Processing support device and processing support system

Abstract

To provide a processing support device configured to be capable of grasping how a processing result varies when a processing condition is changed.SOLUTION: When a requirement specification is input by operation of a user interface 20, a PU 12 of a process support device 10 searches for a process condition satisfying the requirement specification to a predetermined degree or more. The searching process is executed by the PU 12 using a trained model which takes the process condition as input and outputs the process result. The PU 12 displays the searched-for process condition and process result on a display device 22. When an instruction to change the process condition is given by operation of the user interface 20, the PU 12 displays the process result according to the change by the instruction on the display device 22.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a machining support device and a machining support system. [Background technology]

[0002] For example, the following Patent Document 1 describes a machining condition search device. This device outputs data on optimal machining conditions according to input required specifications. This device finds optimal machining conditions based on data on machining results output by a machining condition model section when various machining conditions are input to the model section. The machining condition model section is equipped with a trained model through machine learning. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2008-36812 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, in the case of the above-mentioned device, since the optimum machining conditions for the required specifications are simply output, it is not possible to grasp how the machining results will change when the machining conditions are changed. [Means for solving the problem]

[0005] Means for solving the above problems and their effects will be described below. 1. A method for producing a processing result output map comprising an execution device and a storage device, the storage device storing processing result output mapping data which is data for defining a processing result output mapping, the processing result output mapping being a mapping which takes values ​​of processing condition variables as input and outputs values ​​of processing result variables, the processing condition variables being variables which indicate processing conditions used by a processing device, the processing result variables being variables which indicate processing results used by the processing device, the execution device being configured to execute a request acquisition process, a processing result calculation process, a search process, a processing condition notification process, a processing condition change acceptance process, and a change result notification process, the request acquisition process being a process for acquiring required specifications for the processing result, the processing result calculation process being a process for calculating the processing result by calculating the processing result by converting the values ​​of the processing condition variables into the processing result values, a processing support device which is a process for determining the value of the processing result variable when input into a result output mapping, wherein the search process is a process for searching for a value among the values ​​of the processing result variable obtained by the processing result calculation process that satisfies a required specification or more to a specified extent, wherein the processing condition notification process is a process for notifying the processing conditions indicated by the search results of the search process and the corresponding processing result, wherein the processing condition change acceptance process is a process for accepting an instruction to change the processing conditions to the processing conditions notified by the processing condition notification process, and wherein the change result notification process is a process for notifying the processing result corresponding to the value of the processing result variable obtained as an input to the processing result calculation process based on the value of the processing condition variable in accordance with the instruction to change.

[0006] The processing support device searches for processing conditions that satisfy the required specifications to a predetermined degree or higher. The processing support device then notifies the user of the search results. This allows the user to understand the processing conditions that satisfy the required specifications to a predetermined degree or higher. The processing support device also notifies the user of the processing results in accordance with the user's instruction to change the searched processing conditions. This allows the user to understand how the processing results will change when the processing conditions are changed.

[0007] 2. The processing result has a plurality of items, the processing conditions have a plurality of items, and the execution device is configured to execute a specified item reception process, a contribution degree calculation process, and a contribution degree notification process, wherein the specified item reception process is a process of accepting an item specified by a user among the plurality of items related to the processing result, the contribution degree calculation process is a process of calculating the contribution degree of an item of the processing conditions to the item of the processing result accepted by the specified item reception process, and the contribution degree notification process is a process of notifying information indicating the contribution degree of the item of the processing conditions to the item of the processing result accepted by the specified item reception process.

[0008] The processing support device notifies the user of the contribution of the processing condition items to the item of the processing result designated by the user, thereby making it possible to suppress the user from issuing an instruction to change an item of the processing condition that has a small contribution.

[0009] 3. A processing support device as described in 2 above, wherein the contribution calculation process includes a change direction calculation process that determines how to change the item of the processing result accepted by the specified item acceptance process to make it more conform to the required specifications, and the contribution notification process includes a process that notifies information indicating the contribution of the item of the processing result together with information indicating how to change it to make it more conform to the required specifications.

[0010] The processing support device notifies the user of the contribution of each item to the processing result and also notifies the user of information on how to change the item to better conform to the required specifications, thereby preventing the user from giving instructions to change the processing conditions in a direction that deviates from the required specifications.

[0011] 4. The machining support device according to any one of 1 to 3 above, wherein the machining result variables include a variable indicating machining accuracy and a variable indicating a load applied to the machining. Even if the machining accuracy meets the required specifications, if the load on the machining is excessively large, it will lead to increased costs and machining times that ignore delivery dates, and so on, making the machining conditions less practical. Therefore, the machining support device includes a variable indicating the load on the machining in addition to a variable indicating the machining accuracy in the machining result variable. This makes it possible to prevent machining conditions with excessive loads from being searched for.

[0012] 5. The processing device is a grinding device, and the processing condition variables include a variable indicating the cutting depth by the grinding process, a variable indicating the grinding wheel speed of the grinding process, and a variable indicating the grinding fluid flow rate by the grinding process. This is the processing support device described in any one of 1 to 4 above.

[0013] 6. The processing support device described in any one of 1 to 5 above, wherein the storage device stores combination data, the combination data being data indicating a combination of the processing result and the processing conditions, and the search process includes an acquisition process for acquiring the processing conditions corresponding to the processing result that is closest to the required specifications among the processing results included in the combination data, and a change process for changing the processing conditions to be searched for among the processing conditions acquired by the acquisition process.

[0014] When there are many items indicating the processing results and when there are many items indicating the processing conditions, the processing load for searching for processing conditions that satisfy the required specifications is likely to be large. Therefore, the processing support device acquires processing conditions that are candidates to be searched for using combination data. Since the combination data is data indicating a combination of processing results and processing conditions, it can be used as data that identifies processing that has actually been performed in the past. This makes it possible to quickly acquire highly reliable processing conditions as candidates to be searched for.

[0015] 7. The processing support device described in 6 above, wherein the processing conditions have a plurality of items, and the change process includes a process for increasing the probability of changing, for one item of the processing conditions acquired by the acquisition process, a condition close to the acquired condition for the same item, compared to the probability of changing to a condition far from the acquired condition.

[0016] When the combination data reflects the actual machining that satisfies the required specifications, generally, conditions that are closer to the machining conditions acquired by the acquisition process are more likely to conform to the required specifications than conditions that are farther from the machining conditions. However, when the machining conditions have multiple items, the influence of each item on the machining result may vary depending on other items of the machining conditions. In that case, the machining conditions that conform to the required specifications are not necessarily close to the machining conditions acquired by the acquisition process. Therefore, the above-mentioned machining support device includes conditions that are far from the machining conditions acquired by the acquisition process as candidates to be searched for with a low probability. This increases the possibility of searching for machining conditions that satisfy the required specifications, even when the influence of one item of the machining conditions on the machining result varies greatly depending on another item of the machining conditions.

