Numerical control device, learning device, estimation device, and operation screen display method for numerical control device

Abstract

A numerical control device (1) is provided with: a learning data acquisition unit that acquires learning data including operation log information indicating operation by a user on an operation screen of the numerical control device (1) that controls a machine tool and information on the operation screen to be operated; and operation state information indicating the state of the machine tool when the operation indicated by the operation log information is performed. And a model generation unit that uses the learning data and, on the basis of the operation log information and the operation state information, generates a trained model for estimating screen display data for displaying a new operation screen including information required by a user extracted from a plurality of existing operation screens.

Description

Technical Field

[0001] The present invention relates to a numerical control device, a learning device, an inference device, and a method for displaying an operation screen of a numerical control device. Background Art

[0002] An operation screen of a numerical control device that controls a machine tool is used in various situations. An operation screen that is easy to operate according to the situation to be used is required. For example, Patent Document 1 discloses a numerical control device that uses machine learning and can reorder menus in an operation screen according to an operator.

[0003] Patent Document 1: Japanese Patent No. 6898479 Gazette Summary of the Invention

[0004] However, according to the above prior art, when information requested by a user is spread over a plurality of operation screens, the user has to search for the requested information from the plurality of operation screens. Therefore, there still exists a problem that the user's operation sometimes becomes complicated.

[0005] The present invention has been made in view of the above circumstances, and an object thereof is to obtain a numerical control device that can improve the work efficiency of a user even when information requested by the user is spread over a plurality of existing operation screens.

[0006] In order to solve the above problems and achieve the object, the numerical control device according to the present invention is characterized by having: a learning data acquisition unit that acquires learning data including operation log information representing an operation of a user on an operation screen of a numerical control device that controls a machine tool and information on an operation screen of an operation object, and action state information representing a state of the machine tool when the operation shown in the operation log information is performed; and a model generation unit that uses the learning data and generates a trained model for inferring screen display data for displaying a new operation screen including information extracted from a plurality of existing operation screens, based on the operation log information and the action state information.

[0007] Advantages of the Invention

[0008] The numerical control device according to the present invention has the following effect, that is, it can provide a numerical control device that can improve the work efficiency of a user even when information requested by the user is spread over a plurality of existing operation screens. Brief Description of the Drawings

[0009] Figure 1 It is a diagram showing a functional configuration of the numerical control device according to Embodiment 1.

[0010] Figure 2This is a diagram showing the detailed structure of a learning device related to a numerical control device of a machine tool.

[0011] Figure 3 This is a flowchart related to the learning process of the learning device.

[0012] Figure 4 This is a structural diagram of an inference device related to a numerical control device of a machine tool.

[0013] Figure 5 This is a flowchart related to the inference process of the inference device.

[0014] Figure 6 This is an explanatory diagram related to the effects of Embodiment 1.

[0015] Figure 7 This is a diagram showing an example of a new operation screen displayed by the numerical control device related to Embodiment 3.

[0016] Figure 8 This is a diagram showing the functional structure of the numerical control device related to Embodiment 4.

[0017] Figure 9 This is a diagram showing an example of a method for sharing a trained model in Embodiment 5.

[0018] Figure 10 This is a diagram for explaining the functions of the numerical control device related to Embodiment 6.

[0019] Figure 11 This is a diagram showing an example of a new operation screen displayed by the numerical control device related to Embodiment 7.

[0020] Figure 12 This is a diagram showing the functional structure of the numerical control device related to Embodiment 11.

[0021] Figure 13 This is a diagram showing dedicated hardware for implementing the functions of the numerical control device related to Embodiments 1 to 11.

[0022] Figure 14 This is a diagram showing the structure of a control circuit for implementing the functions of the numerical control device related to Embodiments 1 to 11. Detailed Embodiments

[0023] Hereinafter, based on the accompanying drawings, a numerical control device, a learning device, an inference device, and a method for displaying an operation screen of the numerical control device related to the embodiments of the present invention will be described in detail.

[0024] Embodiment 1.

[0025] Figure 1This is a diagram showing the functional structure of the numerical control device 1 according to Embodiment 1. The numerical control device 1 has a control unit 11, an operation log acquisition unit 12, a data acquisition unit 13, a learning device 14, a trained model storage unit 15, an inference device 16, and an output unit 17. The numerical control device 1 is a device that controls a machine tool according to a machining program. The numerical control device 1 controls a drive unit 2 provided in the machine tool, thereby controlling the operation of the machine tool. The drive unit 2 includes an amplifier, a motor, and the like. The numerical control device 1 can also control peripheral equipment 3. In addition, the numerical control device 1 can use a display device 4 to output an operation screen for operating the numerical control device 1. Further, in Figure 1 this case, the display device 4 is a device separate from the numerical control device 1, but the display device 4 may also be built into the numerical control device 1. In addition, in Figure 1 this case, the input device 5 is a device separate from the numerical control device 1, but the input device 5 may also be built into the numerical control device 1.

[0026] The control unit 11 is connected to the drive unit 2 and the peripheral equipment 3 that are the control objects of the numerical control device 1. The control unit 11 has functions of controlling the amplifier and the motor provided in the drive unit 2 and controlling the peripheral equipment 3, and acquires operation state information indicating the operation states of the respective devices from the drive unit 2 and the peripheral equipment 3, and outputs the acquired operation state information to the data acquisition unit 13. The operation state information can include, for example, an operation mode such as being in machining, in changeover adjustment, or in alarm generation, the state of a PLC (Programmable Logic Controller) signal, the state of a machining program for controlling the drive unit 2 of the machine tool, and the like. In addition, the operation state information can include the type of alarm, the content of the alarm, and the like when the operation mode is in alarm generation.

[0027] The operation log acquisition unit 12 acquires operation log information indicating the operations of the user on the operation screen of the numerical control device 1. The operation log information includes information on the operation screen that is the operation object. The information on the operation screen can include, for example, identification information for identifying the operation screen, hierarchical structure information indicating the relationship with other operation screens when the operation screen has a hierarchical structure, screen display data for displaying the operation screen, and the like. The operation log acquisition unit 12 can acquire the operation log information based on the input information obtained from the input device 5. The input device 5 is, for example, a keyboard that accepts key input, a pointing device that accepts the designation of a coordinate position on the operation screen, a microphone that accepts voice input, and the like. In addition, the pointing device is, for example, a mouse, a touch sensor, and the like. The touch sensor can be a touch panel in which the input device 5 and the display device 4 are integrated, or a single touchpad.

[0028] The data acquisition unit 13 acquires various data to be used by the numerical control device 1. The data acquisition unit 13 can output the acquired data to the learning device 14 and the inference device 16 respectively. The data acquisition unit 13 can acquire operation log information from the operation log acquisition unit 12 and acquire operation state information from the control unit 11.

