DE112018007741T5 - MACHINE LEARNING DEVICE, DEVICE FOR GENERATING PROGRAMS FOR NUMERICALLY CONTROLLED MACHINING AND MACHINE LEARNING PROCEDURES - Google Patents

Abstract

A machine learning device (50) comprising a state observation unit (51) which includes, as state variables, machining shape data, information about a machining end shape of a machined object and information about a material of a blank, a method for generating a program for numerically controlled machining for automatically generating a Program for a numerically controlled machining, which comprises a plurality of machining processes for machining a workpiece into the machined object, and an editing of a first program for a numerically controlled machining observed with reference to the method for generating a program for a numerically controlled machining and a learning unit (52) which learns the method for generating a program for numerically controlled processing on the basis of a data record generated on the basis of the state variables.

Description

area

The present invention relates to a machine learning apparatus, an apparatus for generating a program for numerically controlled machining, and a method for machine learning, each of which is used for learning a method for generating a program for numerically controlled machining, which is used in the automatic generation of a program for a numerically controlled machining is to be used for the numerical control of a machine tool.

background

In the field of numerically controlled (NC) machine tools, multi-function machines have recently been increasingly used in which different types of tools are interchangeably arranged on the tool rest in order to enable different types of machining processes such as turning, drilling and milling that allow automatic machining for Making it possible to manufacture products with a wide variety of complex shapes.

Patent Document 1 discloses an apparatus for generating a program for numerically controlled machining, which stores in a memory unit the change in the machining conditions made by a user with respect to an automatically generated machining program, so that the changed machining conditions can be used again when another user would like to generate a similar machining program at a later point in time.

List of citations

Patent literature

Patent Document 1: Japanese Patent Laid-Open No. 2002-132313

Brief description

Technical problem

However, the apparatus for generating programs for numerically controlled machining according to the above-mentioned Patent Document 1 only enables the reuse of machining conditions such as the rotational speed of the main shaft for rotating the blank to be machined, the relative cutting feed rate between the turning tool and the blank to be machined, and the chamfer, although it does not support the automatic generation of a machining program with regard to the machining sequence and machining method.

The present invention has been made in view of the above, and an object of the present invention is to provide a machine learning apparatus which can automatically learn a method of generating a program for numerically controlled machining in terms of machining order and / or machining method enables the automatic generation of a machining program.

Solution to the problem

In order to solve the above problem and to achieve the object, a machine learning device according to the present invention comprises: a state observation unit to, as state variables, machining shape data, which includes information about a machining end shape of a machined object and information about a blank material, a method of generating a program for a numerically controlled machining for automatically generating a program for a numerically controlled machining, which comprises a plurality of machining operations for machining a workpiece into the machined object, and to observe an editing of a first program for a numerically controlled machining, which with reference to the Method for generating a program for numerically controlled machining was generated; and a learning unit for learning the method of generating a program for numerically controlled machining on the basis of a data set generated on the basis of the state variables.

Advantageous Effects of the Invention

The present invention offers the advantage that a machine learning device is provided which can automatically learn a method for generating a program for numerically controlled machining which enables an automatic generation of a machining program with regard to the machining sequence and / or machining method.

Figure list

• 1 Fig. 13 is a block diagram showing an example of the configuration of a numerical controller according to a first embodiment of the present invention.
• 2 FIG. 13 is a diagram for illustrating the principle of an action value table.
• 3 FIG. 13 is a diagram showing an example of machining knowledge and editing-integrated machining knowledge stored in the function update unit of the machine learning device according to the first embodiment of the present invention.
• 4th FIG. 10 shows a flow chart to illustrate the sequence of a method for updating an action value function Q (s, a) in the first embodiment of the present invention.
• 5 Fig. 13 is a flowchart showing a detail of a procedure for creating an NC machining program performed in the NC machining program creating unit of the NC machining program creating apparatus according to the first embodiment of the present invention.
• 6th Fig. 13 is a perspective view showing an example of a three-dimensional model of the machining end shape represented by CAD (Computer-Aided Design) data according to the first embodiment of the present invention.
• 7th Fig. 13 is a perspective view showing an example of a blank shape surrounding the machining end shape represented by CAD data according to the first embodiment of the present invention.
• 8th Fig. 13 is a perspective view showing an example of a sectional shape generated from the machining end shape and the blank shape represented by CAD data according to the first embodiment of the present invention.
• 9 FIG. 13 is a diagram showing the blank shape, a rotary cross-sectional shape generated from the machining shape, and coordinate values according to the first embodiment of the present invention.
• 10 Fig. 13 is a diagram showing a turning cross-sectional shape in a facing step according to the first embodiment of the present invention.
• 11 Fig. 13 is a diagram showing a turning cross-sectional shape in a drilling turning step according to the first embodiment of the present invention.
• 12th Fig. 13 is a diagram showing a turning cross-sectional shape in a rod material turning machining step according to the first embodiment of the present invention.
• 13th Fig. 13 is a diagram showing a turning cross-sectional shape in a rod material turning machining step according to the first embodiment of the present invention.
• 14th Fig. 13 is a diagram showing an example of a machining start point and machining end point in the facing step according to the first embodiment of the present invention.
• 15th Fig. 13 is a diagram showing an example of a machining start point and machining end point in the drilling turning step according to the first embodiment of the present invention.
• 16 Fig. 13 is a diagram showing an example of a cutting point, machining start point, and machining end point in the rod material turning machining step according to the first embodiment of the present invention.
• 17th Fig. 13 is a diagram showing an example of a cutting point, machining start point, and machining end point in the rod material turning machining step according to the first embodiment of the present invention.
• 18th Fig. 13 is a diagram showing a turning cross-sectional shape in the drilling turning step according to the first embodiment of the present invention.
• 19th Fig. 13 is a diagram showing a turning cross-sectional shape in the facing step according to the first embodiment of the present invention.
• 20th Fig. 13 is a diagram showing a turning cross-sectional shape in the rod material turning machining step according to the first embodiment of the present invention.
• 21 Fig. 13 is a diagram showing a turning cross-sectional shape in the rod material turning machining step according to the first embodiment of the present invention.
• 22nd Fig. 13 is a diagram showing an example of a machining start point and machining end point in the drilling turning step according to the first embodiment of the present invention.
• 23 Fig. 13 is a diagram showing an example of a machining start point and machining end point in the facing step according to the first embodiment of the present invention.
• 24 Fig. 13 is a diagram showing an example of a cutting point, machining start point, and machining end point in a bar face turning step according to the first embodiment of the present invention.
• 25th Fig. 13 is a diagram showing an example of a cutting point, machining start point, and machining end point in a bar groove turning step according to the first embodiment of the present invention.
• 26th Fig. 13 is a diagram showing an example of a neural network model used by the machine learning apparatus according to the first embodiment of the present invention.
• 27 Fig. 13 is a diagram showing a hardware configuration of the NC machining program creating apparatus according to the first embodiment of the present invention.

Description of an embodiment

In the following, a machine learning device, a device for generating a program for numerically controlled machining and a method for machine learning according to an embodiment of the present invention are described in detail with reference to the figures. It should be noted that this embodiment is not intended to limit the scope of the invention.

First embodiment

1 Fig. 13 is a block diagram showing an example of a configuration of a numerical controller 1 according to a first embodiment of the present invention. As in 1 shown includes the numerical control 1 a processing unit for interactive operation 10 , a display unit 20th , an instruction input unit 30th and an NC machining program generating device 40 . The NC machining program generating device 40 has a machine learning device 50 on. The machine learning device 50 comprises a condition monitoring unit 51 and a learning unit 52 . The numerical control 1 is attached to or connected to a numerically controlled machine tool (not shown) in order to numerically control the operation of the numerically controlled machine tool with the aid of an NC machining program.

The processing unit for interactive operation 10 is an interface unit between the numerical control 1 and a user and an interface unit between the NC machining program creation device 40 and the user. The processing unit for interactive operation 10 transmits from the user via the instruction input unit 30th instruction information input to the NC machining program creating apparatus 40 . The processing unit for interactive operation 10 also shows the user via the instruction input unit 30th instruction information inputted on the display unit 20th at.

With the display unit 20th it is a display terminal such as a liquid crystal display, the display unit 20th CAD data 100 , an NC machining program and that of the user through the instruction input unit 30th instruction information entered. The display unit 20th also displays various types of information related to that from the numerical control 1 and the NC machining program generating device 40 refer to the processing carried out.

The instruction input unit 30th is realized by input devices such as a mouse and a keyboard, receives instruction information from the user, and transmits the instruction information to the interactive operation processing unit 10 .

The NC machining program generating device 40 generates an NC machining program that is used to machine a workpiece by machining into the shape of a machined object that is determined by the CAD data 100 is represented from outside the numerical control 1 into the NC machining program generating device 40 can be entered. An NC machining program is a numerical machining program that is used to operate an NC machine tool (not shown). In addition, the numerical control controls 1 the NC machine tool according to the NC machining program to perform machining of the workpiece to produce the machined object.

The NC machining program generating device 40 comprises a machining shape data input unit 41 , a machining shape data storage unit 42 , an NC machining program generating unit 43 , an NC machining program storage unit 44 , an NC machining program editing unit 45 and a decision making unit 46 . The components of the NC machining program generating device 40 are designed so that information can be sent to and received from one another.

The machining shape data input unit 41 receives the CAD data 100 which are machining shape data received from an external device outside of the numerical controller 1 into the NC machining program generating device 40 entered and transfers the CAD data 100 to the machining shape data storage unit 42 . The machining shape data includes information about the machining end shape of the machined object and information about the material of the blank. The machining end shape is a three-dimensional shape to be created by machining in order to obtain the machined object. The blank is the workpiece that is to be machined into the final machining shape, which is determined by the CAD data 100 is represented. However, the machining shape data is not based on the CAD data 100 limited. The machining shape data can be any data obtained by the NC machining program generating device 40 can be interpreted.

The machining shape data storage unit 42 stores the machining shape data received from the machining shape data input unit 41 were transferred.

