Specific Cutting Force Estimation System Utilizing a Knowledge Model

Abstract

To provide a specific cutting resistance estimation system utilizing a knowledge model that is able to estimate a specific cutting resistance highly accurately by constructing and utilizing the knowledge model related to the specific cutting resistance.SOLUTION: When at least machining conditions for cutting, specifications of a tool T, and specifications of a workpiece W are defined as parent factors A, B, and C, a specific cutting resistance is a child factor D, and factor values of the parent factors A, B, C are conditioned, a knowledge model M comprises a conditional probability table M1 in which probabilities P11-P18, P21-P28, P31-P38 corresponding to a factor value of the child factor D are defined. Specific cut resistance estimation systems 1, 100 include: a knowledge model storage unit 22 that stores a knowledge model M; and a specific cut resistance output unit 23 that, when the factor values of the parent factors A, B, C are acquired, performs forward probability inference by means of the conditional probability table M1, thereby outputting an estimated value of the specific cut resistance as the child factor D.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present invention relates to a specific cutting resistance estimation system that utilizes a knowledge model. [Background technology]

[0002] Patent Documents 1 and 2 describe that in order to suppress chatter (also called chatter vibration) during cutting, machining conditions are determined using a stability limit diagram that defines the relationship between the rotational frequency of the tool or workpiece and the maximum depth of cut at which chatter does not occur. In Patent Documents 1 and 2, the stability limit diagram is generated using the dynamic characteristics represented by the mass (M), damping coefficient (C), and rigidity (K) of the tool or workpiece, and the specific cutting resistance. The specific cutting resistance is the value obtained by dividing the cutting resistance by the cutting cross-sectional area, that is, the cutting resistance per unit cross-sectional area.

[0003] It is also known that in order to determine the processing conditions for a processing device, the knowledge of experts is compiled into a database and the database is utilized. For example, Patent Document 3 describes a knowledge model compiled into a database. The knowledge model is a model that is composed of a plurality of factors defined by various technical terms and the relationships between the factors. It is said that the processing conditions can be optimized by using the knowledge model.

[0004] Patent Document 4 describes the application of machine learning when defining the relationship between factors in a knowledge model. In particular, as a means for defining the relationship between factors, a means for defining the correspondence relationship (rank-rank relationship) between the rank value of a reference factor and the rank value of a factor to be connected, and a means for defining the correspondence relationship (rank-range relationship) between the rank value of a reference factor and the range value of a factor to be connected. Furthermore, when a reference factor has a relationship with multiple factors, the degree of influence of multiple factors on the reference factor can be determined by setting the contribution degree. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent No. 5862111 [Patent Document 2] JP 2012-200844 A [Patent Document 3] JP 2020-177547 A [Patent Document 4] Patent Publication No. 2021-071856 Summary of the Invention [Problem to be solved by the invention]

[0006] The specific cutting resistance may be a value according to the material of the workpiece. A detailed analysis revealed that the specific cutting resistance varies not only depending on the material of the workpiece, but also on the shape of the workpiece, the material and shape of the tool, the machining conditions, etc. However, it is not easy to estimate the specific cutting resistance with high accuracy, which is affected by various factors such as the specifications of the workpiece, the specifications of the tool, and the machining conditions.

[0007] The present invention has been made in consideration of such problems, and aims to provide a specific cutting resistance estimation system that utilizes a knowledge model that can estimate specific cutting resistance with high accuracy by constructing and utilizing a knowledge model regarding specific cutting resistance. [Means for solving the problem]

[0008] One aspect of the present invention is a specific cutting resistance estimation system that includes a processor and a storage device and utilizes a knowledge model in cutting a workpiece with a tool, the storage device includes a knowledge model storage unit for storing the knowledge model; The knowledge model is configured to include a conditional probability table in which at least the machining conditions of the cutting process, the specifications of the tool, and the specifications of the workpiece are parent factors, and a specific cutting resistance is a child factor, and a probability that the factor value of the child factor corresponds to a conditional factor value of the parent factor is defined; The processor is in a specific cutting resistance estimation system that utilizes a knowledge model, and is provided with a specific cutting resistance output unit that, when it obtains a factor value of the parent factor, performs forward probabilistic inference using the conditional probability table that constitutes the knowledge model, thereby outputting an estimated value of the specific cutting resistance as the child factor. Effect of the Invention

[0009] In the specific cutting resistance estimation system, a conditional probability table is configured as a knowledge model. The conditional probability table is defined by the probability that the factor value of a child factor corresponds to the factor value of a parent factor when the factor value of a parent factor is set as a condition. In other words, the conditional probability table does not directly relate the factor values ​​of a parent factor and a child factor to each other, but expresses them using probabilities. Therefore, the relationship between factors can be flexibly defined.

[0010] The parent factors include the machining conditions of the cutting process, the tool specifications, and the workpiece specifications. The child factor is the specific cutting resistance. By making the knowledge model a conditional probability table, even if there are many types of parent factors, it is easy to define the probability that the factor value of the child factor, which is the specific cutting resistance, corresponds to the conditional factor value of multiple parent factors.

[0011] Then, when the specific cutting resistance output unit obtains factor values ​​related to the machining conditions, tool specifications, and workpiece specifications as parent factors, it can obtain an estimated value of the specific cutting resistance as a child factor by performing forward probabilistic inference using the conditional probability table that constitutes the knowledge model.

[0012] As described above, according to the above aspect, by constructing and utilizing a knowledge model regarding the specific cutting resistance, it is possible to provide a specific cutting resistance estimation system that utilizes a knowledge model that can estimate the specific cutting resistance with high accuracy. [Brief description of the drawings]