[0029] The learning device 14 generates a trained model. This trained model uses machine learning to infer the screen display data for displaying a new operation screen of the numerical control device 1. The learning device 14 stores the trained model as the result of learning in the trained model storage unit 15. The inference device 16 uses the trained model to infer the screen display data for displaying a new operation screen, and outputs the screen display data as the inference result to the output unit 17.

[0030] Based on the screen display data output by the inference device 16, the output unit 17 outputs a new operation screen using the display device 4. In addition, the screen display data can include at least one of the design of the new operation screen, the information output timing, the information output method, and the operation guidance which is information for assisting the user's operation. The design of the operation screen can include information such as the position on the screen where the information required by the user is displayed and the size of the information required by the user to be displayed. The information output timing represents the information about the time when information is output to the user. The information output method represents the presence or absence of emphasized display, animation display, and screen switching. The operation guidance is supplementary information for the information output to the user, information such as the operation method of the screen, etc., for assisting the user's operation. In addition, if new screen display data is generated, a new operation screen can be displayed. Therefore, in the following description, sometimes generating new screen display data is expressed as "generating a new operation screen". In addition, the timing at which the numerical control device 1 generates a new operation screen using the inference device 16 can be, for example, the timing of a change in the operation mode of the machine tool or the timing when the user operates the operation screen.

[0031] <Learning stage>

[0032] Figure 2 This is a diagram showing the detailed structure of the learning device 14 related to the numerical control device 1 of the machine tool. The learning device 14 has a learning data acquisition unit 141 and a model generation unit 142.

[0033] The learning data acquisition unit 141 acquires operation log information and operation state information as learning data.

[0034] The model generation unit 142 learns the screen display data based on the learning data including the operation log information and the action status information. That is, a trained model for inferring the screen display data is generated, and the screen display data is used to display a new operation screen based on the operation log information and the action status information of the numerical control device 1 of the machine tool. In addition, the "new operation screen" includes information extracted from a plurality of existing operation screens. Here, the "information extracted from a plurality of existing operation screens" is, for example, the following information, that is, in the operation log information, the information displayed by the user in the past operations, etc., is information determined to be required by the user based on the operation log information and the action status information. In addition, the "information extracted from a plurality of existing operation screens" means that the extracted "information" is explored from the information included in the existing "plurality of" operation screens, and includes not only the case where the information extracted as a result is separately included in the information of the existing plurality of operation screens, but also the case where the information extracted as a result is included in one of the existing operation screens. In addition, the "new operation screen" includes information that helps improve the operation efficiency of the operation using the operation screen. Among the "information that helps improve the operation efficiency", information for reducing the operation burden of the user is included. As the operation burden of the user, for example, the number of operations, the number of screen conversions, and the operation time are cited. As the information for reducing the operation burden of the user, for example, the information extracted from the existing operation screens, the information obtained by combining the information extracted from the existing operation screens, operation guidance, etc. are cited. In addition, among the "information that helps improve the operation efficiency", for example, information that has a good influence on the operation result is included. As the information that has a good influence on the operation result, for example, operation guidance and other information for preventing the user's operation errors are cited.

[0035] The learning algorithm used by the model generation unit 142 can use known algorithms such as supervised learning, unsupervised learning, and reinforcement learning. As an example, the case of applying reinforcement learning will be described. In reinforcement learning, an agent as an acting entity in a certain environment observes the environmental parameters representing the current state and determines the action to be taken. The environment changes dynamically through the actions of the agent, and a reward is given to the agent in response to the change of the environment. The agent repeats the above actions and learns the action policy that obtains the most rewards after a series of actions. As a representative method of reinforcement learning, Q-learning (Q-Learning) and TD-learning (TD-Learning) are known. For example, in the case of Q-learning, the general update formula of the action value function Q(s, a) is represented by Equation (1).

[0036]

Equation 1

[0037]

[0038] In equation (1), s t represents the state of the environment at time t, and a t represents the action at time t. Through the action a t , the state changes to s t+1 . r t+1 represents the reward brought about by the change in its state, γ represents the discount rate, and α represents the learning coefficient. In addition, γ is a value in the range of 0 < γ ≤ 1, and α is a value in the range of 0 < α ≤ 1. The operation log information becomes the action a t , and the action state information becomes the state s t . The best action a t for the state s t at time t is learned.

[0039] Through the update formula represented by equation (1), if the action value Q of the action a with the highest Q value at time t + 1 is greater than the action value Q of the action a executed at time t, the action value Q is increased. In the opposite case, the action value Q is decreased. In other words, the action value function Q(s, a) is updated so that the action value Q of the action a at time t approaches the best action value at time t + 1. Thus, the best action value in a certain environment is continuously propagated in sequence as the action value in its previous environment.

[0040] As described above, in the case of generating a trained model through reinforcement learning, the model generation unit 142, as Figure 2 shown, has a reward calculation unit 143 and a function update unit 144.

[0041] The reward calculation unit 143 calculates the reward based on the operation log information and the action status information. The reward calculation unit 143 calculates the reward r based on the reward criteria. The reward criteria includes a reward increase criterion for increasing the reward r and a reward decrease criterion for decreasing the reward r. Here, it is preferable to increase or decrease the reward r according to the user's operation efficiency. Therefore, the reward criteria can be based on information indicating the user's operation efficiency, such as a criterion based on at least one of the number of operations, the number of screen transitions, and the operation time. The counting of the number of operations, the number of screen transitions, and the operation time is performed for each action mode during the operation of the machine tool. For example, when the number of operations decreases, the reward r is increased by giving a reward of "1", and when the number of operations increases, the reward r is decreased by giving a reward of "-1". In addition, the reward r can be increased when the number of screen transitions decreases, and the reward r can be decreased when the number of screen transitions increases. In addition, the reward r can be increased when the operation time decreases, and the reward r can be decreased when the operation time increases. In addition, here, it is determined whether to increase or decrease the reward based on one of the number of operations, the number of screen transitions, and the operation time, but it can also be determined based on two of the number of operations, the number of screen transitions, and the operation time, or it can be determined based on all of the number of operations, the number of screen transitions, and the operation time. In addition, information other than the number of operations, the number of screen transitions, and the operation time can be further used to determine whether to increase or decrease the reward r. The reward calculation unit 143 only needs to increase the reward r when it can be determined that the user's operation efficiency has improved, and decrease the reward r when it can be determined that the user's operation efficiency has decreased.