The NC machining program generation unit 43 Generates an NC machining program for machining the blank into the final machining form, which is determined by the CAD data 100 is represented. Specifically, the NC machining program generating unit sets 43 on the basis of the machining shape data of the machined object, a machining sequence and a machining method with respect to a plurality of machining processes with reference to machining knowledge or editing-integrated machining knowledge that is provided via the decision-making unit 46 from the function update unit 54 was obtained. The NC machining program generation unit 43 then generates a first NC machining program based on the specified machining sequence and machining method 111 , which is the NC machining program for machining the workpiece into the machined object. Furthermore, the NC machining program generation unit links 43 the CAD data 100 , which are the machining shape data, the machining knowledge to which the first NC machining program was created 111 Reference was made, and the first NC machining program 111 with each other and transfers the CAD data 100 , the first NC machining program 111 and the machining knowledge to the NC machining program storage unit 44 and stores them in the NC machining program storage unit 44 from.

Machining knowledge is a set of determination conditions that are used to determine the sequence of machining the blank and / or the method of machining the blank with respect to several machining operations for machining the blank into the machining machined object is referred to when the NC machining program generating unit 43 an NC machining program is automatically generated to convert a blank into the final machining shape of the machined object, which is determined by the CAD data 100 is represented to be machined. Machining knowledge is also information that specifies the method for generating an NC machining program.

More specifically, machining knowledge is a set of conditions that specifies the order of machining the blank and / or the method of machining the blank based on the dimensions of the cut shape, i.e. H. the shape that is to be machined in order to bring the blank into the machined object. The order of machining the blank is the order of the machining operations when several different machining operations are combined in order to bring the blank into the machined object. The material processing method is a processing method that includes the processing direction.

As will be described later, by editing the automatically in the NC machining program generating unit 43 generated first NC machining program 111 generated by the user an editing-integrated processing knowledge, which is a new processing knowledge that includes the editing. Accordingly, the new machining knowledge integrates and accumulates the editing that reflects the knowledge and experience of the user. Therefore, it enables an NC machining program to be generated by the NC machining program generation unit 43 with reference to the machining knowledge, the rapid automatic generation of a user-desired NC machining program that is similar to an NC machining program created by the user himself.

The NC machining program storage unit 44 saves the CAD data 100 , which is the machining shape data, the first NC machining program 111 which is the one in the NC machining program creation unit 43 NC machining program generated is, and the machining knowledge on which the generation of the first NC machining program 111 Was referred to, linked together. The NC machining program storage unit 44 also stores a second NC machining program 112 , which is an NC machining program created in the NC machining program editing unit 45 has been edited. It should be noted that the first NC machining program 111 by the second NC machining program 112 can be updated.

The NC machining program editing unit 45 reads the CAD data 100 , the first NC machining program 111 and that with the first NC machining program 111 Linked machining knowledge from the NC machining program storage unit 44 and transfers the first NC machining program 111 to the processing unit for interactive operation 10 . The NC machining program editing unit 45 edits the first NC machining program 111 based on the user via the instruction input unit 30th inputted editing instruction information and creates the second NC machining program 112 . The NC machining program editing unit 45 transfers the generated second NC machining program 112 to the NC machining program storage unit 44 and save it there. In addition, the NC machining program editing unit transmits 45 the CAD data 100 , which is the machining shape data, the first NC machining program 111 , the one with the first NC machining program 111 linked machining knowledge and the second NC machining program 112 to the condition observation unit described later 51 .

If for the first NC machining program 111 no editing operation was carried out, ie no second NC machining program was performed 112 is generated, transmits the NC machining program editing unit 45 the CAD data 100 , the first NC machining program 111 and that with the first NC machining program 111 linked processing knowledge to the condition monitoring unit 51 .

The decision-making unit 46 searches several different machining knowledge that is in the function update unit 54 are stored, and determine and refer to the CAD data 100 most suitable machining knowledge in response to a request from the NC machining program generation unit 43 . The machining knowledge needed for the CAD data 100 The most suitable is the machining knowledge that is acquired after the creation of the first NC machining program 111 among the machining knowledge that the function update unit 54 stored CAD data 100 correspond to the smallest amount of editing operations on the first NC machining program 111 has required. That means that the machining knowledge is necessary for the CAD data 100 The most suitable machining knowledge is that of the machining knowledge that includes the CAD data 100 that best reflected the knowledge and experience of the user.

The decision-making unit 46 draws from the multiple machining knowledge that is contained in the function update unit 54 stored CAD data 100 correspond to the machining knowledge with the highest action value Q and transmits this machining knowledge to the NC machining program generation unit 43 . It should be noted that when in the feature update unit 54 only one processing knowledge is stored, which is the CAD data 100 corresponds to the decision-making unit 46 determines that this is in the feature update unit 54 stored machining knowledge for the CAD data 100 is the most appropriate machining knowledge.

A machining knowledge is linked to the machined object, for example by assigning a reference number that is used together with the machined object. The machining knowledge is also, for example, with a combination of the through the CAD data 100 represented machining end shape and the material of the blank linked. This allows the decision-making unit 46 look for the machining knowledge that the CAD data 100 and from the NC machining program generating unit 43 can be specified, and further select the machining knowledge with the highest action value Q among these machining knowledge, using the reference number or the combination of information about the machining end shape and information about the material of the blank as a search condition.

The machine learning device 50 , which is the condition observation unit 51 and the learning unit 52 can use any learning algorithm. An example of an application of reinforcement learning is described below.

Reinforcement learning is performed in such a way that an agent who is supposed to take action in a particular environment observes the current state and determines which action to take. The agent learns how to obtain the greatest reward through a series of actions, with the selection of an action leading to receiving a reward in the area. The typical well-known algorithms of reinforcement learning include Q-learning and TD-learning. In Q-learning, for example, a typical update equation of the action value function Q (s, a) can be expressed by the following mathematical formula (1).
[Formula 1] $$Q\left({\mathrm{S.}}_t,a_t\right)\leftarrow Q\left({\mathrm{S.}}_t,a_t\right)+\mathrm{\alpha }\left(r_{t+1}+\mathrm{\gamma }\underset{a}{max}Q\left({\mathrm{S.}}_{t+1},a_{t+1}\right)-Q\left({\mathrm{S.}}_t,a_t\right)\right)$$

In the mathematical formula (1), s _{t means} the state at time t; at is the action performed at time t; s _{t + 1} means the state at time t + 1 or, in other words, the state that results from performing action at; and r _{t + 1} means the reward given depending on the environment that has changed as a result of the action at, that is, r _{t + 1} means the reward given by the change in state. In the mathematical formula (1), the “max” factor represents the maximum value of the action value Q in the state s _{t + 1} , ie the action value that relates to the best action a _{t + 1} . The parameter γ is the discount factor, and it is set so that it satisfies the condition 0 <γ≤1. The parameter γ is set, for example, in the range from 0.9 to 0.99 (including limits). The parameter α is the learning rate, and it is set so that it satisfies the condition 0 <α≤1. The parameter α is set, for example, in the range from 0.05 to 0.2 (including limits).

The update equation expressed by the mathematical formula (1) in Q-learning increases the action value Q at time t when the action value of the best action a _{t + 1} at time t + 1 is greater than the action value Q of at time t performed action at, but otherwise reduces the action value Q at time t. In other words, the action value function Q (s _t , a _t ) is updated in order to bring the action _{value Q of the action a t} at time t closer to the best action value at time t + 1. In this way, the best action value in a given environment is sequentially carried over to the action values in previous environments.

The machine learning device 50 according to the first embodiment is thus able to carry out reinforcement learning according to a previously known Q-learning algorithm. The action value Q, which is assigned to a combination of a specific status variable and an action "a", is not known. The machine learning device 50 selects actions “a” at random for different state variables, executes actions “a” and accumulates the rewards given as a result of actions “a” in order to update the action value function Q (s, a). It should be noted that the NC machining program generating device 40 According to the first embodiment, an “action a” corresponds to machining knowledge, ie information that specifies the method for generating an NC machining program. In addition, a "state s" corresponds to the editing of the first NC machining program 111 .

It should be noted that the action value function Q (s, a) is also referred to as an action value table. 2 FIG. 13 is a diagram for illustrating the principle of an action value table. An action value table is a data set stored in tabular form in which an arbitrary action s is linked to its action value Q. In Q-learning, a data record represents a so-called Q-table. A Q-table in the form of a two-dimensional table comprises various states “s” in rows and possible actions “a” in columns, with an action value Q being entered in each cell , which represents the evaluation value for the combination of the action "a" and the state "s". Due to the limitation of the storable number of states “s” and actions “a” when managing the action value function Q (s, a) by means of a Q table, however, the action values Q are calculated using the above mathematical formula (1) and not calculated using a table format when there are many states "s" and many actions "a".

In the first embodiment, when a Q table format is adopted, a Q table is stored for each set of machining shape data, and each Q table is a machining knowledge corresponding to the action “a” which is information indicating a method for Specify the generation of an NC machining program and the editing of the first NC machining program corresponding to the status "s" 111 both of which are linked to the machining shape data. That is, in the first embodiment, a data record is a set of data of machining shape data, machining knowledge data which is information indicating a method of generating an NC machining program, and data of editing the first NC machining program 111 acts.

The numerical control 1 has machine learning functionality to learn the method of generating an NC machining program. That means the numerical control 1 the machine learning device 50 that learns the method of generating an NC machining program. The machine learning device 50 comprises the condition monitoring unit 51 and the learning unit 52 . The learning unit 52 learns the procedure for generating an NC machining program based on the data record generated on the basis of status variables. The learning unit 52 accordingly, learns a relationship among the machining shape data, the method of generating an NC machining program, and the editing of the first NC machining program 111 according to the data record generated on the basis of status variables, whereby the method for generating an NC machining program is learned. The learning unit 52 has a reward calculation unit 53 that calculates the reward associated with learning the machining knowledge corresponding to the NC machining program generation method, and a function update unit 54 which updates the function in connection with learning the machining knowledge corresponding to the NC machining program generation method and determines the action value Q which is the evaluation value. It should be noted that the machine learning device 50 into the NC machining program generating device 40 can be integrated or installed in a cloud server that is associated with the NC machining program generating device 40 can communicate.

The condition observation unit 51 observes the CAD data as status variables 100 , with the CAD data 100 the machining shape data is the machining knowledge which is information indicating the method of creating a numerically controlled machining program and that of the NC machining program creating unit 43 during the creation of the first NC machining program 111 Reference was made, and the editing of the first NC machining program 111 . The condition observation unit 51 so observed for each set of CAD data 100 which is the machining shape data obtained from the decision making unit 46 certain machining knowledge and the editing of the first NC machining program 111 that are in the NC machining program editing unit 45 during the generation of the second NC machining program 112 was carried out.