[0017] 8. A processing support device as described in any one of 1 to 7 above, wherein the processing conditions have a plurality of items, the execution device is configured to execute a processing condition designation acceptance process, the processing condition designation acceptance process is a process of accepting a user's designation of some of the items related to the processing conditions, and the search process is a process of searching for the processing conditions under constraints that satisfy the specified items.

[0018] A user may not be able to realize all machining conditions. For example, a user may want to know other machining conditions on the assumption that a specific machining device is used. In addition, for example, the machining tools available to the user may be limited. In this case, when a user specifies some items of the machining conditions, the machining support device searches for machining conditions under constraints that satisfy the specified items. In this way, the machining support device can search for machining conditions that satisfy the required specifications from among the machining conditions that the user can actually achieve, and notify the user of the machining conditions.

[0019] 9. A processing support device as described in any one of 1 to 8 above, wherein the processing conditions have a plurality of items, relationship definition data is stored in the storage device, the relationship definition data is data that defines the relationship between the plurality of items of the processing conditions, and the search process is a process that searches for the processing conditions that satisfy the relationship defined by the relationship definition data.

[0020] In the field of machining, there are engineering theories. Some of these theories stipulate the relationships between multiple items of machining conditions. In addition, experts who have been engaged in machining for many years have knowledge about appropriate relationships between multiple items of machining conditions. It is unlikely that machining conditions that deviate from these relationships will satisfy the required specifications. Therefore, the machining support device searches for machining conditions that satisfy the relationships stipulated by the relationship stipulation data. In this way, by reflecting the relationships based on at least one of the engineering theories and the knowledge of experts in the relationship stipulation data, appropriate search candidates can be efficiently determined.

[0021] 10. The execution device is configured to execute a result acquisition process, a re-learning process, and an update process, the result acquisition process being a process of acquiring the processing result by the processing device and the corresponding processing conditions, the re-learning process being a process of re-learning the processing result output mapping using data acquired by the result acquisition process as training data, and the update process being a process of updating the processing result output mapping data stored in the storage device to the re-learned data, the processing support device described in any one of 1 to 9 above.

[0022] The processing support device updates the processing result output mapping data by executing a re-learning process. This allows the user to actually use the processing device to perform processing and update the processing result output mapping data using the processing results, etc. This makes it possible to improve the accuracy of notification of processing results that meet the required specifications.

[0023] 11. The processing device is a grinding device, and the processing result variables include at least one of five variables: a knock mark variable, a feed mark variable, a burn variable, a scratch variable, and a corner sagging variable, and the knock mark variable is a variable indicating a pattern that appears on the surface of the workpiece in a direction perpendicular to the processing direction, the feed mark variable is a variable indicating a pattern that appears on the surface of the workpiece in the processing direction, the burn variable is a variable indicating the degree of discoloration of the surface of the workpiece due to processing heat, the scratch variable is a variable indicating the degree of scratches on the surface of the workpiece, and the corner sagging variable is a variable indicating the degree to which the peripheral portion of the workpiece has been excessively removed. This is the processing support device described in any one of 1 to 10 above.

[0024] The above variables indicate items that are usually qualitatively evaluated by an operator in a grinding process, and therefore the machining support device can include items that are usually qualitatively evaluated by an operator in a grinding process in the required specifications.

[0025] 12. A processing support system comprising the execution device and the memory device in the processing support device described in any one of 1 to 11 above, wherein the execution device comprises a first execution device and a second execution device, the first execution device is configured to execute at least the request acquisition process and the processing condition change acceptance process, and the second execution device is configured to execute at least the processing result calculation process.

[0026] In the above processing support system, by having the second executing unit execute the processing result calculation process, it is possible to reduce the calculation load on the first executing unit, compared to a case in which the processing result calculation process is executed by the first executing unit.

[0027] 13. A processing support system comprising the execution device and the memory device in the processing support device described in 10 above, wherein the execution device comprises a first execution device and a second execution device, the first execution device is configured to execute at least the result acquisition process, and the second execution device is configured to execute at least the re-learning process.

[0028] In the above processing support system, the second execution unit executes the re-learning process. Therefore, for example, by providing the second execution unit in a device other than the user terminal, it becomes easy to execute the re-learning process using data acquired by the result acquisition process by each of the multiple user terminals.

[0029] 14. A processing support device comprising the first execution device in the processing support system described in 12 above. 15. A processing support device comprising the first execution device in the processing support system described in 13 above.

[0030] In the processing support device according to any one of 1 to 11 above and the processing support system according to any one of 12 and 13 above, the processing result may have a plurality of items, and the search process may be a process of searching for a plurality of Pareto optimal solutions.

[0031] In addition, in the processing support device described in 1 to 11 above and the processing support system described in 12 or 13 above, the value of the processing result variable may include a variable indicating a probability distribution of the processing result, the search process may be a process of considering the probability distribution of the processing result to satisfy the required specification to a predetermined degree or more when the probability that the probability of the processing result not satisfying the required specification is less than a threshold value, the processing condition notification process may include a process of notifying the probability distribution of the processing result corresponding to the processing conditions searched for by the search process, and the change result notification process may include a process of notifying the probability distribution of the processing result according to the value of the processing result variable obtained as an input of the processing result calculation process using the value of the processing condition variable in accordance with the change instruction. [Brief description of the drawings]

[0032] [Figure 1] 1 is a block diagram showing a processing support device according to a first embodiment. [Diagram 2] FIG. 13 is a diagram showing a processing result output map according to the embodiment. [Diagram 3] 4 is a flowchart showing a procedure of a process executed by the machining support device according to the embodiment. [Figure 4] 4 is a flowchart showing a procedure of a process executed by the machining support device according to the embodiment. [Diagram 5] 13A and 13B are diagrams showing examples of displays for supporting a grinding process according to the embodiment; [Figure 6] 13A and 13B are diagrams showing examples of displays for supporting a grinding process according to the embodiment; [Figure 7] 13A and 13B are diagrams showing examples of displays for supporting a grinding process according to the embodiment; [Figure 8] 13 is a flowchart showing a procedure of a re-learning process of the processing result output mapping data according to the embodiment. [Figure 9] FIG. 11 is a block diagram showing a configuration of a processing support system according to a second embodiment. [Figure 10] 13 is a flowchart showing a procedure of a re-learning process of the processing result output mapping data according to the embodiment. [Figure 11]FIG. 13 is a block diagram showing a configuration of a processing support system according to a third embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0033] <First embodiment> Hereinafter, the first embodiment will be described with reference to the drawings. "Prerequisite configuration" FIG. 1 shows a machining support device according to this embodiment.

[0034] 1 receives a user's intention in response to an input operation via a user interface 20. The processing support device 10 also displays a calculation result on a display device 22.

[0035] The processing support device 10 includes a PU 12 and a storage device 14. The PU 12 is a software processing device such as a CPU, a GPU, or a TPU. The storage device 14 may be an electrically rewritable non-volatile memory, a disk medium, or other storage medium.