[0042] The function update unit 144 updates the function for determining the screen display data according to the reward r calculated by the reward calculation unit 143, and outputs it to the trained model storage unit 15. For example, in the case of Q-learning, the action value function Q(s t , a t ) represented by the formula (1) is used as the function for calculating the screen display data.

[0043] The learning device 14 repeats the above learning. The trained model storage unit 15 stores the updated action value function Q(s t , a t ), that is, the trained model.

[0044] Next, Figure 3 , the process learned by the learning device 14 will be described. Figure 3 is a flowchart related to the learning process of the learning device 14.

[0045] The learning data acquisition unit 141 acquires operation log information and operation status information as learning data (step S10).

[0046] The model generation unit 142 calculates the reward r based on the operation log information and the operation status information (step S11). Specifically, the reward calculation unit 143 can calculate the reward r using a predetermined reward criterion, for example. The reward criterion includes a reward increase criterion for increasing the reward r and a reward decrease criterion for decreasing the reward r. As described above, the reward criterion can be based on information indicating the work efficiency of the user, for example, a criterion based on at least one of the number of operations, the number of screen transitions, and the operation time. For example, in the case of using a reward criterion based on the number of operations, the reward calculation unit 143 counts the number of operations for each action pattern based on the operation log information and the operation status information. When the number of operations decreases, it is determined that the reward increase criterion is satisfied, and when the number of operations increases, it is determined that the reward decrease criterion is satisfied.

[0047] When the reward calculation unit 143 determines that the reward increase criterion is satisfied (step S11: reward increase criterion), it increases the reward r (step S12). When the reward calculation unit 143 determines that the reward decrease criterion is satisfied (step S11: reward decrease criterion), it decreases the reward r (step S13).

[0048] The function update unit 144 updates the action value function Q(s t , a t ) represented by the arithmetic expression (1) stored in the trained model storage unit 15 based on the reward r calculated by the reward calculation unit 143 (step S14).

[0049] The learning device 14 repeatedly executes the above steps S10 to S14, and stores the generated action value function Q(s t , a t ) as a trained model.

[0050] The learning device 14 according to the present embodiment stores the trained model in the trained model storage unit 15 provided outside the learning device 14, but the trained model storage unit 15 may also be provided inside the learning device 14.

[0051] <Effective use phase>

[0052] Figure 4 It is a structural diagram of the inference device 16 related to the numerical control device 1 of the machine tool. The inference device 16 has an inference data acquisition unit 161 and an inference unit 162.

[0053] The inference data acquisition unit 161 acquires operation log information and operation status information.

[0054] The inference unit 162 performs inference on the screen display data using the trained model. That is, by inputting the operation log information and operation status information acquired by the inference data acquisition unit 161 into the trained model, it is possible to infer the screen display data suitable for the operation log information and operation status information.

[0055] In addition, in the present embodiment, it has been described that the trained model learned by the learning device 14 provided in the numerical control device 1 of the machine tool is used to output the screen display data. However, it is also possible to obtain the trained model from other numerical control devices 1 and output the screen display data based on the trained model.

[0056] Next, use Figure 5 to explain the process of using the trained model to obtain the screen display data. Figure 5 is a flowchart related to the inference process of the inference device 16.

[0057] The inference data acquisition unit 161 acquires operation log information and operation status information as inference data (step S20).

[0058] The inference unit 162 inputs the operation log information and operation status information into the trained model stored in the trained model storage unit 15 (step S21) and acquires the screen display data. The inference unit 162 outputs the obtained screen display data to the output unit 17 (step S22).

[0059] The numerical control device 1 uses the output screen display data to cause the display device 4 to display the operation screen (step S23). The operation screen displayed here is a new operation screen that includes the information required by the user extracted from the existing multiple operation screens. Therefore, even when the multiple pieces of information required by the user are separately included in multiple screens in the existing operation screens, it is possible to summarize and confirm the multiple pieces of information through a single new operation screen.

[0060] In addition, in the present embodiment, the case where reinforcement learning is applied to the learning algorithm used by the inference unit 162 has been described. However, it is not limited thereto. Regarding the learning algorithm, in addition to reinforcement learning, supervised learning, unsupervised learning, or semi-supervised learning, etc. can also be applied.

[0061] In addition, as a learning algorithm used in the model generation unit 142, deep learning that learns the extraction of feature quantities themselves can also be used, and machine learning can also be executed according to other known methods, such as neural networks, genetic programming, functional logic programming, support vector machines, etc.

[0062] In addition, the learning device 14 and the inference device 16 can be connected to the numerical control device 1 via a network, for example, and are devices separate from the numerical control device 1. In addition, the learning device 14 and the inference device 16 can also be Figure 1 as shown, built into the numerical control device 1. Also, the learning device 14 and the inference device 16 can exist on a cloud server.

[0063] In addition, the model generation unit 142 can use the learning data obtained from multiple numerical control devices 1 to learn the screen display data. In addition, the model generation unit 142 can obtain the learning data from multiple numerical control devices 1 used in the same area, or can also use the learning data collected from multiple numerical control devices 1 that operate independently in different areas to learn the screen display data. In addition, it is also possible to add the numerical control device 1 that collects the learning data to the object midway or remove it from the object. Also, the learning device 14 after learning the screen display data for a certain numerical control device 1 can be applied to other numerical control devices 1, and re-learning can be performed on the screen display data for the other numerical control devices 1 to update.

[0064] Here, the operation screen generated using the trained model will be described. Figure 6 It is an explanatory diagram related to the effects of the first embodiment. As Figure 6As shown in the upper part of , the existing operation screen has a hierarchical structure. Below the top screen, there are existing screen #1 and existing screen #2. In addition, below existing screen #1, there are existing screen #1-1, existing screen #1-2, and existing screen #1-3. Below existing screen #2, there are existing screen #2-1 and existing screen #2-2. Here, the information required by the user is information #1, information #2, and information #3. Information #1 is included in existing screen #1-1, information #2 is included in existing screen #1-3, and information #3 is included in existing screen #2-2. In this case, if only the existing operation screen is used, after the user performs the operation (1) of switching from the top screen to existing screen #1, then performs the operation (2) of switching from existing screen #1 to existing screen #1-1, and then can reach information #1. Then, the user performs the operation (3) of returning from existing screen #1-1 to existing screen #1, performs the operation (4) of switching from existing screen #1 to existing screen #1-3, and then can reach information #2. And the user further performs the operation (5) of returning from existing screen #1-3 to existing screen #1, performs the operation (6) of switching from existing screen #1 to the top screen, performs the operation (7) of switching from the top screen to existing screen #2, performs the operation (8) of switching from existing screen #2 to existing screen #2-2, and then can reach information #3. That is, in order to reach all of information #1, information #2, and information #3 from the top screen, 8 screen switching operations are required. In contrast, according to the numerical control device 1 according to Embodiment 1, a new operation screen, that is, generation screen #1, including information #1, information #2, and information #3 can be displayed. Thus, the user can confirm a plurality of pieces of information required by the user through 1 screen switching operation.