The condition observation unit 51 relates to the editing of the first NC machining program 111 into the machining knowledge on which the generation of the first NC machining program 111 Was referred to, and thus generates an editing-integrated processing knowledge. Editing-integrated machining knowledge is a method for generating an editing-integrated program for numerically controlled machining, which results from the integration of the editing of the first NC Machining program 111 in the method for generating a numerically controlled machining program results, on which during the generation of the first NC machining program 111 Has been referred to. The editing-integrated machining knowledge takes over the reward “r” of the one with the first NC machining program 111 linked processing knowledge. The condition observation unit 51 transfers the machining knowledge to that during the generation of the first NC machining program 111 Reference was made, and the editing-integrated processing knowledge as an observation result to the reward calculation unit 53 .

When the condition observer 51 If an editing-integrated machining knowledge is generated, the reference number that is assigned to the original machining knowledge or the combination of the information about the final machining shape and the information about the material of the blank is adopted and assigned. The machining knowledge integrated into the editing, which is generated on the basis of the machining knowledge, and the machining knowledge both correspond to the final machining shape of a machined object. The editing-integrated processing knowledge corresponds accordingly to the final processing form of a machined object linked to the processing knowledge from which the editing-integrated processing knowledge was generated. Therefore, the reference number that is assigned to the machining knowledge from which the editing-integrated machining knowledge was generated, or the combination of the information about the machining end shape and the information about the material of the blank is transferred and assigned to the editing-integrated machining knowledge. This means that the editing-integrated processing knowledge is linked to the CAD data 100 that are assigned to it.

If the CAD data 100 , the first NC machining program 111 , the one with the first NC machining program 111 linked machining knowledge and the second NC machining program 112 from the NC machining program editing unit 45 can be transmitted, the condition monitoring unit 51 the editing of the first NC machining program 111 extract. In this case, the machining knowledge integrated into the editing becomes the new machining knowledge that results from the inclusion of the editing of the first NC machining program 111 in connection with the CAD data 100 in the machining knowledge results on the generation of the first NC machining program 111 Has been referred to.

On the other hand, if only the CAD data 100 , the first NC machining program 111 and that with the first NC machining program 111 linked machining knowledge from the NC machining program editing unit 45 and the second NC machining program 112 is therefore not transferred, there is no editing of the first NC machining program 111 . In this case, the machining knowledge that is integrated into the editing process is the machining knowledge itself that was used when the first NC machining program was generated 111 Has been referred to.

Accordingly, the observation of the CAD data means 100 , which is the machining shape data, the machining knowledge, which is information indicating the method of creating a program for numerical control machining and referred to by the NC machining program creating unit 43 during the creation of the first NC machining program 111 And the editing of the first NC machining program 111 The following as status variables: Editing of the first NC machining program 111 in connection with the CAD data 100 is extracted in order to obtain, as the observation result, the machining knowledge that was accessed during the generation of the first NC machining program 111 Reference was made, and the editing-integrated machining knowledge for each set of the CAD data 100 to obtain.

The learning unit 52 learns the relationship among the machining shape data, the method of creating an NC machining program, and the editing of the first NC machining program 111 based on the dataset of the CAD data 100 , which is the machining shape data, the machining knowledge, which is information indicating the method of creating a program for numerical control machining and referred to by the NC machining program creating unit 43 during the creation of the first NC machining program 111 And the editing of the first NC machining program 111 , thereby learning the NC machining program generation method. That means that the learning unit 52 the processing knowledge based on that of the condition monitoring unit 51 observed state variables. In one embodiment, the learning unit learns 52 the processing knowledge according to the principle of reinforcement learning.

The reward calculation unit 53 calculates the reward r on the basis of the editing-integrated processing knowledge that is obtained from the condition monitoring unit 51 transmitted observation result is. Specifically, the reward calculation unit calculates 53 The reward r of the machining knowledge integrated into the editing process on the basis of the machining knowledge on the basis of the generation of the first NC machining program 111 Reference was made, and the editing-integrated processing knowledge. The reward calculation unit 53 calculates the reward r of the machining knowledge integrated into the editing based on the degree of equality between the machining knowledge to that during the creation of the first NC machining program 111 Reference was made, and the editing-integrated processing knowledge, wherein the degree of equality serves as a reward allocation criterion for the allocation of the reward r. The reward calculation unit 53 links the reward r to the CAD data 100 and the editing-integrated processing knowledge and then transmits the calculated reward r together with the CAD data 100 and the editing-integrated processing knowledge to the function update unit 54 .

If the machining knowledge is based on the generation of the first NC machining program 111 Reference was made to the editing-integrated processing knowledge is identical, increases the reward calculation unit 53 the reward r of the editing-integrated processing knowledge. If the machining knowledge is based on the generation of the first NC machining program 111 Reference was made, is identical to the editing-integrated processing knowledge, the reward calculation unit 53 the editing-integrated processing knowledge, for example, a "positive" reward r, such as B. a reward r of "+1". The reward r of the machining knowledge integrated into the editing is thus calculated by adding to the reward r that was taken over from the machining knowledge during the generation of that when the first NC machining program was generated 111 Was referred to, "+1" is added. Thus, the reward r for the machining knowledge integrated in the editing becomes greater than the reward r that is adopted from the machining knowledge during the generation of the first NC machining program 111 Has been referred to.

If the machining knowledge is based on the generation of the first NC machining program 111 Reference was made from the editing-integrated processing knowledge, reduces the reward calculation unit 53 the reward r of the machining knowledge, which is based on the generation of the first NC machining program 111 Has been referred to. If the machining knowledge is based on the generation of the first NC machining program 111 Reference was made, differs from the editing-integrated processing knowledge, the reward calculation unit 53 the machining knowledge that was used when creating the first NC machining program 111 Reference was made to, for example, a "negative" reward r, such as B. a reward r of "-1". For example, if a first NC machining program 111 that is generated so that facing is performed before drilling is machined so that drilling is performed before facing decreases the reward calculation unit 53 the reward r and indicates the machining knowledge that was acquired during the generation of the first NC machining program 111 Reference was made, for example, to a reward r of "-1". The reward r of the machining knowledge, which is based on the creation of the first NC machining program 111 Was referenced, is calculated by adding “-1” to the current reward r. The reward r of the machining knowledge, which is gained during the creation of the first NC machining program 111 Reference was made, is thus decreased to a value that is under the reward r at the reference time during the generation of the first NC machining program 111 lies.

If the machining knowledge is based on the generation of the first NC machining program 111 Reference was made from the editing-integrated processing knowledge, reduces the reward calculation unit 53 In addition, as described above, the reward r for the machining knowledge, which is based on the generation of the first NC machining program 111 Is referred to, and then increases the reward r of editing-integrated processing knowledge. If the machining knowledge is based on the generation of the first NC machining program 111 Reference was made, differs from the editing-integrated processing knowledge, the reward calculation unit 53 for example, to the editing-integrated processing knowledge a “positive” reward r, such as a reward r of “+1”. The reward r of the machining knowledge integrated into the editing is thus calculated by adding to the reward r that was added during the generation of the machining knowledge to that during the generation of the first NC machining program 111 Was referenced, was adopted, "+1" is added. The reward r for the machining knowledge integrated in the editing is thus greater than the reward r that was adopted from the machining knowledge during the generation of the first NC machining program 111 Has been referred to.

The numerical value of the reward r is not limited to “+1” or “-1”. In addition, the reward can be r depending on the editing of the first NC machining program 111 , ie the editing to generate the editing-integrated machining knowledge from the machining knowledge, to the generation of the first NC machining program 111 A different value should be assigned.

The feature update unit 54 updates the function for determining the method for NC machining program generation based on that from the reward calculation unit 53 calculated reward r of the editing-integrated processing knowledge. For example, the function update unit updates 54 the action value function Q (s, a) and determines the action value Q which is the evaluation value. The feature update unit 54 links the determined action value Q with the editing-integrated processing knowledge and stores the action value Q together with the editing-integrated processing knowledge. Thus, in addition to the initially stored machining knowledge, the edit-integrated machining knowledge and the action value Q associated with the edit-integrated machining knowledge are sequentially accessed as the machining knowledge to which the first NC machining program was generated 111 Should be referenced in the functional update unit 54 saved. 3 FIG. 13 shows a representation to illustrate an example of machining knowledge and editing-integrated machining knowledge that is in the function update unit 54 the machine learning device 50 according to the first embodiment of the present invention. 3 illustrates an example of the storage of machining knowledge A. 121 , a processing knowledge B 122 , machining knowledge C 123 , an editing-integrated processing knowledge A 131 , an editing-integrated processing knowledge B 132 and an editing-integrated processing knowledge C 133 . It is pointed out that a machining knowledge storage unit can be provided separately, which stores the machining knowledge, the editing-integrated machining knowledge and the action values Q which are linked to the editing-integrated machining knowledge.

Next, a reinforcement learning method for updating the action value function Q (s, a) will be described. 4th Fig. 13 is a flowchart illustrating a procedure for updating the action value function Q (s, a) according to the first embodiment of the present invention.

At step ST110, the NC machining program generating unit reads 43 the CAD data 100 stored in the machining shape data storage unit 42 is saved and saves the CAD data 100 in a shape memory area of a memory unit (not shown) of the NC machining program generation unit 43 .

Next, the NC machining program generation unit refers 43 the CAD data in step ST120 100 corresponding processing knowledge about the decision-making unit 46 from the function update unit 54 to create an NC machining program. The decision-making unit 46 determines the machining knowledge based on the learning result of the machine learning device 50 . That means the decision-making unit 46 multiple machining knowledge searches that are in the function update unit 54 are stored, the machining knowledge with the highest action value Q from the plurality of machining knowledge that is based on the request of the NC machining program generation unit 43 relate, determine and relate and the related machining knowledge to the NC machining program generation unit 43 transmits.

The following is an example of the machining knowledge that is transferred to the NC machining program generation unit 43 be transmitted.