[0013] [Figure 1] FIG. 1 is a diagram showing a configuration of a specific cutting resistance estimation system according to a first embodiment. [Diagram 2]FIG. 1 is a diagram showing a cutting model. [Diagram 3] 2 is a diagram showing a configuration of a management device constituting the specific cutting resistance estimation system of FIG. 1. [Figure 4] 4 is a diagram showing a configuration of a specific cutting resistance generating unit constituting the management device of FIG. 3. [Diagram 5] 5 is a diagram showing a configuration of a knowledge model storage unit constituting the specific cutting resistance generating unit of FIG. 4. [Figure 6] FIG. 6 is a diagram showing the concept of a conditional probability table which is a part of the knowledge model storage unit in FIG. 5. [Figure 7] FIG. 6 is a diagram showing a conditional probability table which is a part of the knowledge model storage unit in FIG. [Figure 8] 6 is a diagram showing a probability table of factor values ​​for each factor, which is part of the knowledge model storage unit in Fig. 5. (a) is a probability table for factor A, (b) is a probability table for factor B, and (c) is a probability table for factor C. [Figure 9] 5 is a diagram showing a configuration of a knowledge model generation unit constituting the specific cutting resistance generation unit of FIG. 4. [Figure 10] 10 is a conditional probability table generated based on the parameters of the prior distribution stored in a prior distribution storage unit constituting the knowledge model generation unit of FIG. 9. [Figure 11] 10 is a conditional probability table generated based on a posterior distribution generated by a conditional probability table generating unit constituting the knowledge model generating unit in FIG. 9. [Figure 12] 1 shows a procedure for updating a conditional probability table, in which the initial state of the conditional probability table is generated using only the parameters of the prior distribution. [Figure 13] 1 shows a procedure for updating a conditional probability table, in which an updated conditional probability table is generated based on a posterior distribution. [Figure 14] 13 shows a procedure for updating a conditional probability table, in which a further updated conditional probability table is generated based on a new posterior distribution. [Figure 15] 5 is a flowchart showing a process performed by a specific cutting resistance output unit constituting the specific cutting resistance generating unit of FIG. 4 when performing forward probability inference. [Figure 16]5 is a flowchart showing a process performed by a specific cutting resistance output unit constituting the specific cutting resistance generating unit of FIG. 4 when performing backward probability inference. [Figure 17] FIG. 11 is a diagram showing the configuration of a specific cutting resistance estimation system according to a second embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0014] (Embodiment 1) 1. Configuration of specific cutting resistance estimation system 1 The configuration of a specific cutting resistance estimation system 1 of the first embodiment will be described with reference to Fig. 1. The specific cutting resistance estimation system 1 is applied to cutting of a workpiece W by a tool T. In particular, the specific cutting resistance estimation system 1 utilizes a knowledge model M for estimating the specific cutting resistance in the cutting. That is, the specific cutting resistance estimation system 1 estimates the specific cutting resistance by using the knowledge model M based on the machining conditions of the cutting, the specifications of the tool T, and the specifications of the workpiece W.

[0015] Furthermore, the specific cutting resistance estimation system 1 can generate a stability limit diagram and determine machining conditions for suppressing the occurrence of chatter using the estimated specific cutting resistance. Also, the specific cutting resistance estimation system 1 can obtain machining conditions, specifications of the tool T, and specifications of the workpiece W that result in a target specific cutting resistance by using the knowledge model M.

[0016] As shown in Fig. 1, the specific cutting resistance estimation system 1 is configured to include a management device 2 that is mainly operated by a manager, a display device 3 that is held by the manager and the worker, and a plurality of processing devices 4, 5 for performing cutting processing. The management device 2, the display device 3, and the processing devices 4, 5 form a network and are configured to be able to communicate with each other. In other words, the management device 2, the display device 3, and the processing devices 4, 5 are devices that form a network.

[0017] The management device 2 is configured with a processor 2a, a storage device 2b, a display device 2c, and an input device 2d. The processor 2a performs arithmetic processing using a knowledge model M, and the storage device 2b stores the knowledge model M and the like. The management device 2 can also function as a server that manages processing devices 4 and 5. The processing devices 4 and 5 are, for example, lathes, machining centers, milling machines, gear processing machines, boring machines, and the like.

[0018] 2. Machining model The cutting model will be described with reference to Fig. 2. Fig. 2 shows a model using turning as an example, in which a workpiece W rotates around a central axis L and a tool T is fed in the axial direction of the workpiece W. However, the cutting model shown in Fig. 2 is a model that can be applied to cutting in general, and is also applicable to the case where cutting is performed by rotating the tool T.

[0019] As shown in FIG. 2, when cutting a workpiece W with a tool T, a cutting resistance F occurs at a cutting point P. The cutting resistance F is represented by a principal force F1, a thrust force F2, and a feed force F3. In FIG. 2, the principal force F1 corresponds to a component in the tangential direction of a circle centered on the central axis L of the workpiece W, the thrust force F2 corresponds to a component in the normal direction of the cutting point P, and the feed force F3 corresponds to a component in the feed direction of the tool T (the axial direction of the workpiece W). The principal force F1, thrust force F2, and feed force F3 are mutually orthogonal. The cutting resistance F can also be decomposed into the principal force F1, thrust force F2, and feed force F3 in cutting processes other than those using a lathe.

[0020] 3. Configuration of Management Device 2 The configuration of the management device 2 will be described with reference to Fig. 3. As shown in Fig. 3, the management device 2 includes, as a processor 2a and a storage device 2b, at least a compliance transfer function generating unit 11, a specific cutting resistance generating unit 12, a stability limit diagram generating unit 13, a recommended condition determining unit 14, and a teaching unit 15. Furthermore, as shown in Fig. 3, the management device 2 includes, in addition to the above-mentioned units 11 to 15, a display device 2c.

[0021] The compliance transfer function generating unit 11 generates a compliance transfer function of at least one of the tool T and the workpiece W. The compliance transfer function corresponds to the dynamic characteristics of at least one of the tool T and the workpiece W. In other words, the compliance transfer function corresponds to the dynamic characteristics represented by the mass (M), damping coefficient (C), and stiffness (K) of the tool T and the workpiece W.

[0022] More specifically, the compliance transfer function is a transfer function of a response displacement when an excitation force is applied to the tool T, a transfer function of a response displacement when an excitation force is applied to the workpiece W, and a transfer function related to both the tool T and the workpiece W. The compliance transfer function is a frequency response function. The compliance transfer function can be generated, for example, by a hammering test, a simulation, a theoretical calculation, or the like. Note that a method for generating the compliance transfer function is well known, and therefore a detailed description thereof will be omitted. The compliance transfer function generating unit 11 is configured by the processor 2a.

[0023] The specific cutting resistance generating unit 12 generates the specific cutting resistance based on the machining conditions, the specifications of the tool T, and the specifications of the workpiece W. The specific cutting resistance corresponds to a value obtained by dividing the cutting resistance F by the cutting cross-sectional area. However, in this embodiment, the thrust force F2 of the cutting resistance F has a large effect on chatter. Therefore, the specific cutting resistance is set to a value obtained by dividing the thrust force F2 (shown in FIG. 2) constituting the cutting resistance F by the cutting cross-sectional area.

[0024] As will be described later, the specific cutting resistance generating unit 12 generates an estimated value of the specific cutting resistance by performing forward probabilistic inference using a conditional probability table M1 constituting the knowledge model M. The specific cutting resistance generating unit 12 can also determine the machining conditions of the cutting work, the specifications of the tool T, and the specifications of the workpiece W by performing backward probabilistic inference. The specific cutting resistance generating unit 12 is composed of a processor 2a and a storage device 2b.

[0025] The stability limit diagram generating unit 13 generates a stability limit diagram for cutting processing using the compliance transfer function generated by the compliance transfer function generating unit 11 and the estimated value of the specific cutting resistance generated by the specific cutting resistance generating unit 12. The stability limit diagram is a diagram that defines the relationship between the rotational frequency of the tool T or the workpiece W and the maximum cutting depth at which chatter does not occur. The lower region of the stability limit diagram corresponds to a stable region where chatter does not occur, and the upper region corresponds to an unstable region where chatter may occur. The stability limit diagram generating unit 13 is composed of the processor 2a.