[0065] Here, a specific example of the operation screen generated by the numerical control device 1 will be described. For example, consider the case where an abnormality occurs in the machine tool or the peripheral device 3. First, the user confirms the state of the device where the abnormality has occurred and the content of the abnormality, and then confirms the countermeasure method for the occurrence of the abnormality. Therefore, in the first example where it is assumed that an alarm notifying an abnormality of the machine tool occurs during machining, for example, an operation screen capable of confirming the countermeasure method corresponding to the state of the machine tool, the detailed information of the alarm, and the content of the alarm is generated as a new operation screen. In this case, "the state of the machine tool" corresponds to Figure 6In the example of Figure 6 "Information #1", "Details of the alarm" is equivalent to "Information #2", and "Response method corresponding to the content of the alarm" is equivalent to "Information #3". Also, in the second example assuming a situation where communication with the peripheral device 3 cannot be performed well, an operation screen is generated that can confirm communication status such as the status of the machine tool, the peripheral device 3, and the number of communication errors of the peripheral device 3, and response methods, etc., as a new operation screen. In this case, "Status of the machine tool, peripheral device 3" is equivalent to

[0066] "Information #1" in the example of

[0067] "Communication status such as the number of communication errors of the peripheral device 3" is equivalent to "Information #2", and "Response method" is equivalent to "Information #3".

[0068] In addition, when an abnormality occurs, if the information required to judge the information requested by the user is insufficient, fault finding can be performed in the form of a dialogue with the user. In the third example, if the input device 5 such as a keyboard or a microphone is used to input the points that the user is troubled about in natural language, the numerical control device 1 can generate a new operation screen corresponding to the input text and voice and display it on the display device 4. For example, if the user inputs "I hope to back up the parameters", the numerical control device 1 generates an operation screen including information and operation guides required to solve the user's problem, such as a parameter backup screen, and makes the display device 4 display it. The dialogue with the user can be repeated multiple times to gradually screen out the problem. In the above first, second, and third examples, the operation screens in the case of an abnormality occurring in a machine tool, etc. have been described. However, in the fourth example assuming normal times other than when an abnormality occurs, the numerical control device 1 generates a new operation screen including the best information extracted from the existing operation screens of the numerical control device 1 according to what the user wants to do, and makes the display device 4 display it. For example, if the input device 5 such as a keyboard or a microphone is used to input what the user wants to do, the numerical control device 1 generates a new operation screen that can achieve the user's purpose from the functions of the numerical control device 1.As described above, the numerical control device 1 according to Embodiment 1 is characterized by including: a learning data acquisition unit 141 that acquires learning data including operation log information indicating operations on an operation screen of the numerical control device 1 for controlling a machine tool and information on the operation screen of the operation target, and motion state information indicating the state of the machine tool when the operations indicated by the operation log information are performed; and a model generation unit 142 that uses the learning data to generate a trained model for inferring screen display data for displaying a new operation screen including information extracted from a plurality of existing operation screens based on the operation log information and the motion state information. Further, the numerical control device 1 is characterized by including: an inference data acquisition unit 161 that acquires inference data including operation log information indicating operations on an operation screen of the numerical control device 1 for controlling a machine tool and information on the operation screen of the operation target, and motion state information indicating the state of the machine tool when the operations indicated by the operation log information are performed; an inference unit 162 that outputs screen display data for displaying a new operation screen based on the operation log information and the motion state information by using the trained model for inferring screen display data for displaying a new operation screen including information extracted from a plurality of existing operation screens; and an output unit 17 that outputs the new operation screen to a display device 4 based on the screen display data. Thereby, an operation screen that is easy to use can be provided to the user according to the motion state of the machine tool. For example, as described in Figure 6 the case where the information required by the user is spread over multiple screens in the existing operation screen, the information required by the user is extracted to generate a new operation screen, thereby reducing the number of user operations, screen transitions, operation time, etc. Thereby, the work efficiency of the user can be improved. In addition, in the method of preparing in advance operation screens corresponding to the motion state of the machine tool, etc., there are multiple combinations of the information required by the user depending on the state of the machine tool, so the number of operation screens prepared in advance increases. Therefore, by generating a new operation screen based on the operation log information and the motion state information as in the numerical control device 1, the number of operation screens prepared in advance can be reduced, and the effect of reducing the memory usage amount can be achieved.

[0069] In addition, a new operation screen that is displayed based on the screen display data inferred by the trained model can aggregate and display multiple pieces of information separately included in multiple existing operation screens. This operation log information obtained as learning data not only represents operations on one operation screen but also can represent the operation history of the user across multiple existing operation screens. For example, when multiple existing operation screens form a hierarchical structure, the operation log information represents the operation history of the user across the hierarchical structure, that is, multiple existing operation screens.

[0070] In addition, the model generation unit 142 includes a reward calculation unit 143 and a function update unit 144. The reward calculation unit 143 increases the reward r when at least one of the number of operations, the number of screen transitions, and the operation time decreases, and decreases the reward r when at least one of the number of operations, the number of screen transitions, and the operation time increases. Therefore, the trained model can infer the screen display data for displaying a new operation screen that reduces at least one of the number of operations, the number of screen transitions, and the operation time.

[0071] In addition, the new screen display data generated by the numerical control device 1 includes at least one of the design of the new operation screen, the information output timing, the information output method, and the operation guidance, which is information for assisting the user's operation.

[0072] Embodiment 2.

[0073] In Embodiment 2, the information used as input by the learning device 14 and the inference device 16 is different from that in Embodiment 1. Regarding the basic structure of the numerical control device 1, it is the same as that in Figure 1 Embodiment 1 shown, so the reference numerals Figure 1 are used, and mainly the parts different from Embodiment 1 will be described.

[0074] The operation log acquisition unit 12 acquires time-series data such as the coordinate positions input by the user using a pointing device such as a touch panel, the key codes of keyboard operations, the identification information of the operation screen of the operation object, and the hierarchical structure of the operation screen as operation log information. By using the operation log information as described above, the numerical control device 1 can generate a trained model capable of inferring the screen display data for displaying a new operation screen designed to reduce at least one of the scrolling operation, the zoom-in operation, and the zoom-out operation. For example, the reward calculation unit 143 of the model generation unit 142 increases the reward r when the number of operations of dynamic operations such as scrolling operations, zoom-in operations, and zoom-out operations decreases, and decreases the reward r when the number of operations of dynamic operations increases. Thus, the information required by the user can be displayed in a size and configuration that can reduce the number of dynamic operations.