[0036] The storage device 14 stores a processing support program 14a, processing result output image data 14b, combination data 14c, relationship definition data 14d, and a re-learning program 14e.

[0037] The machining support device 10 executes a process for supporting the grinding process by having the PU 12 execute the machining support program 14a. This process includes a process for displaying machining conditions related to grinding that meet the user's required specifications. This process uses a machining result output map, which is a map defined by the machining result output map data 14b.

[0038] "About trained models" Figure 2 shows the output map of the processing results. As shown in FIG. 2, the machining result output map is a map that inputs machining conditions and outputs machining results. FIG. 2 shows examples of machining conditions, such as the type of grinding device, the type of machining tool, the left-right feed rate, the amount of cut, the cut rate, the machining allowance, the grinding wheel rotation speed, and the grinding fluid flow rate. Here, the type of machining tool is, for example, the type of grinding wheel and the size of the grinding wheel. The cut rate is the relative displacement speed of the machining tool with respect to the workpiece. The left-right feed rate is the relative displacement speed of the machining tool with respect to the workpiece in a direction perpendicular to the cut rate. The machining allowance is the amount that is ultimately removed from the workpiece. The grinding fluid flow rate is the amount of grinding fluid supplied per unit time.

[0039] Figure 2 shows examples of the machining results, including geometric accuracy, removal amount, surface quality, surface roughness Ra, knock marks, feed marks, burns, scratches, corner sagging, and machining time. Geometric accuracy is, for example, flatness and roundness. Removal amount is the thickness of material removed by one cutting process. Surface quality indicates the roughness of the surface of the object after machining. In particular, surface quality indicates the difference between the maximum and minimum values ​​of the surface height. Surface roughness Ra is the arithmetic mean roughness.

[0040] A hammer mark is a pattern on the surface of a workpiece caused by the vibration of the grinding machine or the surface shape of the grindstone, and it mainly appears in a direction perpendicular to the processing direction. Here, the processing direction is the relative movement direction between the workpiece and the grindstone. The degree of hammer marks is usually judged visually by an expert grinder. However, here, the degree judged by the expert is quantified.

[0041] Feed marks are patterns that appear on the surface of the workpiece due to the motion accuracy of the grinding process and the rigidity of the grindstone, and they mainly appear in the feed direction of the material. Feed marks are usually judged by skilled grinders visually. However, here, the degree judged by the skilled grinders is quantified.

[0042] Burn marks indicate a state in which the surface of a workpiece has been altered by the effects of processing heat. The degree of burn marks is usually judged by an expert grinder by visually inspecting the surface for discoloration caused by heat. However, here the degree judged by the expert is quantified.

[0043] A scratch is a line of damage on the surface of a workpiece that occurs when abrasive grains fall off the grinding wheel or when debris from the grinding fluid gets mixed into the grinding area. Corner sagging indicates the degree to which the peripheral portion of the workpiece is excessively scraped away.

[0044] The output map of the machining object is a trained model. The results of grinding actually performed by an expert and the machining conditions at that time are used as training data for the output map of the machining object. In order to increase the number of training data, training data may be added by actually performing grinding under several machining conditions determined by experimental design to obtain machining results. Among the training data, the above-mentioned knock marks, feed marks, burns, and scratches are generated by an expert visually determining numerical values. In addition, the type of grinding device, the type of machining tool, etc. are defined by label variables.

[0045] The processing object output map is a regression model. In particular, the processing object output map is a map that outputs a probability distribution for each item of the processing result. One example of the probability distribution is a normal distribution. In this case, the processing result output map is, for example, a map that outputs variables indicating the mean and variance for each item of the processing result.

[0046] In this embodiment, as an example, the processing object output map is configured by a neural network. More specifically, the neural network is a fully connected forward propagation type network. Also, as an example, the neural network has 2 to 7 intermediate layers. Also, as an example, the number of nodes in each intermediate layer is 30 to 200. The number of nodes in the intermediate layers may be the same or different.

[0047] "Grinding support" Figures 3 and 4 show the procedure of the process executed by the machining support device 10. The process shown in Figures 3 and 4 is realized by the PU 12 repeatedly executing the machining support program 14a, for example, at a predetermined interval.

[0048] 3 and 4, the PU 12 first judges whether or not the required specifications for the grinding process have been input by operating the user interface 20 (S10). The required specifications indicate the requirements for the above-mentioned machining results. That is, the required specifications specify the acceptable range for the machining results, such as a surface quality of "x" or less.

[0049] When the PU 12 determines that the required specifications have been input (S10: YES), it temporarily stores the required specifications in the storage device 14 (S12). Next, the PU 12 determines whether or not some of the processing conditions have been specified by the user through the operation of the user interface 20 (S14). This is a process for grasping the processing conditions that the user can realize, for example, when the grinding devices that the user can use or the grinding tools that the user uses are limited.

[0050] When the PU 12 determines that some of the machining conditions have been designated by the user (S14: YES), the PU 12 temporarily stores the designated machining conditions in the storage device 14 (S16). When the PU 12 completes the process of S16 or makes a negative determination in the process of S14, the PU 12 calculates a first component and a second component obtained by dimensionally compressing a vector indicating the required specifications (S18).

[0051] Here, the process of converting the vector indicating the required specifications into the first and second components is defined using the above combination data 14c. The combination data 14c is data indicating a pair of the processing result in the grinding process actually performed by the skilled worker in the past and the processing conditions at that time. The combination data 14c may be training data for learning the processing result output mapping. In the process of S18, the PU 12 first extracts only the items corresponding to the required specifications from the processing results of the combination data 14c. That is, for example, when the processing time is not included in the required specifications, the PU 12 extracts items other than the processing time from the combination data 14c. Then, the PU 12 calculates the first and second principal components of the vector between the extracted processing result items and the processing result of the required specifications.

[0052] Next, the PU 12 searches for machining conditions that correspond to the machining results of the combination data 14c that are closest to the required specifications. The closeness between the machining results and the required specifications may be quantified by the norm between the points defined by the first and second principal components in the process of S18. The PU 12 then selects the closest machining conditions found as candidates for machining conditions that satisfy the required specifications (S20).

[0053] However, when the PU 12 executes the process of S16, in the process of S20, the PU 12 excludes, from among the sets of machining conditions and machining results indicated by the combination data 14c, those sets that do not match the conditions specified by the process of S16. In detail, from among the sets of machining conditions and machining results indicated by the combination data 14c, the PU 12 sets the machining conditions indicated by the sets other than the sets that do not match the conditions specified by the process of S16 as candidates for machining conditions that satisfy the required specifications.

[0054] Furthermore, the PU 12 excludes from candidates for processing conditions those pairs of processing conditions and processing results indicated by the combination data 14c that do not satisfy the relationship between the processing conditions defined by the relationship defining data 14d.