(Machining knowledge A 121 )

Blank material: S45C

Machining sequence: The drilling-turning step is carried out first, and then the facing step is carried out when the machining allowance for the end face is 20 mm or more. The facing step is carried out first and then the drilling and turning step if the blank surface allowance does not meet this condition.

(Machining knowledge B 122 )

Blank material: S45C

Part to be machined: The end face is defined as the part to be machined if “lateral length of the shape of the open area / longitudinal length of the shape of the open area” is “less than 1”. The outer circumference is defined as the part to be machined if “lateral length of the shape of the open area / longitudinal length of the shape of the open area” is “1 or greater”. The term “open area” refers to the part of the rotary cross-sectional shape t that is exposed on the surface of the blank.

(Machining knowledge C 123 )

Blank material: S45C

Machining method: A groove turning step is performed when the longitudinal length of the turning cross-sectional shape is 10 mm or less and the lateral length of the turning cross-sectional shape is 10 mm or less. A bar stock turning machining step is performed when either the longitudinal length of the turning cross-sectional shape or the lateral length does not meet these conditions.

Next, the NC machining program creating unit creates 43 in step ST 130, an NC machining program, that is, the first NC machining program 111 , with reference to the related machining knowledge.

Next, the NC machining program generating unit arranges 43 in step ST140, the CAD data 100 , the machining knowledge that was used when generating the first NC machining program 111 Reference was made, and the first NC machining program 111 to each other and transfers the CAD data 100 , the first NC machining program 111 and the machining knowledge to the NC machining program storage unit 44 . The NC machining program storage unit 44 then saves the CAD data 100 , the first NC machining program 111 and the machining knowledge.

Next, the NC machining program editing unit reads 45 in step ST150, the CAD data 100 , the first NC machining program 111 and that with the first NC machining program 111 Linked machining knowledge from the NC machining program storage unit 44 and transfers the first NC machining program 111 to the processing unit for interactive operation 10 . The processing unit for interactive operation 10 shows the first NC machining program 111 that is produced by the NC machining program editing unit 45 was transmitted, then on the display unit 20th and prompts the user to start the first NC machining program 111 to confirm.

When the first NC machining program 111 on the display unit 20th is displayed, the user confirms the first NC machining program 111 and gives the editing instruction information for the first NC machining program 111 into the instruction input unit 30th on. The editing instruction information is received from the instruction input unit 30th via the processing unit for interactive operation 10 to the NC machining program editing unit 45 the NC machining program generating device 40 transfer.

Next, the NC machining program editing unit receives 45 in step ST 160, the edit instruction information and edit the first NC machining program 111 based on the editing instruction information to create the second NC machining program 112 to create. The second NC machining program 112 is an NC machining program created by editing the first NC machining program 111 is generated based on the editing instruction information. The NC machining program editing unit 45 then transfers the CAD data 100 , the first NC machining program 111 , the one with the first NC machining program 111 linked machining knowledge and the second NC machining program 112 to the condition monitoring unit 51 .

Next, the state observer compares 51 the first NC machining program in step ST170 111 and the second NC machining program 112 processed by the NC machining program editing unit 45 and receives the editing of the first NC machining program 111 . In this way, the condition monitoring unit receives 51 the CAD data as status variables 100 , which are the machining shape data, the machining knowledge acquired with the first NC machining program 111 linked, and the editing of the first NC machining program 111 .

Next, the state observer takes 51 in step ST 180 the editing of the first NC machining program 111 in the machining knowledge on that when generating the first NC machining program 111 Was referred to, and thus generates the editing-integrated processing knowledge. The condition observation unit 51 transfers this with the first NC machining program 111 linked processing knowledge and the processing knowledge integrated in the editing as an observation result to the reward calculation unit 53 .

Next, the reward calculation unit compares 53 in step ST190 that with the first NC machining program 111 Linked machining knowledge and the editing-integrated machining knowledge to determine whether they are identical. If these are identical to one another, i. h in step ST 190 Yes, the procedure goes to step ST200. If these are different from one another, ie No in step ST 190, the procedure continues with step ST210.

In step ST200, the reward calculation unit calculates 53 the reward r that increases the reward r of the edit-integrated machining knowledge to cause an increase in the action value Q for the reward r of the edit-integrated machining knowledge, and adds the calculated reward r to the reward r of the edit-integrated machining knowledge adopted from the edit knowledge, the one with the first NC machining program 111 is linked. This increases the reward r of the machining knowledge incorporated in the editing process, so that it is greater than the reward r that was acquired from the machining knowledge that was acquired with the first NC machining program 111 is linked. This increases the action value Q of the editing-integrated processing knowledge.

If that's with the first NC machining program 111 linked machining knowledge is identical to the machining knowledge integrated into the editing, this is the case with the first NC machining program 111 linked processing knowledge itself the editing-integrated processing knowledge. Since no editing of the first NC machining program 111 has been carried out, it is assumed that the editing-integrated processing knowledge is high-quality editing-integrated processing knowledge. The one with the first NC machining program 111 Linked processing knowledge should therefore be high-quality processing knowledge.

For example, if the editing-integrated machining knowledge is as follows, that with the first NC machining program 111 linked machining knowledge are all identical to the corresponding edit-integrated machining knowledge, and accordingly the rewards r of the edit-integrated machining knowledge are increased in step ST200. The editing-integrated processing knowledge A 121a is editing-integrated processing knowledge which is based on the processing knowledge A 121 is produced. The editing-integrated processing knowledge B 121b is editing-integrated processing knowledge, which is based on the processing knowledge B 122 is produced. The editing-integrated processing knowledge C 121c is an editing-integrated processing knowledge, which is based on the processing knowledge C 123 is produced.

(Editing-integrated processing knowledge A 121a)

Blank material: S45C

Machining sequence: The drilling-turning step is carried out first, and then the facing step is carried out if the blank surface allowance is 20 mm or more. The facing step is carried out first and then the drilling and turning step if the blank surface allowance does not meet this condition.

(Editing-integrated processing knowledge B 121b)

Blank material: S45C

Part to be machined: The end face is defined as the part to be machined if “lateral length of the shape of the open area / longitudinal length of the shape of the open area” is “less than 1”. The outer circumference is defined as the part to be machined if “lateral length of the shape of the open area / longitudinal length of the shape of the open area” is “1 or greater”. The term “open area” refers to the area of a rotary cross-sectional shape t exposed on the surface of the blank.

(Editing-integrated processing knowledge C 121c)

Blank material: S45C

Machining method: A groove turning step is performed when the longitudinal length of the turning cross-sectional shape is 10 mm or less and the lateral length of the turning cross-sectional shape is 10 mm or less. A bar stock turning machining step is performed when either the longitudinal length or the lateral length of the turning cross-sectional shape does not meet these conditions.

Meanwhile, the reward calculating unit calculates 53 in step ST210, the reward r that decreases the machining knowledge reward r to cause decrease in the action value Q for the machining knowledge reward r, and adds the calculated reward r to the machining knowledge reward r. This lowers the reward r of the machining knowledge so that it is smaller than the reward r at the reference time during the generation of the first NC machining program 111 . This decreases the action value Q of the machining knowledge.

For example, if the machining knowledge integrated into the editing is as follows, that with the first NC machining program is different 111 linked machining knowledge are all of the corresponding edit-integrated machining knowledge, and accordingly, the rewards r of the machining knowledge are decreased in step ST210. The editing-integrated processing knowledge A 131 is an editing-integrated processing knowledge, which is based on the processing knowledge A 121 is produced. The editing-integrated processing knowledge B 132 is an editing-integrated processing knowledge, which is based on the processing knowledge B 122 is produced. The editing-integrated processing knowledge C 133 is an editing-integrated processing knowledge, which is based on the processing knowledge C 123 is produced.

(Editing-integrated processing knowledge A 131 )

Blank material: S45C

Machining sequence: Drill turning is performed first, followed by facing step when the blank surface allowance is 10 mm or more. The facing step is carried out first and then the drilling and turning step if the blank surface allowance does not meet this condition.

(Editing-integrated processing knowledge B 132 )

Blank material: S45C

Part to be machined: The end face is determined as the part to be machined when the “lateral length of the shape of the open area / longitudinal length of the shape of the open area” is “1.0 or less”. The outer circumference is defined as the part to be machined if the “lateral length of the shape of the open area / longitudinal length of the shape of the open area” is “greater than 1.0”.

(Editing-integrated processing knowledge C 133 )

Blank material: S45C

Machining method: A groove turning step is performed when the longitudinal length of the turning cross-sectional shape is 10 mm or less and the lateral length of the turning cross-sectional shape is 20 mm or less. A bar stock turning machining step is performed when either the longitudinal length or the lateral length of the turning cross-sectional shape does not meet these conditions.

If that turns out to be the case with the first NC machining program 111 The editing of the first NC machining program distinguishes linked machining knowledge from the editing-integrated machining knowledge 111 carried out. It is therefore assumed that the first NC machining program 111 linked processing knowledge is of lower quality than the editing-integrated processing knowledge.

In contrast, it is assumed that the machining knowledge integrated into the editing process is of a higher quality than that with the first NC machining program 111 linked machining knowledge. Therefore, the reward calculation unit calculates 53 after step ST210, in step ST200, the reward r that the Reward r of the edit-integrated machining knowledge increased to cause an increase in the action value Q for the reward r of the edit-integrated machining knowledge, and adds the calculated reward r to the reward r of the edit-integrated machining knowledge obtained from that with the first NC machining program 111 linked processing knowledge has been adopted. As a result, the reward r for the machining knowledge integrated in the editing increases to a value which is greater than the reward r that was taken over from the machining knowledge that was obtained with the first NC machining program 111 is linked. This increases the action value Q of the editing-integrated processing knowledge.

Next, the function update unit updates 54 in step ST220 the action value function Q (s, a) based on the reward r of the editing-integrated machining knowledge calculated in step ST200 or in step ST210 and determines the action values Q, which are the evaluation values of the Machining program 111 linked machining knowledge and the editing-integrated machining knowledge. The feature update unit 54 assigns the action value Q determined for the editing-integrated processing knowledge to the editing-integrated processing knowledge and stores the action value Q as processing knowledge together with the editing-integrated processing knowledge.

