints, multiple action conditions are generated, which are groups of values ​​of the parameters selected in the parameter selection step.

Images