[0026] The recommended condition determination unit 14 determines the recommended values ​​of the machining conditions using the stability limit diagram generated by the stability limit diagram generation unit 13. In detail, the recommended condition determination unit 14 sets the machining conditions located in the lower region of the stability limit diagram as the recommended values. The recommended condition determination unit 14 is configured by the processor 2a. The recommended condition determination unit 14 can also determine whether the current machining conditions are located in a stable region.

[0027] The teaching unit 15 teaches the recommended values ​​of the machining conditions determined by the recommended condition determination unit 14. As a means for teaching the recommended values ​​of the machining conditions, the teaching unit 15 displays them on the display device 2c constituting the management device 2, or on the display device 3. The display device 3 displays the recommended values ​​of the machining conditions acquired from the teaching unit 15 through communication.

[0028] Further, the teaching unit 15 transmits the recommended values ​​of the machining conditions to the machining devices 4 and 5 through communication. In this case, the machining devices 4 and 5 can obtain the recommended values ​​of the machining conditions taught by the teaching unit 15 and perform cutting processing using the obtained recommended values ​​of the machining conditions. That is, the machining devices 4 and 5 autonomously change the machining conditions and perform cutting processing based on the changed machining conditions. The machining results when the machining devices 4 and 5 perform cutting processing are transmitted to the specific cutting resistance generating unit 12 as observation data. The specific cutting resistance generating unit 12 can generate or update the knowledge model M using the observation data in the machining devices 4 and 5. The teaching unit 15 is configured by the processor 2a.

[0029] 4. Configuration of specific cutting resistance generating unit 12 The configuration of the specific cutting resistance generating unit 12 will be described with reference to Fig. 4 to Fig. 8. As shown in Fig. 4, the specific cutting resistance generating unit 12 is configured with a knowledge model generating unit 21, a knowledge model storage unit 22, and a specific cutting resistance output unit 23. The knowledge model generating unit 21 is configured with a processor 2a and a storage device 2b, the specific cutting resistance output unit 23 is configured with the processor 2a, and the knowledge model storage unit 22 is configured with the storage device 2b.

[0030] The knowledge model generation unit 21 generates the knowledge model M. The knowledge model storage unit 22 stores the knowledge model M generated by the knowledge model generation unit 21.

[0031] As shown in Fig. 5, the knowledge model M includes a conditional probability table M1, and probability tables M2, M3, and M4 for factors A, B, and C. The conditional probability table M1 is a model for estimating the specific cutting resistance. As described above, chatter is significantly affected by the thrust force F2 of the cutting resistance F in Fig. 2. Therefore, the specific cutting resistance to be estimated is the value obtained by dividing the thrust force F2 of the cutting resistance F by the cutting cross-sectional area.

[0032] The conditional probability table M1 defines the probability of the factor value of the child factor D being met when the factor values ​​of at least the parent factors A, B, and C are taken as conditions. Here, the specific cutting resistance is affected by the machining conditions of the cutting process, the specifications of the tool T, the specifications of the workpiece W, etc. In other words, the specific cutting resistance is a value that differs depending on the machining conditions of the cutting process, the specifications of the tool T, and the specifications of the workpiece W.

[0033] Therefore, in the conditional probability table M1, the parent factors A, B, and C are factors regarding the machining conditions of the cutting process, factors regarding the specifications of the tool T, and factors regarding the specifications of the workpiece W. The conditional probability table M1 may further include factors other than the above factors A, B, and C as parent factors. The child factor D is the specific cutting resistance.

[0034] That is, the conditional probability table M1 constituting one of the knowledge models M has at least the cutting conditions, the specifications of the tool T, and the specifications of the workpiece W as parent factors A, B, and C, and the specific cutting resistance as a child factor D. The conditional probability table M1 is a model for estimating the specific cutting resistance based on the cutting conditions, the specifications of the tool T, and the specifications of the workpiece W.

[0035] 6, factors of the machining conditions include, for example, cutting speed, feed rate, cutting depth, etc. Specifications of the tool T include, for example, the material (substance) of the tool T, the shape of the tool T (cutting edge angle, etc.), the presence or absence of coating and the type, etc. Specifications of the workpiece W include, for example, the material (substance) of the workpiece W, the shape of the workpiece W, the presence or absence of heat treatment and the type, etc.

[0036] Furthermore, the machining conditions, the specifications of tool T, and the specifications of workpiece W are each expressed by a numerical factor value. In the case of machining conditions as parent factor A, the numerical values ​​representing the cutting speed, feed rate, cutting depth, etc. are each the factor values ​​of parent factor A. In the case of the specifications of tool T as parent factor B, the physical property values ​​representing the material of tool T, the cutting edge angle representing the shape of tool T, and the physical property values ​​representing the type of coating are each the factor values ​​of parent factor B. In the case of the specifications of workpiece W as parent factor C, the same is true for the specifications of tool T.

[0037] The conditional probability table M1 is conceptually expressed as shown in Fig. 7. For example, the conditional probability table M1 defines probabilities P11-P18, P21-P28, and P31-P38 corresponding to the factor values ​​D1, D2, and D3 of the child factor D when the factor values ​​A1-A2, B1-B2, and C1-C2 of the parent factors A, B, and C are set as conditions. In the conditional probability table M1 shown in Fig. 7, the sum of the probabilities in the vertical columns is 1. For example, the sum of the probabilities P11, P21, and P31 is 1.

[0038] 7, for example, the factor values ​​A1-A2, B1-B2, C1-C2, and D1-D3 of the factors A-D may be one value or a range of values. For example, the factor values ​​A1 and A2 of the parent factor A are, for example, a cutting speed value, the factor values ​​B1 and B2 of the parent factor B are, for example, a cutting angle value of the tool T, and the factor values ​​C1 and C2 of the parent factor C are, for example, a physical property value representing the material of the workpiece W.

[0039] For example, if the factor values ​​of parent factors A, B, and C are A1, B1, and C1, the probability of child factor D corresponding to factor value D1 is P11, the probability of child factor D corresponding to factor value D2 is P21, and the probability of child factor D corresponding to factor value D3 is P31. For example, P11 is 0.667, P21 is 0.333, and P31 is 0.

[0040] As described above and as shown in Fig. 5, the knowledge model M includes probability tables M2, M3, and M4 for factors A, B, and C in addition to the conditional probability table M1. For example, the probability table M2 for factor A defines probabilities Pa1 and Pa2 corresponding to factor values ​​A1 and A2 of factor A as shown in Fig. 8(a). The sum of the probabilities Pa1 and Pa2 is 1.

[0041] As shown in Fig. 8(b), probability table M3 for factor B defines probabilities Pb1 and Pb2 corresponding to factor values ​​B1 and B2 of factor B. The sum of probabilities Pb1 and Pb2 is 1. As shown in Fig. 8(c), probability table M4 for factor C defines probabilities Pc1 and Pc2 corresponding to factor values ​​C1 and C2 of factor C. The sum of probabilities Pc1 and Pc2 is 1.