The feature update unit 54 also links the action value Q, which is for the with the first NC machining program 111 Linked machining knowledge was determined with this machining knowledge and stores the action value Q as machining knowledge together with the machining knowledge. In this context, the function update unit stores 54 So the one with the first NC machining program 111 Linked machining knowledge, the machining knowledge whose action value Q was determined in step ST220, and the editing-integrated machining knowledge whose action value Q was determined in step ST220, each having a different action value Q than the CAD data 100 corresponding processing knowledge. This allows the decision-making unit 46 in determining the the CAD data 100 corresponding machining knowledge, the editing-integrated machining knowledge, which is machining knowledge with a relatively high action value Q, determine as the most suitable machining knowledge that appropriately reflects the knowledge and experience of the user.

It should be noted that the processing knowledge that is in the function update unit 54 is stored and that during the generation of the first NC machining program 111 Reference was made, in this case with the one with the first NC machining program 111 linked machining knowledge can be updated, from which the action value Q was determined in step ST220.

The procedure from step ST110 to step ST220 is then carried out in connection with a further set of newly read-in CAD data 100 performed iteratively. When generating a new machining knowledge integrated into the editing process, associated information such as the reference number of the original machining knowledge, the information about the final machining shape and the information about the material of the blank and the reward r are adopted.

It should be noted that in iterating the procedure from step ST110 to step ST220, that by the decision making unit 46 certain editing-integrated machining knowledge corresponds to the machining knowledge of step ST120 described above.

In addition, in the function update unit 54 a high-quality editing-integrated processing knowledge with an increased action value Q is stored in the procedure of steps ST200 and step ST220. Accordingly, step ST210 can be omitted.

Next, referring to FIG 5 a detail of the NC machining program creation process in the NC machining program creation unit 43 described in detail. 5 Fig. 13 is a flowchart showing a detail of the procedure of the NC machining program creation method in the NC machining program creation unit 43 the NC machining program generating device 40 illustrated in accordance with the first embodiment of the present invention. 6th Fig. 13 is a perspective view showing an example of one by the CAD data 100 represented three-dimensional model of the machining end shape according to the first embodiment of the present invention. 7th Fig. 13 is a perspective view illustrating an example of a blank shape enveloping the machining end shape identified by the CAD data 100 represents according to the first embodiment of the present invention. 8th FIG. 13 is a perspective view illustrating an example of a sectional shape obtained from FIG. 4 by the CAD data 100 machining end shape represented and the blank shape is produced according to the first embodiment of the present invention. 9 Fig. 13 is a diagram illustrating a blank shape S, a turning cross-sectional shape generated from the machining shape, and coordinate values according to the first embodiment of the present invention. The following description will be given in connection with the creation of an NC machining program for turning in which a blank is machined while it is rotating.

In step ST310 of the NC machining program generation process, the NC machining program generation unit reads 43 the CAD data 100 stored in the machining shape data storage unit 42 are stored creates a three-dimensional machining end shape defined by the in 6th CAD data shown 100 is represented, and positions the machining end shape in the program origin, which is the machining origin in the program coordinate system. The program coordinate system is defined in such a way that the Z axis of the XYZ axes is aligned with an axis of rotation SG, which is the central axis during rotation.

In order to position the machining end shape in the program origin, the NC machining program generation unit positions 43 the machining end shape by moving and rotating the machining end shape to align the rotational axis of the cylindrical surface or the rotational axis of the conical surface having the largest diameter in the machining end shape with the Z-axis; and moves the machining end shape to position its end face in the Z-axis direction at the program origin. This means that the NC machining program generating unit 43 a cylindrical surface or a conical surface, the central axis of rotation of which is aligned with the axis of rotation SG, from the CAD data as the end face of the rotation 100 extracted. In this way, the NC machining program generating unit generates 43 from the CAD data 100 a three-dimensional machining end shape with a central axis of rotation, which is the axis of rotation.

Next, the NC machining program creating unit creates 43 in step ST320, a three-dimensional cylindrical shape enveloping the machining end shape positioned at the program origin, as shown in FIG 7th is shown. The NC machining program generation unit 43 then positions the generated cylindrical shape as a blank shape in the program coordinate system similar to the case of the final machining shape. The term “blank shape” refers to the shape of the blank that is to be turned into the final machining shape. This means that the NC machining program generating unit 43 A three-dimensional blank shape is generated based on the final machining shape from the CAD data 100 was generated.

The dimensions of the cylindrical shape enveloping the machining end shape can be determined from the maximum value and the minimum value of the machining end shape in the X-axis direction, Y-axis direction, and Z-axis direction. It should be noted that, with regard to the end face turning machining, the dimensions of the blank shape in the Z-axis direction in accordance with the machining end shape are each 2 to 3 mm larger than the corresponding dimension of the machining end shape. In addition, the end face of the blank shape is positioned in the direction of the Z-axis in such a way that it protrudes from the machining end shape by 2 to 3 mm from the program origin. It is assumed here that the blank shape has a blank outer diameter of 150 mm and a blank length of 120 mm, and the blank material is S45C.

Next, the NC machining program creating unit creates 43 in step ST330 as in 8th shows a sectional shape, that is, the shape to be cut from the blank shape by turning, from the machining end shape and from the blank shape. 8th shows the final machining shape in the form of dashed lines. The sectional shape can be obtained by subtraction, the volume model of the final machining shape being subtracted from the volume model of the blank shape.

Next, the NC machining program creating unit creates 43 in step ST340, a rotary cross-sectional shape t that is the cross-sectional shape of the sectional shape as shown in FIG 9 is shown. As in 9 the rotary cross-sectional shape t is shown as a half rotary cross-sectional shape which is obtained by projecting the sectional shape onto the + XZ plane, which is part of the XZ plane that is limited to the + X area. The rotary cross-sectional shape t can be obtained by obtaining the cross-sectional shape through a multiplication operation of a cross-sectional model formed from the XZ plane and a volume model formed from the machining shape and limiting this shape to the range of X≥0. In this way, a rotary cross-sectional shape t1 and a rotary cross-sectional shape t2, which are the rotary cross-sectional shape t, are generated as in FIG 9 is shown.

Next, the NC machining program generating unit is given 43 the CAD data in step ST350 100 corresponding processing knowledge about the decision-making unit 46 from the function update unit 54 and refers to this processing knowledge. The NC machining program generation unit 43 receives, for example, the processing knowledge A described above 121 , the machining knowledge B 122 and the machining knowledge C 123 that are in the feature update unit 54 are stored, and refers to this machining knowledge.

The NC machining program generation unit 43 receives, for example, the reference number for the CAD data 100 and that on the CAD data 100 based processing knowledge is assigned jointly, and transmits the reference number to the decision-making unit 46 . This allows the decision-making unit 46 using the reference number as a search condition for that from the NC machining program generating unit 43 search for the specified desired machining knowledge.

As used herein, “facing step” refers to a step of performing end face turning processing in which the protruding area on the end face of the blank is removed using an end face turning machining tool. The term “drilling and turning step” refers to a step for carrying out a drilling and turning operation in which a rotary drill is used to drill along the central axis into a radial central section of the blank. The term “rod material turning machining step” refers to a step for carrying out a rod turning machining in which an outer circumference, an inner circumference, an end face or a rear side of a round rod is turned with a turning tool. The term “groove turning step” refers to a step for carrying out a groove formation by turning, in which an outer circumference, an inner circumference, an end face or a rear face of a round bar is profiled with the aid of a lathe chisel. The term “blank surface allowance” refers to the machining allowance for end face turning.

Next, the NC machining program generating unit divides 43 in step ST360, the turning cross-sectional shape t of the sectional shape according to the obtained machining knowledge A 121 , the machining knowledge B 122 and the machining knowledge C 123 in several different machining steps on corresponding parts in order to generate rotary shapes. The term “turning shape” refers to the turning cross-sectional shape t in each machining step when a blank is machined into the final machining shape in several different machining steps.

A method for dividing the rotary cross-sectional shape t is specifically described below. 10 Fig. 13 is a diagram illustrating a turning cross-sectional shape SH1 of a facing step according to the first embodiment of the present invention. 11 Fig. 13 is a diagram illustrating a turning cross-sectional shape SH2 of a drilling turning step according to the first embodiment of the present invention. 12th Fig. 13 is a diagram illustrating a turning cross-sectional shape SH3 of a rod material turning machining step according to the first embodiment of the present invention. 13th Fig. 13 is a diagram illustrating a turning cross-sectional shape SH4 of a rod material turning machining step according to the first embodiment of the present invention.

The NC machining program generation unit 43 receives information about the blank area allowance: 10 mm from the turning cross-section shape t1. The NC machining program generation unit 43 then refers to the machining knowledge A 121 , determines that the facing step is to be performed first because the allowance is not 20 mm or greater, and separates as in 10 is shown, the turning cross-sectional shape SH1 for the facing step from the turning cross-sectional shape t1.

The NC machining program generation unit 43 then refers to machining knowledge A 121 , determines that the drilling-turning step is to be performed after the facing step, and separates as in FIG 11 is shown, the turning cross-sectional shape SH2 for the drilling turning step from the turning cross-sectional shape t1.

Next, the NC machining program generating unit separates 43 , as in 12th is shown, the turning cross-sectional shape SH3 for the rod material turning machining step from the turning cross-sectional shape t1. The NC machining program generation unit then separates 43 , as in 13th is shown, the turning cross-sectional shape t2 as the turning cross-sectional shape SH4 for the rod material turning machining step.

Next, the NC machining program generating unit arranges 43 In step ST370, the divided machining shapes, that is, the turning cross-sectional shapes, based on the machining knowledge A 121 , the machining knowledge B 122 and the machining knowledge C 123 Turning machining steps to in order to this way to generate the turning machining steps. 14th Fig. 13 is a diagram illustrating an example of the machining start point and the machining end point in the facing step according to the first embodiment of the present invention. 15th FIG. 13 is a diagram illustrating an example of the machining start point and the machining end point in the drilling turning step according to the first embodiment of the present invention. 16 Fig. 13 is a diagram illustrating an example of the cutting point, the machining start point, and the machining end point in the bar material turning machining step according to the first embodiment of the present invention. 17th Fig. 13 is a diagram illustrating an example of the cutting point, the machining start point, and the machining end point in the bar material turning machining step according to the first embodiment of the present invention.