[0075] As described above, according to Embodiment 2, the operation log information includes a history of time series of coordinate positions input by the user. In addition, the trained model infers the screen display data for displaying a new operation screen of a design that reduces the number of at least one of the scrolling operation, the zoom-in operation, and the zoom-out operation. Thus, not only can the number of screen transitions be reduced, but also the number of user operations within one operation screen can be reduced, further improving the user's work efficiency.

[0076] Embodiment 3.

[0077] In Embodiment 3, a method for receiving feedback on the new operation screen is described. The basic structure of the numerical control device 1 is the same as that of Embodiment 1, so the reference numerals Figure 1 shown are used, and mainly the parts different from Embodiment 1 will be described below.

[0078] In Embodiment 3, the new operation screen generated by the numerical control device 1 using the inference device 16 includes a feedback receiving unit that receives feedback from the user on the operation screen.

[0079] Figure 7 is a diagram showing an example of the new operation screen displayed by the numerical control device 1 according to Embodiment 3. Figure 7 The shown operation screen includes a first area for displaying alarm details, a second area for displaying the status information of the machine tool indicating the state of the machine tool, and a third area for displaying the countermeasure method corresponding to the alarm in progress. In addition, in Figure 7 the shown operation screen, as the feedback receiving unit 31 for receiving feedback from the user on the operation screen, it includes an input interface for inputting whether to approve the new operation screen. The feedback receiving unit 31 includes a message for assisting the operation of "Approve the new operation screen? (If you select reject, return to the original screen)", and an "Approve" button and a "Reject" button. For example, after the user observes the displayed new operation screen and believes that it contains the required information, by operating the "Approve" button, the new operation screen can be approved. For example, when the new operation screen is approved, the output unit 17 ends the display of the feedback receiving unit 31 and can display the operation screen including the first area, the second area, and the third area. In addition, when the user believes that the displayed new operation screen does not contain the required information, or believes that the layout of the new operation screen makes it difficult to see the required information, the user operates the "Reject" button, thereby being able to reject the new operation screen. When the new operation screen is rejected, the output unit 17 can, for example, end Figure 7The display of the new operation screen shown returns to the display of the existing screen.

[0080] In addition, although not shown, the feedback reception unit 31 of the new operation screen can include an interface for receiving input of approval or rejection and for receiving comments from the user. Further, in the above, feedback on the new operation screen is received through the two options of "approval" or "rejection", but changes can also be received for a part of the new operation screen according to the user's operation. For example, it is possible to replace only the content of the third region for displaying the method of responding to an alarm among the operation screens shown, or to change the display size of each region of the displayed screen according to the user's operation, or to add changes to a part of the screen display data of the new operation screen. The content received by the feedback reception unit 31 is used as an input to the learning device 14. Figure 7 As described above, according to Embodiment 3, the new operation screen includes the feedback reception unit 31 for receiving feedback from the user on the new operation screen. Further, the feedback reception unit 31 can receive an operation to return the operation screen to be displayed from the new operation screen to the existing operation screen as feedback. Thus, when the new operation screen generated by the numerical control device 1 is different from the operation screen required by the user, the intention of the user can be fed back, and thus the operation screen required by the user can be further approximated.

[0081] As described above, according to Embodiment 3, the new operation screen includes the feedback reception unit 31 for receiving feedback from the user on the new operation screen. Further, the feedback reception unit 31 can receive an operation to return the operation screen to be displayed from the new operation screen to the existing operation screen as feedback. Thus, when the new operation screen generated by the numerical control device 1 is different from the operation screen required by the user, the intention of the user can be fed back, and thus the operation screen required by the user can be further approximated.

[0082] Embodiment 4.

[0083] Figure 8 FIG. is a diagram showing the functional configuration of the numerical control device 1A according to Embodiment 4. The numerical control device 1A has a control unit 11, an operation log acquisition unit 12, a data acquisition unit 13, a learning device 14, a trained model storage unit 15, an inference device 16, an output unit 17, and an operator information storage unit 18. The numerical control device 1A has, in addition to the structure of the numerical control device 1 according to Embodiment 1, the operator information storage unit 18. Hereinafter, mainly the parts different from the numerical control device 1 according to Embodiment 1 will be described.

[0084] The operator information storage unit 18 stores operator information for identifying the user who operates the numerical control device 1A. The operator information storage unit 18 can store the operator information input using the input device 5 and output the operator information to the data acquisition unit 13. The operator information is used as input data in each of the learning device 14 and the inference device 16. Thus, the trained model can infer the screen display data based on the operation log information, the action state information, and the operator information, and can provide an optimized operation screen for each user.

[0085] As described above, according to Embodiment 4, the numerical control device 1A further includes an operator information holding unit 18 that holds operator information for identifying the user of the numerical control device 1A. The learning device 14 and the inference device 16 each use the operator information as input data. Thus, the trained model infers the screen display data based on the operator information in addition to the operation log information and the operation state information. Therefore, an operation screen matching the user can be provided.

[0086] Embodiment 5.

[0087] In Embodiment 5, an example in which a trained model is shared among a plurality of numerical control devices 1 will be described. In Embodiments 1 to 4, the numerical control device 1 includes a learning device 14 and an inference device 16, and uses the trained model generated inside the numerical control devices 1 and 1A. The inference device 16 infers the screen display data and performs a closed process within a single machine. In contrast, in Embodiment 5, a method of sharing the trained model among a plurality of numerical control devices 1 will be described.

[0088] For example, the learning device 14 and the trained model storage unit 15 are provided on a cloud server, whereby a plurality of numerical control devices 1 can share the trained model. Figure 9 FIG. is an example of a method of sharing the trained model in Embodiment 5. Here, by providing the learning device 14 on a cloud server or the like, a plurality of numerical control devices 1-1 to 1-N can communicate with the learning device 14 via a communication path. For example, each of the numerical control devices 1-1 to 1-N is a structure in which the structure of the learning device 14 and the trained model storage unit 15 of the numerical control device 1 shown in Figure 1 is omitted. In this case, the learning data is transmitted from the Figure 1 shown data acquisition unit 13 to the learning device 14, and the trained model is transmitted from the trained model storage unit 15 to the inference device 16 via a communication path such as the Internet.