[0042] 4, the specific cutting resistance output unit 23 outputs an estimated value of the specific cutting resistance using the knowledge model M stored in the knowledge model storage unit 22. In detail, when the specific cutting resistance output unit 23 acquires the factor values ​​of the parent factors A, B, and C, it performs forward probability inference using the conditional probability table M1 constituting the knowledge model M. The forward probability inference is to estimate the child factor D based on the parent factors A, B, and C defined as conditions.

[0043] In this way, the specific cutting resistance output unit 23 outputs an estimated value of the specific cutting resistance as the child factor D. For example, the specific cutting resistance output unit 23 can output an expected value. That is, the factor values ​​D1, D2, and D3 of the child factor D are multiplied by the corresponding probabilities P11, P21, and P31, respectively, and the sum of the obtained values ​​is output as the specific cutting resistance. In this case, the specific cutting resistance is "D1×P11+D2×P21+D3×P31".

[0044] Furthermore, when the specific cutting resistance output unit 23 acquires the target factor value of the child factor D, which is the target value of the specific cutting resistance, it can also perform backward probability inference using the conditional probability table M1 constituting the knowledge model M. The backward probability inference is to estimate the parent factors A, B, and C defined as conditions based on the factor value of the child factor D. In this case, the specific cutting resistance output unit 23 can further use the probability tables M2, M3, and M4 of the factors A, B, and C to output factor values ​​for the machining conditions of the cutting process, the specifications of the tool T, and the specifications of the workpiece W as parent factors.

[0045] 5. Configuration of knowledge model generation unit 21 The knowledge model generation unit 21 generates a knowledge model M as shown in Fig. 4. In particular, the knowledge model generation unit 21 generates a conditional probability table M1 constituting the knowledge model M. Here, the conditional probability table M1 may be set by directly inputting the values ​​of the probabilities P11 to P18, P21 to P28, and P31 to P38. In addition, the conditional probability table M1 may be generated using Bayesian estimation, as described below. Furthermore, the conditional probability table M1 may be updated using Bayesian estimation.

[0046] The configuration of the knowledge model generation unit 21 will be described with reference to Fig. 9 to Fig. 14. As shown in Fig. 9, the knowledge model generation unit 21 includes an observation data acquisition unit 41, an interval determination unit 42, an interval edit processing unit 43, a prior distribution storage unit 44, a conditional probability table generation unit 45, a weight storage unit 46, and a model edit processing unit 47. The observation data acquisition unit 41, the interval determination unit 42, the interval edit processing unit 43, the conditional probability table generation unit 45, and the model edit processing unit 47 are configured by the processor 2a, and the prior distribution storage unit 44 and the weight storage unit 46 are configured by the storage device 2b.

[0047] The observation data acquisition unit 41 acquires observation data 41a on the factor values ​​of parent factors A, B, and C and the factor value of child factor D. The observation data 41a may be data obtained by performing an actual cutting process, or may be data obtained by performing a simulation of the cutting process. The observation data 41a may be data obtained by performing an actual cutting process, or may be data obtained by performing a simulation of the cutting process.

[0048] The observation data acquisition unit 41 is configured to be able to communicate with the processing devices 4, 5, and can therefore acquire observation data 41a when actual cutting processing is performed by the processing devices 4, 5. The observation data acquisition unit 41 can acquire the observation data 41a of the processing devices 4, 5 in real time, and can also acquire a set of observation data 41a for a predetermined period of time rather than in real time. Furthermore, the observation data acquisition unit 41 can acquire the observation data 41a of the multiple processing devices 4, 5 in a centralized manner.

[0049] The observation data acquisition unit 41 acquires the machining conditions used in the actual cutting, the specifications of the tool T, and the specifications of the workpiece W as data obtained by performing actual cutting. Furthermore, when the actual cutting is performed, cutting resistance data is detected by a sensor such as a cutting dynamometer installed on the processing device 4, 5, and a value obtained by dividing the cutting resistance data by the cutting cross-sectional area is calculated as the specific cutting resistance. Then, the observation data acquisition unit 41 acquires the calculated specific cutting resistance.

[0050] In this way, when actual cutting processing is performed, the observation data acquisition unit 41 can acquire observation data 41a regarding the factor values ​​of the parent factors A, B, and C and the factor value of the child factor D. The sensors installed on the processing devices 4 and 5 can acquire the power values ​​(current values, etc.) of each driving motor in addition to the cutting dynamometer, and can also calculate the specific cutting resistance.

[0051] Furthermore, when a simulation of cutting is performed, the observation data acquisition unit 41 can acquire the machining conditions, the specifications of the tool T, and the specifications of the workpiece W as input information for the simulation, and can also acquire the specific cutting resistance as a result of the simulation. In this way, when a simulation of cutting is performed, the observation data acquisition unit 41 can acquire observation data 41a regarding the factor values ​​of the parent factors A, B, and C and the factor value of the child factor D.

[0052] The interval determination unit 42 determines the discrete intervals by discretizing each of the factor values ​​of the parent factors A, B, and C and the factor value of the child factor D in the conditional probability table M1 shown in Fig. 7. The interval determination unit 42 can select one of a plurality of discrete interval determination methods. For example, the interval determination unit 42 can determine an interval in which the factor values ​​of the factors A, B, C, and D are evenly discretized.

[0053] Furthermore, the interval determination unit 42 can determine intervals in which the factor values ​​of each of the factors A, B, C, and D are discretized unevenly. As an example of uneven discretization, the interval determination unit 42 can determine discrete intervals based on mutual information I(X;Y) between each random variable in each of the factors A, B, C, and D. The mutual information I(X;Y) is an amount that represents a measure of the interdependence of two random variables. The mutual information I(X;Y) represents how much one variable can be predicted by understanding the other variable. As another example of uneven discretization, the interval determination unit 42 can determine intervals that are discretized with equal frequency.

[0054] The section edit processing unit 43 is configured to receive input from an operator and edit discrete sections in response to the operator's input. The section edit processing unit 43 makes it possible to adjust the discrete sections determined by the section determination unit 42 based on the operator's input. The section edit processing unit 43 is also capable of creating discrete sections from scratch based on the operator's input.

[0055] The prior distribution storage unit 44 stores a parameter α of a pre-set prior distribution 44a in order to determine the conditional probability table M1. jk In the following, for ease of explanation, the relationship between the parent factor A and the child factor D will be explained. The prior distribution 44a uses the Dirichlet distribution as expressed in formula (1). In formula (1), j is the discrete interval of the parent factor A, and k is the discrete interval of the child factor D.

[0056]

number

[0057] Furthermore, the prior distribution storage unit 44 stores the parameter α jk The expected value θa of the prior distribution 44a generated based on jk The conditional probability table 44b stored in the prior distribution storage unit 44 has element values ​​of the above. Here, the conditional probability table 44b stored in the prior distribution storage unit 44 is generated based on the discrete intervals determined by the interval determination unit .