The NC machining program generation unit 43 as in 14th is shown, the coordinate values of a machining start point P1 and the coordinate values of a machining end point P2 of the facing step from the turning cross-sectional shape SH1 for the facing step and generates a facing step LC1, which is the facing step.

The NC machining program generating unit then receives 43 , as in 15th is shown, the bore diameter of the bore to be drilled, the coordinate values of a machining start point P3 and the coordinate values of a machining end point P4 from the rotary cross-sectional shape SH2 for the drilling and turning step and generates a drilling and turning step LC2, which is the drilling and turning step.

Next, the NC machining program generating unit is given 43 , as in 16 is shown, information on the longitudinal length of the shape of the open area: 30 mm and information on the lateral length of the shape of the open area: 30 mm from the turning cross-sectional shape SH3 for the bar material turning machining step. The NC machining program generation unit 43 then refers to the machining knowledge B 122 and determines that the part to be machined in the rod material turning machining step is the outer circumference because “lateral length of the shape of the open area / longitudinal length of the shape of the open area” is “1.0”. Next, the NC machining program generating unit is given 43 the coordinate values of a cutting point P5, the coordinate values of a machining start point P6 and the coordinate values of a machining end point P7 of the rotary cross-section shape SH3 for the rod material turning machining step from the turning cross-sectional shape SH3 and generates a rod outer circumference turning step LC3, which is a rod material turning machining step in which the outer circumference was determined as the part to be machined.

Next, the NC machining program generating unit is given 43 , as in 17th is shown, information on the longitudinal length of the rotary cross-sectional shape SH4: 10 mm and information on the lateral length of the rotary cross-sectional shape SH4: 20 mm from the rotary cross-sectional shape SH4 for the bar material turning machining step. The NC machining program generation unit 43 then refers to the machining knowledge C 123 and determines that the rod material turning machining step is to be performed because the longitudinal length of the turning cross-sectional shape SH4 is 10 mm or less and the lateral length of the turning cross-sectional shape SH4 is not 10 mm or less.

Next, the NC machining program generating unit is given 43 from the turning cross-sectional shape SH4 for the bar material turning machining step, information on the longitudinal length of the shape of the open area: 10 mm and information on the lateral length of the shape of the open area: 20 mm. The NC machining program generation unit 43 then refers to the machining knowledge B 122 and determines that the part to be machined is the outer circumference because “lateral length of the shape of the open area / longitudinal length of the shape of the open area” is “2.0” and therefore is not “less than 1.0”. Next, the NC machining program generating unit is given 43 the coordinate values of a cutting point P8, a machining start point P9 and a machining end point P10 of the turning cross-sectional shape SH4 from the turning cross-sectional shape SH4 for the rod material turning machining step and generates a rod outer circumference turning step LC4, which is a rod material turning machining step in which the outer circumference was determined as the part to be machined.

• Plandrehschritt LC1: Bearbeitungsstartpunkt P1 (75, -10) und Bearbeitungsendpunkt P2 (0, 0)
• Bohrdrehschritt LC2: Bohrungsdurchmesser 30 mm, Bearbeitungsstartpunkt P3 (0, 0) und Bearbeitungsendpunkt P4 (0, 110)
• Stabaußenumfangsdrehschritt LC3: Schneidpunkt P5 (75, 0), Bearbeitungsstartpunkt P6 (45, 0) und Bearbeitungsendpunkt P7 (45, 30)
• Stabaußenumfangsdrehschritt LC4: Schneidpunkt P8 (75, 60), Bearbeitungsstartpunkt P9 (65, 60) und Bearbeitungsendpunkt P10 (65, 80).

The following is details of the in the NC machining program generating unit 43 generated turning machining steps described.

• Facing step LC1: machining start point P1 (75, -10) and machining end point P2 (0, 0)
• Drilling and turning step LC2: hole diameter 30 mm, machining start point P3 (0, 0) and machining end point P4 (0, 110)
• Bar outer circumference turning step LC3: cutting point P5 (75, 0), machining start point P6 (45, 0) and machining end point P7 (45, 30)
• Bar outer circumference turning step LC4: cutting point P8 (75, 60), machining start point P9 (65, 60) and machining end point P10 (65, 80).

In the above-described method, the NC machining program generating unit refers 43 on the machining knowledge A 121 , the machining knowledge B 122 and the machining knowledge C 123 and can thus automatically generate the several different turning steps for machining a blank into the final machining shape. In addition, the automatic generation of the turning machining steps with reference to the machining knowledge that the editing that reflects the knowledge and experience of the user, integrated and accumulated, enables the automatic generation of the turning machining steps desired by the user, the turning machining steps in the case of generating an NC -Editing program by the user himself are similar.

Furthermore, the above-mentioned learning action performed by the machine learning device 50 according to the first embodiment is carried out continuously each time an NC machining program is generated. Accordingly, the NC machining program creating apparatus performs 40 According to the first embodiment, an autonomous learning of a machining knowledge continuously, which the method for generating an NC machining program in accordance with the CAD data 100 which includes information about the machining end shape of the machined object and information about the material of the blank. That is, the NC machining program generating device 40 who have favourited the machine learning device 50 includes, continuously and autonomously, machining knowledge that offers high machining efficiency and reflects the knowledge and experience of the user based on that in the NC machining program editing unit 45 the editing performed by the NC machining program generation unit 43 generated NC machining program learned.

Next, a detail of a method for generating edit-integrated machining knowledge will be described. In this procedure, the condition monitoring unit refers 51 the editing of the first NC machining program 111 into the machining knowledge on which the generation of the first NC machining program 111 Reference was made in order to generate an edit-integrated machining knowledge. A case will first be described in which the first NC machining program 111 has been edited to execute the facing step LC1 according to the user-inputted editing instruction information after the drilling turning step LC2. The condition observation unit 51 receives information about the blank surface allowance: 10 mm from the turning cross-section shape SH1 for the facing step LC1 of the second NC machining program 112 that is produced by the NC machining program editing unit 45 was transferred.

In this case it became the second NC machining program 112 by editing the first NC machining program 111 generated, in contrast to the case of the criterion of machining knowledge A 121 Carry out the drilling and turning step for the criterion of adding the blank surface area: 10 mm first. The condition monitoring unit relates accordingly 51 the editing of the first NC machining program 111 and generates an editing-integrated processing knowledge A 131 resulting from the change in processing knowledge A 121 according to the following criteria.

(Editing-integrated processing knowledge A 131 )

Blank material: S45C

Machining sequence: First the drilling and turning is carried out and then the facing step is carried out if the blank surface allowance is "10 mm or greater". The facing step and then the drilling and turning step are carried out if the blank surface allowance does not meet this condition.

Next, a description will be given of a case where the first NC machining program 111 has been edited according to the user inputted editing instruction information to change the part to be machined in the rod outer circumference turning step LC3 from the outer circumference to the end face, thus changing the rod outer circumference turning step LC3 to a rod end face turning step LC31. The condition observation unit 51 obtains information on the longitudinal length of the shape of the open area: 30 mm and information on the lateral length of the shape of the open area: 30 mm from the turning cross-sectional shape SH3 for the rod outer circumference turning step LC3 of the second NC machining program 112 that is produced by the NC machining program editing unit 45 is transmitted.

In this case it became the second NC machining program 112 by editing the first NC machining program 111 generated to the part to be machined in contrast to the case of the criterion of machining knowledge B 122 for the criterion "lateral length of the shape of the open area / longitudinal length of the shape of the open area" to be changed to "1.0" in the end face. The condition monitoring unit relates accordingly 51 the editing of the first NC machining program 111 and generates processing knowledge that is integrated into the editing process B 132 obtained when the machining knowledge B 122 is modified to include the following criteria.

(Editing-integrated processing knowledge B 132 )

Blank material: S45C

Part to be machined: The part to be machined is determined as the end face when “lateral length of the shape of the open area / longitudinal length of the shape of the open area” is “1.0 or less”. The part to be machined is determined as the outer circumference if the “lateral length of the shape of the open area / longitudinal length of the shape of the open area” is “greater than 1.0”.

Next, a description will be given of a case where the first NC machining program 111 has been edited to change the rod outer circumference turning step LC4 to a groove turning step LC41 according to the editing instruction information input by the user. The condition observation unit 51 receives information on the longitudinal length of the rotary cross-sectional shape SH4: 10 mm and information on the lateral length of the rotary cross-sectional shape SH4: 20 mm from the rotary cross-sectional shape SH4 for the rod outer circumference turning step LC4 of the second NC machining program 112 that is produced by the NC machining program editing unit 45 was transferred.

In this case it became the second NC machining program 112 by editing the first NC machining program 111 generated to, in contrast to the case of the criterion of machining knowledge C 123 for a criterion “the longitudinal length of the rotary cross-sectional shape SH4 is 10 mm and the lateral length of the rotary cross-sectional shape SH4 is 20 mm”, the groove turning step LC41 is to be carried out. The condition monitoring unit relates accordingly 51 the editing of the first NC machining program 111 and generates an editing-integrated processing knowledge C 133 resulting from the change in machining knowledge C 123 according to the following criteria.

(Editing-integrated processing knowledge C 133 )

Blank material: S45C

Machining method: The groove turning step is performed when the longitudinal length of the turning cross-sectional shape is “10 mm or less” and the lateral length of the turning cross-sectional shape is “20 mm or less”. The rod material turning machining step is carried out when either the longitudinal length or the lateral length of the turning cross-sectional shape does not meet these conditions.

A method for dividing the rotary cross-sectional shape t is specifically described below for a case in which the NC machining program generation unit 43 on the editing-integrated processing knowledge A 131 , the editing-integrated processing knowledge B 132 and the editing-integrated processing knowledge C 133 References and generates an NC machining program. 18th Fig. 13 is a diagram illustrating a turning cross-sectional shape SH5 of the drilling turning step according to the first embodiment of the present invention. 19th Fig. 13 is a diagram illustrating a turning cross-sectional shape SH6 of the facing step according to the first embodiment of the present invention. 20th Fig. 13 is a diagram illustrating a turning cross-sectional shape SH7 of the rod material turning machining step according to the first embodiment of the present invention. 21 Fig. 13 is a diagram illustrating a turning cross-sectional shape SH8 of the rod material turning machining step according to the first embodiment of the present invention.