[0055] The relationship definition data 14d is data that defines the relationships between the items of the processing conditions defined by theories in engineering and the relationships between the items of the processing conditions obtained by the knowledge of an expert.

[0056] The relationship between processing condition items defined by engineering theory may be a relationship that has a certain degree of flexibility beyond the unambiguous relationship obtained from a theoretical formula. In other words, if the relationship y=2·x is theoretically derived between two processing condition items x and y, "2·x-δ≦y≦2·x+δ" may be included in the relationship definition data.

[0057] Similarly, the relationship based on the knowledge of a skilled person includes a relationship expressed by an inequality, etc. In addition to the above, for example, when the processing object is a cemented carbide, the grinding wheel as a processing tool may be diamond.

[0058] Next, the PU 12 generates N candidates for machining conditions by changing the machining conditions selected by the process of S20 (S22). Here, N is a number of more than one. Note that the search target (1), which is one of the candidates for machining conditions, is the machining condition selected by the process of S20. Also, the search targets (2) to (N) are generated by changing the machining conditions selected by the process of S20. In detail, the PU 12 changes the machining conditions for each of the multiple items of the machining conditions selected by the process of S20 under a condition that the probability of a change that reduces the difference with the value of the same item is made larger than the probability of a change that increases the difference with the value of the same item. Note that the condition that the probability of a change that reduces the difference is made larger than the probability of a change that increases the difference may be applied only to continuous variables among the machining conditions. In other words, this condition does not need to be applied to the grinding device, grinding tool, etc. Note that the above condition may also be applied to the items of the machining conditions defined by label variables, etc., by defining the closeness within the same item.

[0059] However, when PU12 executes the process of S16, in the process of S22, it prohibits changing to processing conditions that do not match the conditions specified by the process of S16. Thereby, when the process of S16 is executed, the search targets (1) to (N) satisfy the conditions stored by the process of S16. Also, PU12 prohibits changing to processing conditions that do not satisfy the relationship between the processing conditions specified by the relationship regulation data 14d. Thereby, the search targets (1) to (N) satisfy the relationship specified by the relationship regulation data 14d.

[0060] Next, PU12 assigns "1" to the label variable i that designates the search target (S24). Then, by inputting the search target (i) into the processing result output mapping, PU12 calculates the probability distribution for each item of the processing result (S26).

[0061] Next, PU12 determines whether the label variable i is N (S28). When PU12 determines that the label variable i is not N (S28: NO), it increments the label variable i (S30) and returns to the process of S26.

[0062] On the other hand, when PU12 determines that the label variable i is N (S28: YES), it determines the Pareto optimal solutions (1) to (M) among the N processing results corresponding to the search targets (1) to (N) (S32). That is, since there are multiple items in the processing result, instead of narrowing down to one optimal processing condition that satisfies the requirement specifications, PU12 determines M (<N) Pareto optimal solutions. However, M may be "1".

[0063] 4, the PU 12 calculates the contribution of each item of the processing result for each of the search targets (1) to (N) (S34). That is, for each of the search targets (1) to (N), the PU 12 calculates the degree of contribution of each item of the processing conditions to each item of the processing result. In detail, for each of the search targets (1) to (N), the PU 12 calculates the contribution of a target item to each item of the processing result by changing the value of the target item in a state where items other than the target item among the multiple items of the processing conditions are marginalized. For example, the PU 12 may calculate the contribution of the cutting amount to each item of the processing result by changing the value of the cutting amount in a state where items other than the cutting amount are marginalized for each of the search targets (1) to (N). Specifically, the PU 12 may calculate a SHAP value as the contribution.

[0064] The process of S34 includes a process of identifying how each item of the machining conditions should be changed to make the machining result more conform to the required specifications, i.e., for each item of the machining result, a process of identifying whether the cutting depth should be increased or decreased to make the machining result more conform to the required specifications.

[0065] Next, the PU 12 operates the display device 22 to display the Pareto optimal solutions (S36). FIG. 5 illustrates the process of S36. As shown in FIG. 5, the PU 12 displays on the display device 22 an optimal solution list 30 indicating the Pareto optimal solutions and an existence probability map 32 indicating the probability distribution of the machining result. The optimal solution list 30 is a list of items of the machining conditions (machining systems) corresponding to the Pareto optimal solutions. When there are multiple Pareto optimal solutions, the PU 12 generates an optimal solution list 30 for each of the Pareto optimal solutions and displays it on the display device 22. FIG. 5 illustrates an example in which there are three Pareto optimal solutions. In detail, the PU 12 displays the contents of the currently selected optimal solution list 30, while for the optimal solution list 30 that is not selected, only the outline of the list is displayed. When there are multiple Pareto optimal solutions, the PU 12 displays on the display device 22 an existence probability map 32 corresponding to the currently selected optimal solution. That is, the existence probability map 32 shown in FIG. 5 indicates the probability distribution of the machining result according to the machining conditions, such as the grinding device being "4" and the machining tool being "3".

[0066] The existence probability map 32 is an image including a probability distribution curve CP with the horizontal axis representing the value of each item and the vertical axis representing the probability for each item of the processing result, and a threshold value Lt representing the required specification. The threshold value Lt is a value indicated by the required specification for each item of the processing result. For example, when the required specification for the geometric accuracy of the processing result is "0.1" or less, the PU 12 sets the threshold value Lt of the geometric accuracy to "0.1". When the probability of satisfying the threshold value Lt is equal to or greater than a predetermined rate, the PU 12 determines that the required specification for the item is satisfied at a predetermined level or higher. That is, in this embodiment, the Pareto optimal solutions are all solutions in which the probability that each item of the processing result satisfies the required specification is equal to or greater than a predetermined rate.

[0067] 4, the PU 12 determines whether or not a specific item among the items of the processing results displayed on the existence probability map 32 has been selected by operating the user interface 20 (S38). When the PU 12 determines that a specific item has been selected (S38: YES), the PU 12 operates the display device 22 to display the calculation result for that item obtained by the process of S34 (S40). That is, the PU 12 displays, on the display device 22, the calculation result obtained by the process of S34 that corresponds to the currently selected Pareto optimal solution.

[0068] FIG. 6 illustrates the processes of S38 and S40. 6 illustrates a state in which "geometric accuracy" among the items of the machining result is selected by the cursor 40. As a result, the PU 12 causes the display device 22 to display a contribution graph 34 related to the currently selected Pareto optimal solution.

[0069] The contribution graph 34 positions the selected Pareto optimal solution from the optimal solution list 30 for each item of the machining conditions on the reference line Lr. Then, it displays information on how to change each item of the machining conditions to more reliably satisfy the required specifications. That is, in the example shown in FIG. 6, the bar graph of the horizontal feed speed extends to the right, which indicates that increasing the horizontal feed speed more reliably satisfies the required specifications. Also, the bar graph of the cutting depth amount extends to the left, which indicates that reducing the cutting depth more reliably satisfies the required specifications. Also, the amount by which the bar graph of the horizontal feed speed extends from the reference line Lr is longer than the amount by which the bar graph of the cutting depth amount extends from the reference line Lr, which indicates that the contribution of the horizontal feed speed is greater than the contribution of the cutting depth amount.