[0058] Expected value θa of prior distribution 44a jk is represented by formula (2). In formula (2), Na j is the parameter α of the prior distribution 44a in the discrete interval j. jk is the sum of Na jk is the parameter α of the prior distribution 44a in the discrete interval j, k. jk It is.

[0059]

number

[0060] The conditional probability table 44b shown in FIG. 10 shows the corresponding expected value θa jk Here, the parameter α of the prior distribution 44a is jk is set based on, for example, the worker's experience, information on similar fields, etc. In this case, the conditional probability table 44b based on the prior distribution 44a is a conditional probability table generated based on the worker's experience, information on similar fields, etc.

[0061] 9, the conditional probability table generating unit 45 acquires the observed data 41a acquired by the observed data acquiring unit 41. jk is expressed by the formula (3). jk is expressed by a multinomial distribution. jk The maximum likelihood estimator φ for jk is expressed by equation (4). j is the total number of observation data 41a in the discrete interval j, and N jk is the data frequency β of the observation data 41a in the discrete interval j, k. jk It is.

[0062]

number

[0063]

number

[0064] The conditional probability table generating unit 45 generates the conditional probability table based on the observation data 41 a acquired by the observation data acquiring unit 41 and the parameter α jk Based on this, a posterior distribution 45a is calculated using Bayesian estimation. The posterior distribution 45a is a Dirichlet distribution that is configured with the same discrete intervals as the prior distribution 44a.

[0065] Furthermore, the conditional probability table generating unit 45 calculates the expected value θb of the posterior distribution 45a based on the calculated posterior distribution 45a. jk Here, the conditional probability table M1 is generated based on the discrete intervals determined by the interval determination unit 42. In other words, the conditional probability table M1 generated by the conditional probability table generation unit 45 is composed of the same discrete intervals as the conditional probability table 44b generated based on the prior distribution 44a. The generated conditional probability table M1 is shown in the upper part of FIG. 11. The conditional probability table M1 has the value θb jk is stored.

[0066] The lower part of Fig. 11 shows a diagram for explaining in detail the discrete interval j of the parent factor A in the upper part of Fig. 11. Each element value θb jk is the expected value θb of the posterior distribution 45a in the lower part of FIG. jk Matches.

[0067] Expected value θb of posterior distribution 45a jk is expressed by equation (5). In equation (5), W j is the weight. Na jk is the parameter α of the prior distribution 44a jk In the discrete interval j, j is the parameter α of the prior distribution 44a in the discrete interval j. jk It is the sum of N jk is the data frequency β of observation data 41ajk is the value of the discrete interval j in N j is the total number of data in the observation data 41a in the discrete interval j.

[0068]

number

[0069] That is, the conditional probability table generating unit 45 calculates the observed data 41a and the parameter α jk Based on this, the posterior distribution 45a is calculated using Bayesian estimation, and the expected value θb jk A conditional probability table M1 is generated with each element value being:

[0070] Here, the conditional probability table generating unit 45 uses weights W j In other words, the posterior distribution 45a is based on the observed data 41a and the weights W j Parameter α of the prior distribution 44a jk Based on the above, the Bayesian estimation is used to calculate each element value Nb jk is expressed by the formula (6). jk corresponds to the numerator of equation (5).

[0071]

number

[0072] For comparison, the weight W j The expected value θb of the posterior distribution 45a when not applying jk is expressed by equation (7).

[0073]

number

[0074] In this embodiment, as shown in equation (5), the weight Wj Using this, the expected value θb of the posterior distribution 45a jk The expected value θb of the posterior distribution 45a in this embodiment is calculated. jk represents the prior distribution 44a jk , Na j Weight W j That is, the weight W j is the parameter α of the prior distribution 44a for the observed data 41a. jk From equation (5), the weight W j The larger the parameter α jk The influence ratio of is high, and the weight W j The smaller the parameter α jk The impact rate of

[0075] Weight W j is expressed by the formula (8). In the formula (8), K j is the weight W j This is a coefficient to represent Na j is the parameter α of the prior distribution 44a in the discrete interval j. jk It is the sum of N j is the total number of data in the observation data 41a in the discrete interval j. In other words, the weight W j is Na j , N j It is expressed by:

[0076]

number

[0077] In addition, when the amount of observed data 41a is small, the conditional probability table generating unit 45 uses the observed data 41a and the parameter α jk Based on this, the expected value θb of the posterior distribution 45a is calculated using Bayesian estimation. jkIt is preferable to generate a conditional probability table M1 with each element value as follows. If a large amount of observation data 41a is obtained, the conditional probability table generation unit 45 may generate the conditional probability table M1 based only on the observation data 41a. Therefore, the conditional probability table generation unit 45 is configured to be able to select between generating the conditional probability table M1 using Bayesian estimation and generating the conditional probability table M1 using maximum likelihood estimation based on the observation data 41a.

[0078] In FIG. 9, the weight storage unit 46 stores the weight W j Store the weights W j can be set to different values ​​for each discrete interval of the factor value of the parent factor A, or can be set to the same value. As described above, from equation (5), the weight W j The larger the parameter α jk The influence ratio of is high, and the weight W j The smaller the parameter α jk For example, the weight storage unit 46 stores a plurality of types of weights W j_L ,W j_M ,W j_S Store.

[0079]

number

[0080] Weight W j The meaning of will be explained in detail with specific values. As shown in the above formula (7), the weight W j The expected value θb of the posterior distribution 45a without using jk Here, N j4 is "80", Na j When is "100", the parameter α of the prior distribution 44a jk Compare the case of {1, 2, 2, 5} with the case of {100, 200, 200, 500}. The parameter α jkThe ratio of each is {1:2:2:5}.

[0081] Parameter α of prior distribution 44a jk When {1,2,2,5} (α j4 = 5) the expected value θb of the posterior distribution 45a j4 As shown in formula (10), the result is "0.7727". The calculation formula is "(5+80) / (10+100)". The parameter α of the prior distribution 44a jk When {100,200,200,500} (α j4 = 500) j4 As shown in formula (11), the result is "0.5272". The calculation formula is "(500+80) / (1000+100)".

[0082]

number

[0083]

number

[0084] Thus, the parameter α of the prior distribution 44a jk Even if the ratio of the two is the same {1:2:2:5}, the parameter α jk Depending on the magnitude of the absolute value of j4 As shown in equation (10), the parameter α jk The smaller the absolute value of , the greater the expected value θb j4 is the data frequency β of observation data 41a jk The maximum likelihood estimator φ for j4 On the other hand, as shown in equation (11), the parameter α jk The larger the absolute value of , the greater the expected value θb j4 is the expected value θa of the prior distribution 44a j4The value will be close to 0.5 (=5 / 10).