The NC machining program generation unit 43 receives information about the length of the blank: 120 mm and information about the outer diameter of the blank: 150 mm from the blank shape and information about the blank area allowance: 10 mm from the turning cross-section shape t1. The NC machining program generation unit 43 then relates to the editing-integrated processing knowledge A 131 , determines that the drilling-turning step is to be performed first because the allowance is 10 mm or more, and separates, as in 18th is shown, the turning cross-sectional shape SH5 for the drilling turning step from the turning cross-sectional shape t1.

Next, the NC machining program generating unit takes 43 on the editing-integrated processing knowledge A 131 Reference, determines that the facing step is to be performed after the drilling and turning step, and disconnects as in FIG 19th is shown, the turning cross-sectional shape SH6 for the facing step from the turning cross-sectional shape t1.

Next, the NC machining program generating unit separates 43 , as in 20th is shown, the turning cross-sectional shape SH7 for the rod material turning machining step from the turning cross-sectional shape t1. The NC machining program generation unit then separates 43 , as in 21 is shown, the turning cross-sectional shape t2 as the turning cross-sectional shape SH8 for the bar material turning machining step.

A method for assigning the turning machining steps is described below specifically for the case that the NC machining program generation unit 43 on the editing-integrated processing knowledge A 131 , the editing-integrated processing knowledge B 132 and the editing-integrated processing knowledge C 133 References and generates an NC machining program. 22nd FIG. 13 is a diagram illustrating an example of the machining start point and the machining end point in the drilling turning step according to the first embodiment of the present invention. 23 Fig. 13 is a diagram illustrating an example of the machining start point and the machining end point in the facing step according to the first embodiment of the present invention. 24 Fig. 13 is a diagram illustrating an example of the cutting point, the machining start point, and the machining end point in the bar face turning step according to the first embodiment of the present invention. 25th Fig. 13 is a diagram illustrating an example of the cutting point, machining start point, and machining end point in the bar groove turning step according to the first embodiment of the present invention.

The NC machining program generation unit 43 as in 22nd is shown, the drilling diameter to be removed when drilling turning, the coordinate values of a machining start point P11 and the coordinate values of a machining end point P12 from the rotary cross-sectional shape SH5 for the rotary drilling step and generates a rotary drilling step LC11, which is the rotary drilling step.

The NC machining program generating unit then receives 43 , as in 23 is shown, from the turning cross-sectional shape SH6 for the facing step, the coordinate values for a machining start point P13 and the coordinate values for a machining end point P14 of the facing step and generates a facing step LC12, which is the facing step.

Next, the NC machining program generating unit is given 43 , as in 24 is shown, from the turning cross-sectional shape SH7 for the bar material turning machining step, information on the longitudinal length of the shape of the open area: 30 mm and information on the lateral length of the shape of the open area: 30 mm. The NC machining program generation unit 43 then refers to the editing-integrated processing knowledge B 132 and determines that the part to be machined in the rod material turning machining step is the end face because “lateral length of the shape of the open area / longitudinal length of the shape of the open area” is “1.0”, that is, “1.0 or less”. Next, the NC machining program generating unit is given 43 the coordinate values of a cutting point P15, the coordinate values of a machining start point P16 and the coordinate values of a machining end point P17 of the turning cross-section shape SH7 for the rod material turning machining step from the turning cross-section shape SH7 and generates a rod end face turning step LC3, which is a rod material turning machining step in which the end face was determined as the part to be machined.

Next, the NC machining program generating unit is given 43 , as in 25th is shown, information on the longitudinal length of the rotary cross-sectional shape SH8: 10 mm and information on the lateral length of the rotary cross-sectional shape SH8: 20 mm from the rotary cross-sectional shape SH8 for the bar material turning machining step. The NC machining program generation unit 43 then takes on the editing-integrated processing knowledge C 133 Reference and determines that the grooving step is to be performed because the longitudinal length of the rotary cross-sectional shape SH4 is 10 mm or less and the lateral length of the rotary cross-sectional shape SH4 is 20 mm or less.

Next, the NC machining program generating unit is given 43 From the turning cross-sectional shape SH8 for the bar material turning machining step, information on the longitudinal length of the shape of the open area: 10 mm and information on the lateral length of the shape of the open area: 20 mm. The NC machining program generation unit 43 receives the information about the groove depth: 10 mm from the information about the longitudinal length of the shape of the open area: 10 mm and the information about the groove width: 20 mm from the information about the lateral length of the shape of the open area: 20 mm. The NC machining program generation unit 43 then receives the coordinate values of a machining start point P18 from the turning cross-sectional shape SH8 for the bar material turning machining step and generates a groove turning step LC14.

• . Bohrdrehschritt LC11: Bohrungsdurchmesser 30 mm, Bearbeitungsstartpunkt P11 (0, -10) und Bearbeitungsendpunkt P12 (0, 110)
• . Plandrehschritt LC12: Bearbeitungsstartpunkt P13 (75, -10) und Bearbeitungsendpunkt P14 (15, 0)
• . Stabstirnflächendrehschritt LC13: Schneidpunkt P15 (75, 0), Bearbeitungsstartpunkt P16 (75, 30) und Bearbeitungsendpunkt P17 (45, 30)
• . Nutdrehschritt LC14: Nutbreite 20 mm, Nuttiefe 10 mm und Bearbeitungsstartpunkt P18 (75, 60).

In the following, with reference to the processing knowledge A 131 , the editing-integrated processing knowledge B 132 and the editing-integrated processing knowledge C 133 those in the NC machining program generation unit 43 generated turning machining steps described in detail.

• . Drilling and turning step LC11: hole diameter 30 mm, machining start point P11 (0, -10) and machining end point P12 (0, 110)
• . Facing step LC12: machining start point P13 (75, -10) and machining end point P14 (15, 0)
• . Bar face turning step LC13: cutting point P15 (75, 0), machining start point P16 (75, 30) and machining end point P17 (45, 30)
• . Groove turning step LC14: groove width 20 mm, groove depth 10 mm and machining start point P18 (75, 60).

In the above-described method, the NC machining program generation unit takes 43 on the editing-integrated processing knowledge A 131 , the editing-integrated processing knowledge B 132 and the editing-integrated processing knowledge C 133 Reference and can thus automatically generate the several different turning steps for the machining of a blank into the final machining shape. Further, the automatic generation of turning machining steps with reference to the editing-built machining knowledge that the editing that reflects the knowledge and experience of the user, integrated and accumulated, enables the automatic generation of the turning machining steps desired by the user, which is the turning machining steps in the case of generating a NC machining program by the user himself are similar.

The decision-making unit 46 determines the machining knowledge or the editing-integrated machining knowledge for the query of the NC machining program generation unit 43 using the learning results of the machine learning device 50 . It may be necessary that the NC machining program generation unit 43 Edits the NC machining program after the NC machining program has been generated. Since the machine learning device 50 according to the present embodiment has already learned the editing-integrated processing knowledge, the decision-making unit can 46 from the ones in the function update unit 54 stored machining knowledge, if necessary, that for the query of the NC machining program generation unit 43 determine the most suitable editing-integrated processing knowledge.

The machine learning device 50 is also able to establish a relationship between the material of the blank and the CAD data 100 and to learn machining knowledge. If it is provided that the NC machining program with regard to the material of the blank and the CAD data 100 is to be edited, the numerical control can be designed so that the decision making unit 46 determined to select a different machining knowledge or editing-integrated machining knowledge.

It should be noted that editing-integrated processing knowledge can be stored in such a way that the CAD data 100 themselves are linked to it. For example, the CAD data 100 with the editing-integrated processing knowledge A 131 linked and in the function update unit 54 together with the machining knowledge A 121 saved.

In addition, editing-integrated processing knowledge can also be stored in such a way that information relating to the CAD data 100 are linked to the editing knowledge that is integrated into the editing process. For example, information that relates to the CAD data 100 with the editing knowledge A 131 linked and together with the editing knowledge A 131 saved. Examples of information related to the CAD data 100 include an image of the CAD data and an element of the CAD data. It should be noted that “Information relating to the CAD data 100 In other words, “refers to“ information relating to a combination of information about the final machining shape of the machined object and information about the material of the blank ”.

This allows the decision-making unit 46 using the CAD data 100 or by yourself on the CAD data 100 related information as a search condition a desired, editing-integrated processing knowledge among several in the function update unit 54 Search and receive stored machining knowledge.

Furthermore, for each of the NC machine tools that are controlled by executing the NC machining program, a different set of machining knowledge and machining knowledge integrated into editing can be stored. When a plurality of NC machine tools are used, the NC machine tools often have different functions from each other, for example, in view of a situation where different types of NC machine tools are used, a situation where NC machine tools of the same type are used, but too different optional functions are added to various NC machine tools, or the like. In these cases there are suitable criteria for the machining knowledge on a per NC machine tool basis.

By storing a different set of machining knowledge and editing-integrated machining knowledge for each of the NC machine tools, the editing that reflects the knowledge and experience of the user and is suitable for each of the NC machine tools is integrated and accumulated in an editing-integrated machining knowledge. In this way, the NC machining program generating unit generates 43 an NC machining program with reference to machining knowledge or editing-integrated machining knowledge that is managed on a per NC machine tool basis. This enables an NC machining program that is desired by the user and is suitable for every NC machine tool and that is similar to an NC machining program generated by the user himself to be generated quickly and automatically.

In addition, the numerical control can be designed in such a way that the machining knowledge and the editing-integrated machining knowledge are stored in a storage unit provided in a network server.

As a variant of the machine learning device 50 According to the first embodiment, the machine learning device 50 perform machine learning according to a neural network model. 26th Fig. 13 is a schematic diagram illustrating an example of a neural network model to which the machine learning apparatus 50 according to the first embodiment of the present invention follows. A neural network is composed, for example, of an input layer that includes an input neuron, an intermediate layer that includes a number of m neurons, and an output layer that includes a number of n output neurons. 26th Figure 3 illustrates an example of a neural network model constructed from an input layer with four input neurons, an intermediate layer with five neurons, and an output layer with one output neuron. It should be noted that 26th illustrates only one layer as an intermediate layer, but there may be two or more intermediate layers.