[0070] The contribution degree graph 34 may be a graph in which the processing condition items are arranged in descending order of contribution degree. Returning to Fig. 4, the PU 12 judges whether or not a command to change the machining conditions has been input by operating the user interface 20 (S42). If a command to change the machining conditions has been input (S42: YES), the PU 12 calculates the probability distribution of the machining result by inputting the machining conditions according to the command to the machining result output map (S44). The PU 12 then operates the display device 22 to display the calculation result obtained by the process of S44 (S46).

[0071] FIG. 7 illustrates the processes of S42 to S46. 7 shows an example in which the probability distribution curve CP of the geometric accuracy moves to the left while the probability distribution curve CP of the surface quality moves to the right as a result of a change command to increase the horizontal feed speed. Note that the change command for the machining conditions may be a process of moving the value of the machining condition item corresponding to the Pareto optimal solution selected in the optimal solution list 30 by operating the user interface 20.

[0072] 4, when the PU 12 completes the process of S46 or when the process of S38 or S42 is negative, the PU 12 determines whether or not a command to end the display has been input by operating the user interface 20 (S48). When the PU 12 determines that a command to end the display has not been input (S48: NO), the PU 12 returns to the process of S38.

[0073] On the other hand, if the PU 12 determines that a command to end has been input (S48: YES) or if the determination in the process of S10 is negative, the PU 12 temporarily ends the series of processes shown in FIG. 3 and FIG.

[0074] "Relearning process" A procedure for the re-learning process of the machining result output map is shown in Fig. 8. The process shown in Fig. 8 is realized by the PU 12 repeatedly executing the re-learning program 14e, for example, at a predetermined interval.

[0075] In the series of processes shown in Fig. 8, the PU 12 first judges whether or not a pair of processing conditions and processing results for the actual grinding process has been input by operating the user interface 20 (S50). The above input is made when the user measures the processing results after grinding is performed under the processing conditions actually adopted using the processes in Fig. 3 and Fig. 4. When the PU 12 judges that a pair of processing conditions and processing results for the actual grinding process has been input (S50: YES), it stores the pair of processing conditions and processing results in the storage device 14 (S52).

[0076] The PU 12 judges whether the amount of data stored in the storage device 14 by the process of S52 is equal to or greater than a predetermined amount (S54). The predetermined amount is an amount that determines a trigger for re-learning. When the PU 12 judges that the amount of data is equal to or greater than the predetermined amount (S54: YES), the PU 12 executes re-learning of the processing result output mapping (S56). That is, the PU 12 uses the processing results stored in the storage device 14 to determine the average and variance of each item of the processing results in the training data for re-learning. This process may be a process of correcting the average and variance of each item of the processing results previously used as training data. Then, the parameters of the processing result output mapping data 14b are updated using these averages and variances and the newly obtained processing conditions as training data.

[0077] When the re-learning is completed, the PU 12 updates the processing result output mapping data 14b stored in the storage device 14 with the new parameters obtained by the re-learning (S58). 8 when the process of S58 is completed or when a negative determination is made in the processes of S50 and S54. Incidentally, when the process of S56 is completed, the PU 12 erases the data stored in the process of S52. However, in order to subsequently execute the process of S56 again and obtain the mean and variance from all the training data up to that point, it is desirable to store data for determining the mean and variance in a separate storage area of ​​the storage device 14.

[0078] "Actions and Effects of the Present Embodiment" When the user inputs the required specifications, the PU 12 searches for machining results that satisfy the required specifications to a certain degree or higher among the machining results when various machining conditions are input to the machining result output map. Then, the machining results and machining conditions that satisfy the required specifications to a certain degree or higher are displayed on the display device 22. Here, the machining result output map is a trained model that has been trained using training data generated from actual grinding performed by an expert. Therefore, when the required specifications are given, machining conditions close to the machining conditions that an expert would select are displayed on the display device 22.

[0079] Furthermore, when the user instructs to change the processing conditions, the PU 12 calculates the processing result by inputting the changed processing conditions to the processing result output map. Then, the PU 12 displays the processing result corresponding to the changed processing conditions on the display device 22. This allows the user to understand how the processing result will change when the processing conditions are changed. This allows the user to select the actual processing conditions after understanding the change in the processing result in response to the change in the processing conditions.

[0080] In addition, by displaying the machining results according to the change instructions from the user, it is also possible to train workers who have little experience with grinding. According to the present embodiment described above, the following actions and effects can be obtained.

[0081] (1-1) Instead of displaying only the machining conditions that satisfy the required specifications to a certain degree or more on the display device 22, the PU 12 displays the machining conditions and a probability distribution of the machining result on the display device 22. Here, since grinding processing involves various factors, even if all the items defined as the machining conditions are the same, the machining result is not necessarily determined uniquely. Therefore, by displaying the probability distribution of the machining result, the user can be more accurately notified of information regarding the machining result actually obtained by grinding processing.

[0082] (1-2) When the user instructs to change the machining conditions, the PU 12 displays how the probability distribution changes due to the change in the machining conditions. This allows the user to understand how to change the machining conditions to more reliably satisfy the required specifications.

[0083] (1-3) There may be items that have a trade-off relationship among the multiple items of the processing result, such as when one item among the multiple items of the processing result is set to a more desirable value, another item may deviate from the desirable value. In such a case, it takes a lot of man-hours to create logic that narrows down the processing conditions that an expert ultimately judges to be optimal.

[0084] Therefore, the PU12 is made capable of displaying multiple Pareto optimal solutions for the required specifications. When multiple Pareto optimal solutions are displayed, the user can see the probability distribution of the machining results, or see the machining results when the machining conditions of the optimal solutions are changed. This helps the user determine the optimal machining conditions for the required specifications.

[0085] <Second embodiment> The second embodiment will be described below with reference to the drawings, focusing on the differences from the first embodiment.

[0086] "Prerequisite configuration" A processing support system according to this embodiment is shown in Fig. 9. In Fig. 9, the same reference numerals are used for the members corresponding to those shown in Fig. 1 for the sake of convenience.

[0087] The processing support system includes a plurality of processing support devices 10(1), 10(2), etc. as client terminals. The numbers in parentheses after the processing support devices 10 are given to identify the processing support devices 10. The processing support devices 10(1), 10(2), etc. are terminals located at different processing sites. The processing support devices 10(1), 10(2), etc. may be terminals owned by different processors. In the following, the processing support devices 10(1), 10(2), etc. are collectively referred to as the processing support device 10. In addition, in FIG. 10, the same reference numerals are given to the components corresponding to the processing support device 10 shown in FIG. 1 for convenience.