[0085] And the weight W j can be regarded as the confidence of the prior distribution 44a. In other words, when the confidence of the prior distribution 44a is high, the weight W j It is better to set it to a large value. In this case, the expected value θb of the posterior distribution 45a j4 is the expected value θa of the prior distribution 44a j4 On the other hand, when the confidence of the prior distribution 44a is low, the weight W j It is advisable to set to a small value. In this case, the expected value θb of the posterior distribution 45a j4 is the data frequency β of observation data 41a jk The maximum likelihood estimator φ for j4 In this way, the weight W j By adjusting the expected value θb of the posterior distribution 45a, j4 can be set to a desired value.

[0086] 9, the knowledge model storage unit 22 stores the conditional probability table M1 generated by the conditional probability table generating unit 45 as a knowledge model M representing the relationship between a parent factor A and a child factor D.

[0087] The model edit processing unit 47 is configured to receive input from an operator and edit the conditional probability table M1 constituting the knowledge model M stored in the knowledge model storage unit 22 in response to the operator's input. The model edit processing unit 47 makes it possible to adjust each element value of the conditional probability table M1 shown in the upper part of Fig. 11 by the operator's input. The model edit processing unit 47 is also capable of creating the conditional probability table M1 from scratch by the operator's input.

[0088] Here, in addition to the above processing, when the observation data acquisition unit 41 acquires new observation data 41a, the conditional probability table generation unit 45 can generate a new conditional probability table M1 and update the knowledge model M stored in the knowledge model storage unit 22.

[0089] In this case, the prior distribution storage unit 44 stores each element value Nb obtained in the process of generating the posterior distribution 45a by the conditional probability table generating unit 45. jk (expressed in equation (6)) is the parameter α of the following prior distribution 44a jk Then, the conditional probability table generating unit 45 stores the parameter α jk Based on the newly calculated posterior distribution 45a, the conditional probability table generating unit 45 can calculate a new posterior distribution 45a based on the observation data 41a newly acquired by the observation data acquiring unit 41. jk It is possible to generate a conditional probability table M1 with each element value being:

[0090] 12 to 14, a procedure for updating the conditional probability table M1 constituting the knowledge model M will be described in detail. As shown in FIG. 12, when no observed data 41a exists, the parameter α jk In other words, the element values ​​of the conditional probability table M1 are the expected values ​​θa jk At this point, the conditional probability table M1 constituting the knowledge model M stored in the knowledge model storage unit 22 is a parameter α jk The conditional probability table M1 (44b) is defined by the conditional probability table M1 (44b) represented only by

[0091] Next, assume that observed data 41a is acquired. As shown in FIG. 13, the conditional probability table generating unit 45 calculates the parameter α jk Based on the observation data 41a, the parameter γ of the posterior distribution 45a is calculated using Bayesian estimation. jk Calculate the expected value θb of the posterior distribution 45a jk Then, the conditional probability table generating unit 45 generates the expected value θb jk In this way, the conditional probability table generating unit 45 updates the conditional probability table M1 constituting the knowledge model M stored in the knowledge model storage unit 22.

[0092] Next, the conditional probability table generating unit 45 calculates each element value Nb jk The parameter α of the new prior distribution 44a is stored in the prior distribution storage unit 44. jk That is, as shown in FIG. 14, the parameter α jk However, each element value Nb obtained in the previous process of generating the posterior distribution 45a jk Then, assume that new observation data 41a is acquired. The conditional probability table generating unit 45 updates the parameter α jk and the new observation data 41a, a new posterior distribution 45a is calculated, and the expected value θb' of the new posterior distribution 45a is calculated. jk Then, the conditional probability table generating unit 45 generates the expected value θb' of the new posterior distribution 45a. jk Then, the conditional probability table generating unit 45 again updates the conditional probability table M1 constituting the knowledge model M stored in the knowledge model storage unit 22.

[0093] 6. Configuration of specific cutting resistance output unit 23 4, the specific cutting resistance output unit 23 performs forward probabilistic inference using a conditional probability table M1 constituting the knowledge model M, thereby outputting an estimated value of the specific cutting resistance. The specific cutting resistance output unit 23 also performs backward probabilistic inference using the knowledge model M, thereby outputting factor values ​​for the machining conditions of the cutting work, the specifications of the tool T, and the specifications of the workpiece W. Each process performed by the specific cutting resistance output unit 23 will be described in detail below.

[0094] The process of performing forward probability inference by the specific cutting resistance output unit 23 will be described with reference to Fig. 15. The specific cutting resistance output unit 23 acquires factor values ​​of parent factors A, B, and C (factor value acquisition step S1).

[0095] Next, the specific cutting resistance output unit 23 performs forward probability inference using the conditional probability table M1 constituting the knowledge model M to obtain probabilities P11-P18, P21-P28, P31-P38 (shown in FIG. 7) of each factor value of the child factor D corresponding to the obtained factor values ​​of the parent factors A, B, and C (probability obtaining step S2). For example, when the factor values ​​of the parent factors A, B, and C are A1, B1, and C1, the probability P11 of the factor value D1 of the child factor D, the probability P21 of the factor value D2 of the child factor D, and the probability P31 of the factor value D3 of the child factor D are obtained.

[0096] Next, the specific cutting resistance output unit 23 generates and outputs an estimate of the specific cutting resistance based on the acquired probabilities P11, P21, and P31 (output step S3). For example, the specific cutting resistance output unit 23 can set the expected value obtained from the probabilities P11, P21, and P31 of the factor values ​​D1, D2, and D3 of the child factor D as the estimate of the specific cutting resistance. The specific cutting resistance output unit 23 can also set the factor value of the child factor D with the highest probability as the estimate of the specific cutting resistance. The specific cutting resistance output unit 23 can also calculate the estimate of the specific cutting resistance based on the factor values ​​of a predetermined number of child factors D in ascending order of probability.

[0097] In this way, the specific cutting resistance output unit 23 can output an estimated value of the specific cutting resistance corresponding to the machining conditions as the input parent factors A, B, and C, the specifications of the tool T, and the specifications of the workpiece W.

[0098] Moreover, a process of performing backward probability inference by the specific cutting resistance output unit 23 will be described with reference to Fig. 16. The specific cutting resistance output unit 23 acquires a target factor value of the child factor D, which is a target value of the specific cutting resistance (factor value acquisition step S11).

[0099] Next, the specific cutting resistance output unit 23 performs backward probability inference using the conditional probability table M1 constituting the knowledge model M (probability acquisition step S12). In the backward probability inference, the specific cutting resistance output unit 23 further uses the probability tables M2, M3, and M4 of each of the factors A, B, and C. Then, the specific cutting resistance output unit 23 can obtain the probability of each factor value of the parent factors A, B, and C corresponding to the target factor value of the specific cutting resistance using the conditional probability table M1 and each probability table M2, M3, and M4.

[0100] Next, the specific cutting resistance output unit 23 generates and outputs each factor value of the parent factors A, B, C, i.e., factor values ​​for the machining conditions of the cutting process, the specifications of the tool T, and the specifications of the workpiece W, based on the calculated probability of each factor value of the parent factors A, B, C (output step S13). In this way, the specific cutting resistance output unit 23 can output the factor values ​​of the parent factors A, B, C corresponding to the input target factor value of the specific cutting resistance, i.e., factor values ​​for the machining conditions of the cutting process, the specifications of the tool T, and the specifications of the workpiece W.