[0089] In addition, the structure described here is an example. In Embodiment 5, the method of sharing the trained model among a plurality of numerical control devices 1 is not limited to the above example. For example, the learning device 14 and the trained model storage unit 15 are not limited to the example of being provided on a cloud server, as long as they can communicate with each of the plurality of numerical control devices 1-1 to 1-N via a communication path. Figure 9 The shown learning device 14 can be built into the numerical control device 1. In addition, not only the example of sharing the trained model generated by one learning device 14, but also each of the numerical control devices 1-1 to 1-N may have a learning device 14 and a trained model storage unit 15, and be connected to each other to share the trained model updated by each of the numerical control devices 1-1 to 1-N.

[0090] As described above, according to Embodiment 5, the learning device 14 can generate a trained model based on the operation log information in the plurality of numerical control devices 1-1 to 1-N and the operation state information of the plurality of machine tools. The learning device 14 can be provided on a cloud server or built into the numerical control device 1.

[0091] Embodiment 6.

[0092] In Embodiment 6, there is a function that, when it is determined that the user is unskilled based on the operation log information, a warning is issued for incorrect operations or operations with a potential risk, or operation guidance is output based on the result of learning the operation log information of skilled users. In addition, the basic structure of the numerical control device 1 according to Embodiment 6 is the same as the structure of Embodiment 1 shown below. Therefore, the parts different from Embodiment 1 will be mainly described below. Figure 1 The figure is for explaining the functions of the numerical control device 1 according to Embodiment 6. The numerical control device 1 according to Embodiment 6 has a proficiency determination unit 19. For example, the proficiency determination unit 19 is included in the data acquisition unit 13. The proficiency determination unit 19 determines the proficiency of the user based on the operation log information and the operation state information, and outputs the determination results to the learning device 14 and the inference device 16 respectively. The determination results are used as input data in the learning device 14 and the inference device 16 respectively.

[0093] Figure 10 The figure is for explaining the functions of the numerical control device 1 according to Embodiment 6. The numerical control device 1 according to Embodiment 6 has a proficiency determination unit 19. For example, the proficiency determination unit 19 is included in the data acquisition unit 13. The proficiency determination unit 19 determines the proficiency of the user based on the operation log information and the operation state information, and outputs the determination results to the learning device 14 and the inference device 16 respectively. The determination results are used as input data in the learning device 14 and the inference device 16 respectively.

[0094] The determination result output by the proficiency determination unit 19 only needs to be information that can indicate whether the target user is an unskilled user. For example, the determination result can be information including a flag indicating whether the target user is an unskilled user or a skilled user, or information including a numerical value indicating the proficiency of the target user. The proficiency determination unit 19 can, for example, identify the case of an unskilled user based on the operation log information and the operation state information, according to the time until the desired information is obtained, the time taken for the operation, the number of operation errors, the number of redundant operations, etc.

[0095] The inference device 16 can output screen display data including information corresponding to the determination result obtained by the proficiency determination unit 19. For example, the inference device 16 can output screen display data including warning information and operation guidance, which is information for assisting the operation, for unskilled users.

[0096] As described above, according to Embodiment 6, the numerical control device 1 further includes a proficiency determination unit 19 that determines the proficiency of the user during operation based on the operation log information and the operation status information. In this case, the trained model can infer the screen display data based on the operation log information, the operation status information, and the determination result of proficiency, and the screen display data is used to display a new operation screen including the information required by the user extracted from a plurality of existing operation screens and the information corresponding to the determination result.

[0097] Embodiment 7.

[0098] In Embodiment 7, an example of the screen layout of the new operation screen will be described. Figure 11 FIG. is an example of a new operation screen displayed by the numerical control device 1A according to Embodiment 7. In addition, the basic structure of the numerical control device 1A according to Embodiment 7 is the same as that of Figure 8 Embodiment 4 shown, so the reference numerals shown in Figure 8 are used, and the parts different from Embodiment 4 will be mainly described below.

[0099] The numerical control device 1A, as shown in Figure 11 , can display a new operation screen including: a common display area 32 that displays common information regardless of the user; and a user-classified display area 33 that displays information matching the user during operation. The new screen display data inferred by the trained model can include information specifying the information displayed in the common display area 32 and the information displayed in the user-classified display area 33. For example, as an example of the information displayed in the common display area 32, the operation mode, processing time, time, etc. indicating the state of the numerical control device 1A are cited. In addition, in the user-classified display area 33, when the user during operation is a skilled user, options for advanced menu selections can be displayed, and when the user during operation is an unskilled user, the menu can be simplified and operation guidance can be displayed. In addition, the numerical control device 1A can also accept an operation of specifying an area in the operation screen where the desired display layout is to be maintained. Thereby, unexpected layout changes by the user can be prevented.

[0100] As described above, according to Embodiment 7, the new operation screen displayed by the numerical control device 1A includes: a common display area 32, which is a common display area independent of the user; and a user-classified display area 33, which is a display area for different contents for each user. By having the common display area 32, information required regardless of the user is displayed in the common display area 32. By creating parts with no changes in the display, the load on the user to identify information within the operation screen can be reduced.

[0101] Embodiment 8.

[0102] In Embodiment 8, the operation log information for the custom screen, which is the operation screen created by the user, can be used as input data. The basic structure of the numerical control device 1 according to Embodiment 8 is the same as that of Figure 1 Embodiment 1 shown, so the reference numerals shown in Figure 1 are used. Hereinafter, mainly the parts different from Embodiment 1 will be described.

[0103] The operation log acquisition unit 12 can acquire the operation log information for the custom screen. By learning by including the custom screens with high usage frequency designed by the user in the user screen, operation assistance can be performed through a wider range of use cases.

[0104] Embodiment 9.

[0105] In Embodiment 9, the numerical control device 1 cannot be used in the standard state, but in the case of having an optional function that can be enabled by purchase, the function of proposing the optional function will be described. The basic structure of the numerical control device 1 according to Embodiment 9 is the same as that of Figure 1 Embodiment 1 shown, so the reference numerals shown in Figure 1 are used. Hereinafter, mainly the parts different from Embodiment 1 will be described.

[0106] The trained model generated by the learning device 14 can infer the screen display data for displaying a new operation screen when it is determined that the proposal of the optional function is effective based on the operation log information. The new operation screen includes information on the purchase of the proposed optional function. Here, the new operation screen can include, for example, information indicating the effects obtained by using the optional function. For example, when the user of the numerical control device 1 purchases an optional function and it is detected based on the operation log information that effects such as a reduction in the number of operations, a reduction in the number of screen transitions, and a reduction in the operation time are obtained, it can be determined that the proposal of the optional function is effective. Therefore, the operation log acquisition unit 12 uses the information related to the existing operation screen including the information on the function that has not been activated as input data. Moreover, when the screen display data output by the trained model includes information on a function that the user has not purchased, it can output the screen display data including the information on the purchase of the proposed function.