[0088] 9, the re-learning program 14e is not stored in the storage device 14 of the processing support device 10. The processing support device 10 also includes a communication device 16. The communication device 16 is capable of communicating with a server device 60 included in the processing support system via a network 70.

[0089] The server device 60 is a device that distributes the latest processing result output map data 14b to the processing support device 10. The server device 60 includes a PU 62, a storage device 64, and a communication device 66. The communication device 66 is a device for communicating with the processing support device 10 via a network 70. The PU 62 is a software processing device such as a CPU, a GPU, and a TPU. The storage device 64 may be an electrically rewritable non-volatile memory, a disk medium, or other storage medium.

[0090] The storage device 64 stores the processing result output image data 14b and the re-learning program 14e. "Relearning process" FIG. 10 shows a procedure of processing related to re-learning of the machining result output mapping data 14b. The processing shown in FIG. 10 consists of the following two processes. One is a process realized by the PU 12 repeatedly executing the machining support program 14a, for example, at a predetermined cycle. The other is a process realized by the PU 62 repeatedly executing the re-learning program 14e, for example, at a predetermined cycle. Incidentally, the machining support program 14a in this embodiment includes commands to execute the process of FIG. 10 executed by the PU 12 and the processes of FIG. 3 and FIG. 4. In the following, a series of processes shown in FIG. 10 will be described according to the time sequence of the process for actual re-learning.

[0091] In the series of processes shown in Fig. 10, the PU 12 of the processing support device 10 first judges whether or not a pair of processing conditions for the actual grinding process and a processing result has been input by operating the user interface 20 (S60). This process corresponds to the process of S50. When the PU 12 judges that a pair of processing conditions for the actual grinding process and a processing result has been input (S60: YES), the PU 12 operates the communication device 16 to transmit the pair of data to the server device 60 (S62).

[0092] In response to this, the PU 62 of the server device 60 receives the data relating to the set of processing conditions and processing results (S70). Then, the PU 62 stores the data relating to the set of processing conditions and processing results in the storage device 64 (S72). Next, the PU 62 determines whether the amount of data relating to the set of processing conditions and processing results stored in the storage device 64 is equal to or greater than a predetermined amount (S74). This process corresponds to the process of S54.

[0093] When the PU 62 determines that the amount of the data is equal to or greater than the predetermined amount (S74: YES), the PU 62 executes a re-learning process (S78). This process corresponds to the process of S56. The PU 62 then updates the processing result output mapping data 14b stored in the storage device 64 (S80). The PU 62 then operates the communication device 66 to transmit the updated processing result output mapping data 14b to the processing support device 10 (S82).

[0094] When the PU 62 completes the process of S82, the PU 62 temporarily ends the process executed by the PU 62 shown in FIG. Meanwhile, the PU 12 of the processing support device 10 receives the updated processing result output mapping data 14b (S64). Then, the PU 12 updates the processing result output mapping data 14b stored in the storage device 14 to the received processing result output mapping data 14b (S66).

[0095] When the process of S66 is completed or when a negative determination is made in the process of S60, the PU 12 temporarily ends the process executed by the PU 12 among the processes shown in FIG. "Effects of this embodiment" When grinding is actually performed at a plurality of machining sites, users at each machining site can input machining conditions and machining results at that time to the machining support device 10. When the machining conditions and machining results are input, the PU 12 transmits data indicating them to the server device 60. The PU 62 of the server device 60 updates the machining result output mapping data 14b using the data transmitted from the plurality of machining support devices 10. Then, the PU 62 transmits the updated machining result output mapping data 14b to the machining support devices 10 at the plurality of machining sites.

[0096] In this way, the PU 12 updates the processing result output mapping data 14b using data related to processing actually performed at each of a plurality of processing sites. Therefore, it is easy to increase the amount of data for re-learning. Therefore, it is easy to update the processing result output mapping data 14b to data that predicts the processing result with high accuracy.

[0097] <Third embodiment> The third embodiment will be described below with reference to the drawings, focusing on the differences from the second embodiment.

[0098] "Prerequisite configuration" A processing support system according to this embodiment is shown in Fig. 11. In Fig. 11, the members corresponding to those shown in Fig. 9 are denoted by the same reference numerals for the sake of convenience.

[0099] 11, the memory device 14 of the processing support device 10 stores a client processing support program 14f. Meanwhile, the memory device 64 of the server device 60 stores processing result output mapping data 14b, combination data 14c, relationship definition data 14d, a re-learning program 14e, and a server processing support program 14g.

[0100] The server processing support program 14g includes instructions to receive from the processing support device 10 the required specifications, the designation of processing conditions related to the processing of S14, the selection of the processing result related to S38, and an instruction to change the processing conditions related to the processing of S42. The server processing support program 14g also includes instructions to execute the processes of S10 to S34, S38, S42, and S44. The server processing support program 14g also includes an instruction to operate the communication device 66 to transmit the calculation result to the processing support device 10.

[0101] The client-use processing support program 14f includes instructions to receive the required specifications, the designation of processing conditions related to the processing of S14, the selection of the processing result related to S38, and an instruction to change the processing conditions related to the processing of S42. The client-use processing support program 14f also includes an instruction to operate the communication device 16 and transmit the received input to the server device 60. The client-use processing support program 14f also includes an instruction to receive the calculation result transmitted from the server device 60. The client-use processing support program 14f also includes an instruction to execute a process of displaying an image on the display device 22, such as the processes of S36, S40, and S46.

[0102] As described above, in this embodiment, the processing support device 10 receives a command from a user and transmits the received command to the server device 60. In addition, the processing support device 10 receives a calculation result by the server device 60 and displays the calculation result on the display device 22.

[0103] <Correspondence> The correspondence between the items in the above embodiment and the items described in the above "Means for solving the problem" column is as follows. In the following, the correspondence is shown for each number of the solving means described in the "Means for solving the problem" column. [1] The execution device corresponds to the PU 12 in FIG. 1. The storage device corresponds to the storage device 14 in FIG. 1. The request acquisition process corresponds to the process of S10 and S12. The processing result calculation process corresponds to the process of S26. The search process corresponds to the processes of S20 to S24 and S28 to S32. The processing result notification process corresponds to the process of S36. The processing condition change acceptance process corresponds to the process of S42. The change result notification process corresponds to the process of S46. The "value that satisfies the required specifications to a predetermined degree or more" corresponds to the processing result in which the probability that each item of the processing result satisfies the required specifications is a predetermined rate or more. [2] The designated item acceptance process corresponds to the process of S38. The contribution degree calculation process corresponds to the process of S34. The contribution degree notification process corresponds to the process of S40. [3] Corresponds to FIG. 6. [4] The variables indicating the machining accuracy correspond to the geometric accuracy, removal amount, surface quality, surface roughness Ra, scratches, and corner sagging. The variables indicating the load correspond to the machining time. [5] The variables indicating the grinding speed correspond to the cutting speed and the left-right feed speed. [6] The acquisition process corresponds to the process of S20. The change process corresponds to the process of generating search candidates (2) to (N) in the process of S22. [7] In the process of S22, each item of the machining condition selected in the process of S20 corresponds to changing according to the probability distribution. [8] The machining condition designation acceptance process corresponds to the processes of S14 and S16. [9] The search process corresponds to using the search candidates that satisfy the constraints specified by the relational specification data 14d in the processes of S20 and S22.