[0101] 7.Effects The storage device 2b constituting the specific cutting resistance estimation system 1 described above includes a knowledge model storage unit 22 that stores a knowledge model M. The knowledge model M includes a conditional probability table M1 in which at least the machining conditions of the cutting process, the specifications of the tool T, and the specifications of the workpiece W are defined as parent factors A, B, and C, the specific cutting resistance is defined as a child factor D, and the probabilities P11 to P18, P21 to P28, and P31 to P38 corresponding to the factor value of the child factor D when the factor values ​​of the parent factors A, B, and C are set as conditions are defined.

[0102] The processor 2a constituting the specific cutting resistance estimation system 1 is provided with a specific cutting resistance output unit 23 that, when it obtains the factor values ​​of the parent factors A, B, and C, performs forward probabilistic inference using the conditional probability table M1 constituting the knowledge model M, thereby outputting an estimated value of the specific cutting resistance as the child factor D.

[0103] In this way, in the specific cutting resistance estimation system 1, the conditional probability table M1 is configured as the knowledge model M. The conditional probability table M1 is defined by the probability of the factor value of the child factor D when the factor values ​​of the parent factors A, B, and C are set as conditions. In other words, the conditional probability table M1 does not directly relate the factor values ​​of the parent factors A, B, and C and the child factor D to each other, but is expressed using probabilities P11 to P18, P21 to P28, and P31 to P38. Therefore, the relationship between the factors can be flexibly defined.

[0104] Parent factors A, B, and C include the machining conditions of the cutting process, the specifications of the tool T, and the specifications of the workpiece W. Child factor D is the specific cutting resistance. By making the knowledge model M a conditional probability table M1, even if there are many types of parent factors A, B, and C, it becomes easy to define the probabilities P11 to P18, P21 to P28, and P31 to P38 that correspond to the factor value of child factor D, which is the specific cutting resistance, when the factor values ​​of these multiple parent factors A, B, and C are used as conditions.

[0105] Then, when the specific cutting resistance output unit 23 acquires factor values ​​relating to the machining conditions, the specifications of the tool T, and the specifications of the workpiece W as parent factors A, B, and C, it can obtain an estimated value of the specific cutting resistance as a child factor D by performing forward probabilistic inference using the conditional probability table M1 that constitutes the knowledge model M.

[0106] Therefore, according to the specific cutting resistance estimation system 1, by constructing and utilizing a knowledge model M regarding the specific cutting resistance, it is possible to estimate the specific cutting resistance with high accuracy.

[0107] Moreover, the cutting resistance F occurring in cutting is represented by a principal force F1, a thrust force F2, and a feed force F3. In the specific cutting resistance estimation system 1, the specific cutting resistance corresponds to a value obtained by dividing the thrust force F2 constituting the cutting resistance F by the cutting cross-sectional area. In this way, by setting the specific cutting resistance to a value related to the thrust force F2, which has a large effect on chatter, the specific cutting resistance related to chatter can be derived with high accuracy.

[0108] In addition, in the specific cutting resistance estimation system 1, when the specific cutting resistance output unit 23 acquires the target factor value of the child factor D, which is the target value of the specific cutting resistance, it performs backward probabilistic inference using the conditional probability table M1 constituting the knowledge model M, thereby outputting factor values ​​for the machining conditions of the cutting work, the specifications of the tool T, and the specifications of the workpiece W, which are parent factors A, B, and C. This makes it possible to derive the machining conditions, the specifications of the tool T, and the specifications of the workpiece W that are the target value of the specific cutting resistance.

[0109] In addition, the conditional probability table M1 constituting the knowledge model M is generated using Bayesian estimation. That is, the storage device 2b constituting the specific cutting resistance estimation system 1 stores the parameter α jk The processor 2a constituting the specific cutting resistance estimation system 1 further includes an observation data acquisition unit 41 for acquiring observation data 41a obtained by performing an actual cutting process or a simulation of the cutting process, the observation data 41a being about the factor values ​​of the parent factors A, B, and C and the factor value of the child factor D, and a parameter α jk Based on the observation data 41a, a posterior distribution 45a is calculated using Bayesian estimation, and the expected value θb jk and a conditional probability table generating unit 45 that generates a conditional probability table M1 having element values ​​of the above. The knowledge model storage unit 22 stores the knowledge model M configured by the conditional probability table M1 generated by the conditional probability table generating unit 45.

[0110] In this way, by using Bayesian estimation in generating the conditional probability table M1, the prior distribution 44a and the observed data 41a can be effectively used. In other words, by combining the prior distribution 44a as previously set knowledge with the observed data 41a, the conditional probability table M1 can be made highly accurate.

[0111] In addition, the storage device 2b constituting the specific cutting resistance estimation system 1 stores the parameter α jk Weight W, which represents the influence ratio of j The conditional probability table generating unit 45 is provided with a weight storage unit 46 for storing the observed data 41a and the weights W j Parameter α of the prior distribution 44a jk Based on this, a posterior distribution 45a is calculated using Bayesian estimation, and the expected value θb jk A conditional probability table M1 is generated with each element value being:

[0112] When generating the conditional probability table M1 using Bayesian estimation, depending on the accuracy of the prior distribution 44a, it may be impossible to obtain a high accuracy of the conditional probability table M1. Therefore, the conditional probability table generating unit 45 uses the weights W j Using the expected value θb of the posterior distribution 45a jk By calculating the above, a conditional probability table M1 according to the confidence of the prior distribution 44a can be generated.

[0113] Then, when a large amount of observed data 41a is obtained by the conditional probability table generation unit 45 generating the conditional probability table M1 using Bayesian estimation, the expected value θb jk In addition, when the conditional probability table generating unit 45 sequentially updates the conditional probability table M1 using Bayesian estimation, it is possible to secure a large amount of the observation data 41a, so that the expected value θb jk Therefore, when a large amount of observed data 41a can be secured, a highly accurate conditional probability table M1 can be generated by increasing the influence ratio of highly reliable observed data 41a.

[0114] And the weight W j The weights W can be set to different values ​​for each factor value of the parent factors A, B, and C, i.e., for each discrete interval j of the parent factors A, B, and C. jrepresents the confidence of the prior distribution 44a. The confidence may differ depending on the discrete interval j of the parent factors A, B, and C. For example, in a discrete interval j that the operator has had a lot of experience with, the prior distribution 44a can be determined with a high confidence. However, in other cases, the prior distribution 44a may not be determined with a high confidence. In such cases, the weight W j For each discrete interval j, an appropriate conditional probability table M1 can be generated.