[0107] As described above, according to Embodiment 9, the trained model can infer the screen display data of the operation screen including the information on the purchase of the optional function of the numerical control device 1. The information on the purchase of the proposed optional function can include the information indicating the effects obtained by using the optional function. Thereby, the work efficiency of the user can be further improved.

[0108] Embodiment 10.

[0109] In Embodiment 10, it is possible to display an operation screen corresponding to the number of display devices 4 connected to the numerical control device 1. For example, when the numerical control device 1 can use a plurality of display devices 4 including a main monitor and a sub-monitor, the display using the plurality of display devices 4 is performed according to the situation. In addition, the basic structure of the numerical control device 1 according to Embodiment 10 is the same as that of Embodiment 1 shown in Figure 1 Therefore, the reference numerals shown in Figure 1 are used, and the parts different from Embodiment 1 will be mainly described below.

[0110] The data acquisition unit 13 can acquire information indicating the number of display devices 4 connected to the numerical control device 1 as input data. Here, when the display device 4 is built in the numerical control device 1, the built-in display device 4 is also included in the "display device 4 connected to the numerical control device 1". In addition, the "display device 4 connected to the numerical control device 1" can be a display device 4 dedicated to the numerical control device 1, or a mobile device such as a tablet terminal or a smartphone. The trained model infers the screen display data based on the operation log information, the operation state information, and also based on the number of display devices 4.

[0111] As described above, according to Embodiment 10, the trained model infers the screen display data corresponding to the number of display devices 4 connected to the numerical control device 1. Thereby, the display devices 4 available to the numerical control device 1 can be effectively used, and the work efficiency of the user can be further improved.

[0112] Embodiment 11.

[0113] In Embodiment 11, an example of effectively using external data obtained from the Internet or the like will be described. Figure 12 FIG. is a diagram showing the functional configuration of the numerical control device 1B according to Embodiment 11. The numerical control device 1B includes a control unit 11, an operation log acquisition unit 12, a data acquisition unit 13B, a learning device 14, a trained model storage unit 15, an inference device 16, and an output unit 17.

[0114] The numerical control device 1B is different from the numerical control device 1 according to Embodiment 1 in terms of the data acquired by the data acquisition unit 13B. Hereinafter, the parts different from Embodiment 1 will be mainly described. The data acquisition unit 13B further acquires power information, which is information indicating power supply to the numerical control device 1B, the drive unit 2 of the machine tool, and the peripheral equipment 3, etc., and power consumption in the machine tool. The power information acquired by the data acquisition unit 13B includes, for example, the power consumption of the machine tool acquired from the control unit 11, and information related to power supply included in external data acquired from outside the numerical control device 1B such as the Internet. The external data includes, for example, information indicating electricity charges, availability of renewable energy, weather, etc. These power information are used as input data in the learning device 14 and the inference device 16. Thereby, the trained model can infer the screen display data corresponding to the power information. In the learning device 14, the reward r is calculated based on the power information, and a trained model capable of inferring the screen display data that can reduce power consumption and electricity charges is generated.

[0115] Specifically, the new screen display data can display a new operation screen, which includes at least one piece of environment-related information such as electricity charges, greenhouse gas emissions, renewable energy utilization fees, proposal of machining schedules to avoid peak power usage periods, and messages for users to achieve energy conservation, in addition to the machining time.

[0116] As described above, according to Embodiment 11, the trained model can infer the screen display data corresponding to the power information representing the power supply and the power consumption of the machine tool. Thus, since each factor is complexly related, for an environmental target where it is difficult to make an intuitive action selection, the appropriate action selection can be assisted through the operation screen. For example, when only shortening the processing time is prioritized, the electricity cost increases during the peak period of power usage, or solar power generation, which is a renewable energy source, cannot be utilized due to bad weather. Sometimes, this may have an adverse impact on the achievement of the environmental target for the user. In contrast, by inferring the screen display data corresponding to the power information, the achievement of the environmental target can be assisted.

[0117] Next, the hardware structures of the numerical control devices 1, 1A, and 1B according to Embodiments 1 to 11 will be described. The functions of each part of the numerical control devices 1, 1A, and 1B are implemented using a processing circuit. The processing circuit can be implemented by dedicated hardware or a control circuit using a CPU (Central Processing Unit).

[0118] In the case where the above processing circuit is implemented by dedicated hardware, they are implemented using Figure 13 the processing circuit 90 shown. Figure 13 It is a diagram showing dedicated hardware for implementing the functions of the numerical control devices 1, 1A, and 1B according to Embodiments 1 to 11. The processing circuit 90 is a single circuit, a composite circuit, a programmed processor, a parallel-programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or a combination thereof.

[0119] In the case where the above processing circuit is implemented by a control circuit using a CPU, the control circuit is, for example, Figure 14 the control circuit 91 having the structure shown. Figure 14 It is a diagram showing the structure of the control circuit 91 for implementing the functions of the numerical control devices 1, 1A, and 1B according to Embodiments 1 to 11. As Figure 14As shown, the control circuit 91 includes a processor 92 and a memory 93. The processor 92 is a CPU, and is also referred to as a processing device, an arithmetic device, a microprocessor, a microcomputer, a DSP (Digital Signal Processor), etc. The memory 93 is, for example, a non-volatile or volatile semiconductor memory such as a RAM (Random Access Memory), a ROM (Read Only Memory), a flash memory, an EPROM (Erasable Programmable ROM), an EEPROM (registered trademark) (Electrically EPROM), a magnetic disk, a floppy disk, an optical disk, a compact disk, a minidisk, a DVD (Digital Versatile Disk), etc.

[0120] When the above processing circuit is implemented using the control circuit 91, it is implemented by the processor 92 reading and executing a program corresponding to the processing of each structural element stored in the memory 93. In addition, the memory 93 is also used as a temporary memory in each process executed by the processor 92. The program can be provided in a state stored in a storage medium, or can be provided via a communication path such as the Internet.

[0121] The structures shown in the above embodiments represent an example, and can also be combined with other known technologies, and the embodiments can also be combined with each other. Without departing from the gist, a part of the structure can also be omitted or changed. For example, in the above embodiments, the technology described for the parts different from those in Embodiment 1 can be combined with the functions of the numerical control device 1A related to Embodiment 4, or can be combined with the functions of the numerical control device 1B related to Embodiment 11.