[10] The result acquisition process corresponds to the processes of S50 and S52. The re-learning process corresponds to the process of S56.

[11] Corresponds to FIG. 2. [12,14,15] The first execution unit corresponds to PU12 in Figure 11. The second execution unit corresponds to PU62 in Figure 11. [13,14,15] The first execution unit corresponds to PU12 in Figure 9 and PU12 in Figure 11. The second execution unit corresponds to PU62 in Figure 9 and PU62 in Figure 11.

[0104] <Other embodiments> This embodiment can be modified as follows: This embodiment and the following modifications can be combined with each other to the extent that no technical contradiction occurs.

[0105] "About machining condition variables" The processing condition variables are not limited to the set of variables exemplified in the above embodiment. For example, the processing condition variables may be composed of only some of the variables indicating each of the multiple items exemplified in the above embodiment.

[0106] The processing condition variables are not limited to the variables exemplified in the above embodiment. For example, they may include a variable indicating the rotation speed of the motor that rotates the grinding wheel. They may also include a variable specifying the type of processing aid. They may also include a variable specifying the type of measuring instrument.

[0107] "About processing result variables" The processing result variables are not limited to the set of variables exemplified in the above embodiment. For example, they may be composed of only some of the variables indicating each of the multiple items exemplified in the above embodiment.

[0108] The processing result variable, which is a variable indicating a load, is not limited to processing time. For example, it may be a variable indicating the cost of processing. Also, it may be a variable indicating the power consumption required for processing.

[0109] The processing result variables are not limited to the variables exemplified in the above embodiment. "About the output mapping of processing results" The neural network that configures the processing result output map is not limited to a fully connected forward propagation type neural network. For example, it may be a neural network equipped with an attention mechanism.

[0110] The trained model constituting the processing result output map does not necessarily have to be a neural network. The trained model constituting the processing result output map may be, for example, a gradient boosting decision tree. Also, for example, the trained model constituting the processing result output map may be a support vector regression model.

[0111] The trained model that constitutes the processing result output map does not necessarily have to be a regression model. For example, it may be a discrimination model. In that case, the output of the trained model that constitutes the processing result output map may be the output of a logistic sigmoid function. This allows the output of the trained model to be the value of a random variable that satisfies the required specifications. In that case, it becomes possible to display the change in the probability of satisfying the required specifications when the processing conditions are changed. Therefore, the process of presenting the results of changing the processing conditions is highly useful.

[0112] It is not essential that the trained model constituting the processing result output map indicates a probability distribution. For example, it may be an identification model such as a support vector machine that indicates whether the required specifications are satisfied or not depending on whether the sign of the output is positive or negative. In that case, a processing result that satisfies a predetermined number or more of a plurality of items indicating the processing result may be regarded as a processing result that satisfies the required specifications to a predetermined degree or more.

[0113] "Regarding related regulation data" The relationship definition data 14d does not necessarily have to be data that defines the relationship between the items of the processing conditions defined by the theories in engineering and the relationship between the items of the processing conditions obtained by the knowledge of an expert. For example, the relationship definition data 14d may be composed only of data that defines the relationship between the items of the processing conditions defined by the theories in engineering. Also, for example, the relationship definition data 14d may be composed only of data that defines the relationship between the items of the processing conditions obtained by the knowledge of an expert.

[0114] "About the designated item acceptance process" The designated item receiving process does not necessarily have to be a process for receiving a selection of the display of the corresponding processed result. For example, the designated item receiving process may be a process for receiving a selection of a specific item from the list when the PU 12 displays a list of items of the processed result.

[0115] "About contribution notification processing" The contribution degree notification process does not necessarily have to include information on how to change the processing conditions to more reliably satisfy the required specifications. The contribution degree notification process may be, for example, a process that only indicates the relative magnitude of the contributions between multiple items of the processing conditions.

[0116] The contribution degree notification process does not necessarily have to notify the contribution degree for all of the multiple processing condition items. For example, the process may be limited to notifying the contribution degree of a predetermined number of items having the highest contribution degree.

[0117] "About Acquisition Processing" In the process of S18, the first and second principal components of the vector of the items extracted from the processing result indicated by the combination data 14c and the required specifications are calculated, but this is not limited to this. For example, the PU 12 may extract the first and second principal components of the vector consisting of the extracted items. In this case, the PU 12 calculates the components parallel to the first and second principal components of the vector indicating the required specifications. This allows both the extracted items and the required specifications to be compressed into two dimensions.

[0118] The acquisition process is not limited to a process of selecting a processing result included in the combination data that is close to the required specifications using the first and second principal components. For example, the processing result and the required specifications may be quantified using components parallel to each of three or more principal components. The acquisition process does not necessarily have to be a process that uses principal component analysis. The acquisition process may be a process that acquires candidates for processing conditions that satisfy the required specifications using the cosine similarity between a vector indicating the processing result and a vector indicating the required specifications.

[0119] About modification processing The change process is not limited to a process that has a higher probability of changing to conditions close to each item of the processing conditions selected by the process of S20 than changing to conditions that are far from the selected items. For example, the change process may be a process that selects changed conditions evenly from among conditions whose differences from each item of the processing conditions selected by the process of S20 are within a predetermined range.

[0120] "About the search process" The search process does not necessarily include a process of first obtaining one candidate machining condition that satisfies the required specifications using the combination data 14c. For example, instead of the combination data, a process may be used that uses a mapping that inputs a variable indicating the required specifications and outputs a variable indicating the machining condition. This mapping may be a trained model trained using the combination data 14c as training data.

[0121] It is not essential that the search process includes a process of determining one candidate for the machining conditions by using data generated from the machining conditions and the machining results in the actually performed machining. The search process does not necessarily have to be a process of searching for machining conditions using the specified conditions specified in the processes of S14 and S16 as constraint conditions.

[0122] The search process does not necessarily have to be a process of searching for processing conditions using the relationship defined by the relationship defining data 14d as a constraint condition. The search process does not necessarily have to be a process of searching for a Pareto optimal solution. For example, it may be a process of assigning points to each item of the processing result according to the probability that it satisfies the required specifications, and selecting one processing result according to the total points.

[0123] "About combined data" In the above embodiment, no particular mention was made about updating the combination data 14c when re-learning is performed. However, the combination data 14c may be updated by adding new training data for re-learning to the combination data 14c. As described in the "Regarding search processing" section, in the case of using a mapping that inputs a variable indicating a required specification and outputs a variable indicating a processing condition, this may be updated using training data for re-learning.