[0115] Moreover, the processor 2a constituting the specific cutting resistance estimation system 1 includes a stability limit diagram generating unit 13 that generates a stability limit diagram in cutting processing by using the estimated value of the specific cutting resistance output by the specific cutting resistance output unit 23 and the dynamic characteristics (corresponding to the compliance transfer function) of at least one of the tool T and the workpiece W. Since the estimated value of the specific cutting resistance can be obtained with high accuracy, the stability limit diagram can also be obtained with high accuracy.

[0116] Furthermore, the processor 2a constituting the specific cutting resistance estimation system 1 includes a recommended condition determination unit 14 that determines recommended values ​​of the machining conditions using a stability limit diagram, and an instruction unit 15 that instructs the recommended values ​​of the machining conditions determined by the recommended condition determination unit 14. In this way, by instructing the recommended values ​​of the machining conditions, the operator or manager can grasp the appropriate machining conditions.

[0117] Furthermore, the processing devices 4, 5 may be configured to be able to communicate with the processor 2a, acquire recommended values ​​of the processing conditions taught by the teaching unit 15, and perform cutting processing applying the acquired recommended values ​​of the processing conditions. In this way, the processing devices 4, 5 perform cutting processing autonomously using the taught processing conditions, thereby realizing the desired cutting processing without human intervention.

[0118] (Embodiment 2) The configuration of a specific cutting resistance estimation system 100 according to the second embodiment will be described with reference to Fig. 17. In Fig. 17, the same components as those in the first embodiment are denoted by the same reference numerals.

[0119] 17, the specific cutting resistance estimation system 100 is configured by incorporating the processor 2a and storage device 2b constituting the management device 2 in the embodiment 1 into each of the processing devices 4 and 5. That is, each of the processing devices 4 and 5 has the functions of the processor 2a and storage device 2b in the embodiment 1 as a single unit. The processing devices 4 and 5 are configured to acquire the recommended values ​​of the processing conditions taught by the teaching unit 15 of the processor 2a, and perform cutting processing applying the acquired recommended values ​​of the processing conditions.

[0120] The processing devices 4 and 5 applicable to this embodiment are various processing devices that perform cutting processing, such as a lathe, a machining center, a milling machine, a gear processing machine, and a boring machine.

[0121] (others) The specific cutting resistance estimation system 1,100 uses the specific cutting resistance to generate the stability limit diagram. In addition, the specific cutting resistance estimation system 1,100 can also use the specific cutting resistance to calculate physical quantities related to the tool T, such as the force acting on the tool T (equal to the cutting force), the deflection of the tool T, and the amount of wear of the tool T. Furthermore, the specific cutting resistance estimation system 1,100 can also use the specific cutting resistance to calculate the energy consumed by the processing devices 4 and 5.

[0122] In the specific cutting resistance estimation system 100 of the second embodiment, all the functions of the processor 2a and the storage device 2b in the first embodiment are incorporated into the machining devices 4 and 5, but some of the functions of the processor 2a and the storage device 2b may be incorporated into the machining devices 4 and 5. In this case, the system is configured with a management device 2 and the machining devices 4 and 5 that configure a network, and the management device 2 has some of the functions of the processor 2a and the storage device 2b, and the machining devices 4 and 5 incorporate the remaining some of the functions. [Explanation of symbols]

[0123] 1,100 Specific cutting resistance estimation system 2a Processor 2b Storage device 22 Knowledge model storage 23 Specific cutting resistance output section M Knowledge Model M1 Conditional Probability Table A,B,C parent factor D child factor (specific cutting force) P11~P18, P21~P28, P31~P38 Probability T-tool W Workpiece

Claims

1. A specific cutting resistance estimation system that includes a processor and a storage device and utilizes a knowledge model in the cutting of a workpiece by a tool, wherein the storage device includes a knowledge model storage unit that stores the knowledge model, the knowledge model includes at least a conditional probability table in which machining conditions of the cutting, specifications of the tool, and specifications of the workpiece are used as parent factors, and the specific cutting resistance is used as a child factor, and a probability corresponding to the factor value of the child factor is defined when the factor values of the parent factors are used as conditions, and the processor includes a specific cutting resistance output unit that outputs an estimated value of the specific cutting resistance as the child factor by performing forward probability inference using the conditional probability table that constitutes the knowledge model when obtaining the factor values of the parent factors. The specific cutting resistance estimation system utilizes a knowledge model.

2. The cutting resistance generated during the cutting is represented by a main cutting force, a back cutting force, and a feed cutting force, and the specific cutting resistance corresponds to a value obtained by dividing the back cutting force that constitutes the cutting resistance by the cutting cross-sectional area. The specific cutting resistance estimation system utilizes a knowledge model according to Claim 1.

3. When obtaining a target factor value of the child factor, which is the target value of the specific cutting resistance, the specific cutting resistance output unit performs reverse probability inference using the conditional probability table that constitutes the knowledge model, and outputs factor values for the machining conditions of the cutting, the specifications of the tool, and the specifications of the workpiece as the parent factors. The specific cutting resistance estimation system utilizes a knowledge model according to Claim 1 or 2.

4. The storage device further includes a prior distribution storage unit that stores parameters of a prior distribution preset for determining the conditional probability table, and the processor further includes an observation data acquisition unit that acquires observation data obtained by performing actual cutting or a simulation of the cutting, and the observation data for the factor values of the parent factors and the factor values of the child factor, and a conditional probability table generation unit that calculates a posterior distribution using Bayesian estimation based on the parameters of the prior distribution and the observation data, and generates the conditional probability table having the expected value of the posterior distribution as each element value, and is provided with The knowledge model storage unit stores the knowledge model constituted by the conditional probability table generated by the conditional probability table generation unit. The specific cutting force estimation system using the knowledge model according to any one of claims 1 to 3.

5. The storage device further includes a weight storage unit that stores a weight representing an influence ratio of parameters of the prior distribution with respect to the observation data in the Bayesian estimation. The conditional probability table generation unit calculates the posterior distribution using Bayesian estimation based on the observation data and the parameters of the prior distribution taking into account the weight, and generates the conditional probability table having the expected value of the posterior distribution as each element value. The specific cutting force estimation system using the knowledge model according to claim 4.

6. The processor further includes a stability limit diagram generation unit that generates a stability limit diagram in the cutting process using the estimated value of the specific cutting force output by the specific cutting force output unit and at least one of the dynamic characteristics of the tool and the workpiece. The specific cutting force estimation system using the knowledge model according to any one of claims 1 to 5.

7. The processor further includes a recommended condition determination unit that determines a recommended value of the machining condition using the stability limit diagram, and a teaching unit that teaches the recommended value of the machining condition determined by the recommended condition determination unit. The specific cutting force estimation system using the knowledge model according to claim 6.

8. Furthermore, it includes a machining device configured to be communicable with the processor or configured by incorporating the processor and the storage device. The machining device is configured to acquire the recommended value of the machining condition taught by the teaching unit and perform the cutting process applying the acquired recommended value of the machining condition. The specific cutting force estimation system using the knowledge model according to claim 7.

Images