A neural network learns the relationship between a state variable and an environmental change, e.g. B. when an optimal action is learned intuitively on the basis of a successful experience or an unsuccessful experience of a person. A neural network learns the relationship between a state variable and an environmental change on the basis of a data set that is generated on the basis of a state variable or of state variables that are generated by the state observation unit 51 to be observed. This means that a neural network learns the relationship between a state variable and an environmental change through what is known as supervised learning. In this case, the machine comprises learning device 50 the output layer for outputting machining knowledge corresponding to the machining shape data and better reflecting the knowledge and experience of the user in response to the machining shape data input to the input layer of the neural network and the NC machining program generation method. Then the decision making unit 46 operated to determine an optimal NC machining program generation method. Even in the case of using a neural network model, the machine learning device 50 that is, the relationship among the machining shape data, the NC machining program generating method, and the editing of the first NC machining program 111 learn, thereby learning the method of creating an NC machining program.

It should be noted that, although an embodiment has been described above, in which the machine learning device 50 machine learning using reinforcement learning or using a neural network performs the machine learning device 50 also can perform machine learning according to another known method, such as genetic programming, functional logic programming or a support vector machine.

Furthermore, the learning unit 52 according to one embodiment of the learning of the learning unit 52 learn a machining knowledge to be applied to an NC machining program used in another numerical control. The learning unit 52 Thus, it can be configured to use machining knowledge to be applied to an NC machining program used in another numerical control based on the machining knowledge to be used in generating the NC machining program with respect to certain machining shape data generated for another numerical control and learned on the basis of the editing of the NC machining program. The learning unit 52 can learn machining knowledge to be applied to an NC machining program used in another numerical control operated in the same location, or it can learn machining knowledge to be applied to an NC machining program used in another numerical control is used, which is operated independently at a different location.

27 Fig. 13 is a diagram showing a hardware configuration of the NC machining program creating apparatus 40 illustrated in accordance with the first embodiment of the present invention. In the 27 Hardware shown includes a processor 201 that performs arithmetic operations, a memory 202 from the processor 201 is used as a work area, a storage device 203 that stores a program for operation as a numerical controller or a program converting device, an input device 204 , which represents an input interface for a user, a display device 205 that displays information to the user and a communication device 206 whose function is to communicate with a device to be controlled or with another numerical control and / or with various other devices. The processor 201 , the memory 202 , the storage device 203 , the input device 204 who have favourited the display device 205 and the communication device 206 are via a data bus 207 connected with each other. The processor can do this 201 a processing unit, a computing unit, a microprocessor, a microcomputer, a central processing unit (CPU), a digital signal processor (DSP) or the like. Furthermore, the memory 202 a non-volatile or volatile semiconductor memory such as a random access memory (RAM), a read-only memory (ROM), a flash memory, an erasable programmable ROM (EPROM) or an electrically erasable programmable ROM (EEPROM) (registered trademark); a magnetic disk, a floppy disk, an optical disk, a compact disk, a mini disk, a digital versatile disk (DVD) or the like.

The NC machining program generation unit 43 for example, by executing a program contained in the in 27 memory shown 202 is stored by the processor 201 implemented. The above functionality can be implemented by multiple processors in conjunction with multiple memories. Furthermore, the functionality of the NC machining program generation unit 43 partially implemented as an electronic circuit, with the remaining functionality being provided by the processor 201 and the memory 202 can be implemented.

Furthermore, the NC machining program generating device 40 be designed so that the machining shape data input unit 41 , the NC machining program editing unit 45 , the decision-making unit 46 , the condition observation unit 51 , the reward calculation unit 53 and the function update unit 54 similarly by running one in memory 202 stored program by the processor 201 implemented. In addition, the aforementioned functionality can be implemented by multiple processors in conjunction with multiple memories. Furthermore, the functionality of the machining shape data input unit 41 , the NC machining program editing unit 45 , the decision-making unit 46 , the condition observation unit 51 , the reward calculation unit 53 and the function update unit 54 partially implemented as an electronic circuit, with the remaining functionality through the processor 201 and the memory 202 can be implemented. In addition, the processor and the memory can be used to implement the functionality of the machining shape data input unit 41 , the NC machining program editing unit 45 , the decision-making unit 46 , the condition observation unit 51 , the reward calculation unit 53 and the function update unit 54 be the same processor and the same memory or a different processor and a different memory, which are different from the processor and the memory for implementing the NC machining program generation unit 43 distinguish.

The NC machining program generating device 40 according to the first embodiment described above, refers to machining knowledge that the editing that reflects the knowledge and experience of the user has integrated and accumulated, and automatically generates Turning steps based on CAD data 100 which are the machining shape data. This enables the automatic generation of the turning machining steps desired by the user, which are similar to the turning machining steps in the case of the generation of an NC machining program by the user himself.

In addition, the NC machining program generating device learns 40 who have favourited the machine learning device 50 comprises, according to the first embodiment, a machining knowledge of the method for generating an NC machining program and the CAD data 100 which includes information on the final machining shape of the machined object and information on the material of the blank. The the machine learning device 50 comprehensive NC machining program generating device 40 That is, independently learns a machining knowledge that offers high machining efficiency and reflects the knowledge and experience of the user based on that in the NC machining program editing unit 45 the editing performed by the NC machining program generation unit 43 generated NC machining program. This allows the decision-making unit 46 based on the learning result, the machining knowledge that the CAD data 100 corresponds to, has the highest action value Q and offers the highest processing efficiency, among those in the function update unit 54 stored machining knowledge determine when the CAD data 100 again into the NC machining program generating device 40 can be entered.

The NC machining program generating device 40 According to the first embodiment, therefore, enables the automatic generation of a high-quality NC machining program that offers high machining efficiency and performs turning machining steps close to the turning machining steps desired by the user without accumulating the knowledge and skills of an expert. Since there is no need for the user to create an NC machining program on a trial and error basis, work efficiency in creating an NC machining program is improved. Therefore, the NC machining program creating apparatus 40 according to the first embodiment, it is effective in reducing the time required for automatically generating an NC machining program and in reducing the machining time for a workpiece.

The NC machining program generating device 40 According to the first embodiment further learns the editing of the NC machining program performed by the user and incorporates the editing into the machining knowledge, whereby the scope of the editing operation of the NC machining program by the user can be reduced at a next opportunity of the NC program generation.

Various aspects of the embodiment of the present invention have been described. It is of course possible that a person skilled in the art can achieve a function and effect intended by the present invention by another embodiment. In particular, a component of the above first embodiment can be removed or replaced, or known agents can be added without departing from the scope of the present invention. In addition, it is obvious to a person skilled in the art that the present invention can also be carried out by any combination of features of the embodiment expressly or implicitly disclosed here. The configurations described in the above embodiment are therefore only examples of various aspects of the present invention. These configurations can be combined with known other technology, and moreover, a part of these configurations can be omitted and / or modified without departing from the gist of the present invention.

List of reference symbols

1

numerical control;

10

Processing unit for interactive operation;

20th

Display unit;

30th

Instruction input unit;

40

NC machining program generating device;

41

Machining shape data input unit;

42

Machining shape data storage unit;

43

NC machining program generating unit;

44

NC machining program storage unit;

45

NC machining program editing unit;

46

Decision making unit;

50

Machine learning device;

51

Condition monitoring unit;

52

Learning unit;

53

Reward calculation unit;

54

Function update unit;

100

CAD data;

111

first NC machining program;

112

second NC machining program;

121

Machining knowledge A;

122

Machining knowledge B;

123

Machining knowledge C;

131

editing-integrated processing knowledge A;

132

editing-integrated processing knowledge B;

133

editing-integrated processing knowledge C;

201

Processor;

202

Storage;

203

Storage device;

204

Input device;

205

Display device;

206

Communication device;

207

Data bus.

QUOTES INCLUDED IN THE DESCRIPTION

This list of the documents listed by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Patent literature cited

• JP 2002132313 [0004]

Claims (5)

A machine learning device comprising: a state observation unit to act as state variables Machining shape data that includes information about a machining end shape of a machined object and information about a material of a blank, a numerically controlled machining program generation method for automatically generating a numerically controlled machining program comprising a plurality of machining operations for machining a workpiece into the machined object, and editing a first program for numerically controlled machining generated with reference to the method for generating a program for numerically controlled machining, to observe; and a learning unit to learn the method for generating a program for numerically controlled machining on the basis of a data record generated on the basis of the state variables. Machine learning device according to Claim 1 wherein the learning unit comprises: a reward calculating unit for calculating a reward based on a method of generating an edit-integrated program for numerically controlled machining resulting from the inclusion of the editing in the method of generating the program for numerically controlled machining; and a function update unit for updating a function for determining the method for generating a program for numerically controlled machining based on the calculated reward. Machine learning device according to Claim 2 wherein the reward calculation unit is configured to: increase a reward of the method for generating an edit-integrated program for numerically controlled machining if the method for generating a program for numerically controlled machining is identical to the method for generating an edit-integrated program for numerically controlled machining , and to reduce a reward of the method for generating a program for numerically controlled machining and to increase the reward of the method for generating an edit-integrated program for numerically controlled machining when the method for generating a program for numerically controlled machining of the method for generating an editing-integrated program for numerically controlled machining. Apparatus for generating a program for numerically controlled machining, comprising: the machine learning apparatus according to one of Claims 1 to 3 ; a numerically controlled machining program generating unit for generating the first numerically controlled machining program with reference to the method of generating a numerically controlled machining program determined based on a result of learning of the machine learning device; a numerically controlled machining program editing unit for editing the first numerically controlled machining program to generate a second numerically controlled machining program; and a decision making unit for determining a method of generating a program for numerically controlled machining that corresponds to the machining shape data based on the result of learning of the learning unit of the machine learning device. A machine learning method for learning a method of generating a program for numerically controlled machining to be used in automatic generation of a program for numerically controlled machining in an apparatus for generating a program for numerically controlled machining, the apparatus for generating a program For numerically controlled machining, the program for numerically controlled machining is generated, the program for numerically controlled machining comprising a plurality of machining processes for machining a workpiece into a machined object, the machine learning method comprising: a step in which machining shape data, the information on a final machining shape of a machined object and information on a material of a blank comprise the method for generating a program for a numerically controlled bea processing and an editing of a first numerically controlled machining program generated by the numerically controlled machining program generating apparatus with reference to the numerically controlled machining program generating method can be observed; and a step in which the method for generating a program for numerically controlled machining is learned on the basis of a data record generated on the basis of the state variables.

Images