[0124] "About the execution device" The execution device is not limited to one that executes software processing. For example, it may be equipped with a dedicated hardware circuit such as an ASIC that executes at least a part of the processing executed in the above embodiment. That is, the execution device may include a processing circuit having any of the following configurations (a) to (c). (a) A processing circuit equipped with a processing device that executes all of the above processing in accordance with a program. (b) A processing circuit equipped with a processing device that executes part of the above processing in accordance with a program, and a dedicated hardware circuit that executes the remaining processing. (c) A processing circuit equipped with a dedicated hardware circuit that executes all of the above processing. Here, there may be multiple software execution devices equipped with a processing device and a program storage device. Also, there may be multiple dedicated hardware circuits.

[0125] "About processing support equipment" 1, the re-learning program 14e does not have to be stored in the storage device 14. That is, the processing support device 10 does not have to have a function of re-learning.

[0126] "About the processing support system" In the third embodiment, the server device 60 executes all of the calculations of the search target, the calculation of the probability distribution of the processing result, and the calculation of the contribution degree, but this is not limited to the above. For example, the processing support device 10 may execute the processes of S14 to S20.

[0127] "About processing equipment" The processing device does not necessarily have to be a grinding device. For example, it may be a cutting device. [Explanation of symbols]

[0128] 10…Processing support equipment 30...Optimal solution list 32... Existence Probability Map 34…Contribution graph 40…Cursor 60...Server equipment

Claims

1. An execution device and a storage device, The storage device stores processing result output mapping data, which is data for defining a processing result output mapping, The processing result output mapping is a mapping that receives a value of a processing condition variable as an input and outputs a value of a processing result variable, The processing condition variable is a variable indicating a processing condition by a processing device, The processing result variable is a variable indicating a processing result by the processing device, the execution device is configured to execute a request acquisition process, a processing result calculation process, a search process, a processing condition notification process, a processing condition change acceptance process, and a change result notification process; The requirement acquisition process is a process of acquiring a requirement specification for a processing result, the machining result calculation process is a process for determining a value of the machining result variable when a value of the machining condition variable is input to the machining result output map, the search process is a process of searching for a value that satisfies a required specification to a predetermined degree or more among values ​​of the processing result variable obtained by the processing result calculation process, The processing condition notification process is a process of notifying the processing condition indicated by the search result of the search process and the corresponding processing result, the machining condition change reception process is a process of receiving an instruction to change the machining conditions to the machining conditions notified by the machining condition notification process, The change result notification process is a process of notifying the processing result corresponding to the value of the processing result variable obtained as an input of the processing result calculation process using the value of the processing condition variable in accordance with the change instruction.

2. The processing result has a plurality of items, The processing conditions include a plurality of items, The execution device is configured to execute a designated item receiving process, a contribution degree calculation process, and a contribution degree notification process; the designated item receiving process is a process of receiving an item designated by a user among the plurality of items related to the processing result, the contribution degree calculation process is a process of calculating a contribution degree of the item of the processing condition to the item of the processing result accepted by the designated item acceptance process, 2. The processing support device according to claim 1, wherein the contribution degree notification process notifies information indicating a contribution degree of the processing condition item to the processing result item accepted by the designated item acceptance process.

3. the contribution degree calculation process includes a change direction calculation process for determining a way to change the item of the processing result accepted by the designated item acceptance process so as to make it more suitable for the required specifications, 3. The processing support device according to claim 2, wherein the contribution degree notification process includes a process of notifying information indicating the contribution degree of each item of the processing result together with information indicating how to change the processing result to better conform to the required specifications.

4. 3. The machining support device according to claim 2, wherein the machining result variables include a variable indicating machining accuracy and a variable indicating a load applied to the machining.

5. The processing device is a grinding device, 3. The machining support device according to claim 2, wherein the machining condition variables include a variable indicating a cutting amount in the grinding process, a variable indicating a grinding wheel speed in the grinding process, and a variable indicating a grinding fluid flow rate in the grinding process.

6. The storage device stores combination data, the combination data is data indicating a combination of the processing result and the processing conditions, The processing support device according to claim 1, wherein the search process includes an acquisition process for acquiring the processing conditions corresponding to the processing results included in the combination data that are closest to the required specifications, and a change process for changing the processing conditions to be searched for among the processing conditions acquired by the acquisition process.

7. The processing conditions include a plurality of items, The processing support device according to claim 6, wherein the change process includes a process for making the probability of changing, for one item of the processing conditions acquired by the acquisition process, a condition close to the acquired condition for the same item greater than the probability of changing to a condition far from the acquired condition.

8. The processing conditions include a plurality of items, The execution device is configured to execute a processing condition designation reception process, The processing condition designation receiving process is a process of receiving a user's designation of some items related to the processing conditions, 2. The machining support device according to claim 1, wherein the search process searches for the machining conditions under constraints that satisfy the specified items.

9. The processing conditions include a plurality of items, The storage device stores related definition data, The relationship definition data is data defining a relationship between a plurality of items of the processing conditions, 2. The machining support device according to claim 1, wherein the search process searches for the machining conditions that satisfy the relationship defined by the relationship defining data.

10. The execution device is configured to execute a result acquisition process, a relearning process, and an update process; the result acquisition process is a process of acquiring the processing result by the processing device and the corresponding processing conditions, the re-learning process is a process of re-learning the processing result output mapping using the data acquired by the result acquisition process as training data, 2. The machining support device according to claim 1, wherein the update process is a process of updating the machining result output map data stored in the storage device to the re-learned data.

11. The processing device is a grinding device, The processing result variable includes at least one of five variables: a hammering mark variable, a feed mark variable, a burn mark variable, a scratch variable, and a corner sagging variable; The knock mark variable is a variable indicating a pattern that appears on the surface of the object to be machined in a direction perpendicular to the machining direction, the feed stitch variable is a variable indicating a pattern appearing in a machining direction on the surface of the workpiece, The burnt parameter is a parameter indicating a degree of discoloration of the surface of the object to be processed due to processing heat, The scratch variable is a variable indicating the degree of damage to the surface of the processing object, The machining support device according to claim 1 , wherein the corner sagging variable is a variable indicating a degree to which a peripheral portion of the workpiece has been excessively cut.

12. The processing support device according to claim 1, further comprising the execution device and the storage device, the execution device comprises a first execution device and a second execution device; The first execution unit is configured to execute at least the request acquisition process and the processing condition change acceptance process, The second execution unit is a processing support system configured to execute at least the processing result calculation process.

13. The processing support device according to claim 10, further comprising the execution device and the storage device, the execution device comprises a first execution device and a second execution device; The first execution device is configured to execute at least the result acquisition process; The second execution unit is a machining support system configured to execute at least the relearning process.

14. A machining support device comprising the first execution device in the machining support system according to claim 12.

15. A machining support device comprising the first execution device in the machining support system according to claim 13.

Images