[0122] Explanation of reference numerals

[0123] 1, 1A, 1B, 1-1 to 1-N numerical control devices, 2 drive unit, 3 peripheral device, 4 display device, 5 input device, 11 control unit, 12 operation log acquisition unit, 13, 13B data acquisition unit, 14 learning device, 15 trained model storage unit, 16 inference device, 17 output unit, 18 operator information storage unit, 19 proficiency determination unit, 31 feedback reception unit, 32 common display area, 33 user-classified display area, 90 processing circuit, 91 control circuit, 92 processor, 93 memory, 141 learning data acquisition unit, 142 model generation unit, 143 reward calculation unit, 144 function update unit, 161 inference data acquisition unit, 162 inference unit.

Claims

1. A numerical control device, characterized in that, comprising: a learning data acquisition unit that acquires learning data, the learning data including operation log information representing operations of a user on an operation screen of a numerical control device for controlling a machine tool and information of the operation screen being an operation object, and operation state information representing a state of the machine tool when the operations shown in the operation log information are performed; and a model generation unit that uses the learning data to generate a trained model for inferring screen display data for displaying a new operation screen including information extracted from a plurality of existing operation screens, based on the operation log information and the operation state information.

2. A numerical control device, characterized in that, comprising: an inference data acquisition unit that acquires inference data, the inference data including operation log information representing operations of a user on an operation screen of a numerical control device for controlling a machine tool and information of the operation screen being an operation object, and operation state information representing a state of the machine tool when the operations shown in the operation log information are performed; an inference unit that uses the trained model for inferring screen display data for displaying a new operation screen including information extracted from a plurality of existing operation screens, based on the operation log information and the operation state information, to infer screen display data for displaying a new operation screen based on the operation log information and the operation state information acquired by the inference data acquisition unit; and an output unit that outputs the new operation screen to a display device based on the screen display data.

3. The numerical control device according to claim 1 or 2, wherein the new operation screen includes information for reducing the operation burden of the user.

4. The numerical control device according to claim 1 or 2, wherein the new operation screen includes information that has a favorable influence on the operation result of the user.

5. The numerical control device according to claim 1 or 2, wherein the new operation screen includes a plurality of information separately included in a plurality of existing operation screens.

6. The numerical control device according to any one of claims 1 to 5, wherein the operation log information represents the operation history of the user across a plurality of existing operation screens.

7. The numerical control device according to any one of claims 1 to 6, wherein a plurality of existing operation screens form a hierarchical structure, and the operation log information represents the operation history of the user across a plurality of existing operation screens that are a hierarchical structure.

8. The numerical control device according to any one of claims 1 to 7, wherein the trained model infers screen display data for displaying the new operation screen that reduces at least one of the number of operations, the number of screen transitions, and the operation time.

9. The numerical control device according to any one of claims 1 to 8, wherein The screen display data includes at least one piece of information among the design of the new operation screen, the information output timing, the information output method, and the operation guidance information for assisting the operation of the user.

10. The numerical control device according to any one of claims 1 to 9, characterized in that The operation log information includes the chronological resume of the coordinate positions input by the user.

11. The numerical control device according to any one of claims 1 to 10, characterized in that The trained model infers the screen display data for the new operation screen that is designed to reduce the number of operations of at least one of the scroll operation, the zoom-in operation, and the zoom-out operation.

12. The numerical control device according to any one of claims 1 to 11, characterized in that The new operation screen includes a feedback receiving unit for the user to receive feedback on the new operation screen.

13. The numerical control device according to claim 12, characterized in that The feedback receiving unit receives an operation to return the operation screen to be displayed from the new operation screen to the existing operation screen as the feedback.

14. The numerical control device according to any one of claims 1 to 13, characterized in that It further has an operator information holding unit that holds operator information for identifying the user, The trained model infers the screen display data based on the operation log information, the action state information, and the operator information.

15. The numerical control device according to claim 1, characterized in that The model generation unit generates the trained model based on the operation log information in multiple numerical control devices and the action state information of multiple machine tools.

16. The numerical control device according to any one of claims 1 to 15, characterized in that It further has a proficiency determination unit that determines the proficiency of the user based on the operation log information and the action state information, The trained model infers the screen display data for displaying a new operation screen based on the operation log information, the action state information, and the determination result of the proficiency. The new operation screen includes the information required by the user extracted from the existing multiple operation screens and the information corresponding to the determination result.

17. The numerical control device according to any one of claims 1 to 16, characterized in that The new operation screen includes a common display area that does not depend on the user and a display area with different contents for each user.

18. The numerical control device according to any one of claims 1 to 17, characterized in that The operation log information includes the operation log information for a custom screen created by the user.

19. The numerical control device according to claim 1 or 2, characterized in that The trained model can infer the screen display data of the operation screen including information for purchasing optional functions of the numerical control device.

20. The numerical control device according to claim 19, wherein the information for purchasing the optional function includes information indicating the effects obtained by using the optional function.

21. The numerical control device according to claim 1 or 2, wherein the trained model infers the screen display data corresponding to the number of display devices connected to the numerical control device.

22. The numerical control device according to claim 1 or 2, wherein the trained model infers the screen display data corresponding to the power information indicating the power supply and the power consumption of the machine tool.

23. A learning device, characterized in that, comprising: a learning data acquisition unit that acquires learning data, the learning data including operation log information indicating operations on an operation screen of a numerical control device for controlling a machine tool and information on the operation screen of the operation object, and action state information indicating the state of the machine tool when the operations shown in the operation log information are performed; and a model generation unit that uses the learning data and generates a trained model for inferring screen display data based on the operation log information and the action state information, the screen display data being for displaying a new operation screen including information extracted from a plurality of existing operation screens.

24. An inference device, characterized in that, comprising: an inference data acquisition unit that acquires inference data, the inference data including operation log information indicating operations on an operation screen of a numerical control device for controlling a machine tool and information on the operation screen of the operation object, and action state information indicating the state of the machine tool when the operations shown in the operation log information are performed; and an inference unit that uses a trained model for inferring screen display data for displaying a new operation screen including information extracted from a plurality of existing operation screens based on the operation log information and the action state information, and infers the screen display data for displaying the new operation screen based on the operation log information and the action state information obtained through the inference data acquisition unit.

25. A method for displaying an operation screen of a numerical control device that controls a machine tool, the method for displaying an operation screen of the numerical control device being characterized by including the following steps: acquiring inference data, the inference data including operation log information indicating operations on an operation screen of the numerical control device and information on the operation screen of the operation object, and action state information indicating the state of the machine tool when the operations shown in the operation log information are performed; using a trained model for inferring screen display data for displaying a new operation screen including information extracted from a plurality of existing operation screens based on the operation log information and the action state information, and inferring the screen display data for displaying the new operation screen based on the operation log information and the action state information obtained in the step of acquiring the inference data; and Based on the screen display data, output the new operation screen to the display device.

Images