WO2023149765A1

WO2023149765A1 - Machine tool thermal displacement compensation device and compensation method therefor

Info

Publication number: WO2023149765A1
Application number: PCT/KR2023/001641
Authority: WO
Inventors: 신향기
Original assignee: 주식회사 디엔솔루션즈
Priority date: 2022-02-07
Filing date: 2023-02-06
Publication date: 2023-08-10
Also published as: KR20230119360A

Abstract

The present invention relates to a machine tool thermal displacement compensation device and compensation method comprising: using a temperature sensor so as to calculate a direct compensation amount through direct calculation; calculating an indirect compensation amount through machine learning with respect to indirect factors, so as to calculate a final compensation amount through a non-reflected portion, unlike a conventional method, by means of the direct compensation amount and the indirect compensation amount, thereby automatically calculating a compensation amount in real time according to the thermal displacement of a structure unit; and immediately reflecting the calculated compensation amount so that a workpiece can be machined.

Description

Thermal displacement compensating device and compensating method for machine tools

The present invention relates to an apparatus and method for compensating for thermal displacement of a machine tool, and more particularly, by machine learning a direct factor and an indirect factor that changes in real time in addition to the direct factor through machine learning to determine a correction amount according to thermal displacement of a structural part in real time. An apparatus and method for compensating thermal displacement of a machine tool that automatically calculates and immediately reflects the calculated correction amount to process a workpiece.

In general, a machine tool refers to a machine used for the purpose of processing a metal/non-metal workpiece into a desired shape and dimension using an appropriate tool by various cutting processing methods or non-cutting processing methods.

Various types of machine tools, including turning centers, vertical/horizontal machining centers, door type machining centers, Swiss turns, electrical discharge machines, horizontal NC boring machines, and CNC lathes, are widely used in various industrial fields to suit the purpose of the work.

In addition, the machine tool has a table on which the material, which is a workpiece, is seated and transported for processing, a pallet for preparing the workpiece before processing, a spindle that rotates with a tool or workpiece combined, a tailstock for supporting the workpiece, etc. provide etc.

In general, in a machine tool, a table, a tool post, a main shaft, a tailstock, a resting ball, etc. are provided with a feed unit that feeds along a feed axis to perform various machining operations.

Also, in general, machine tools use a plurality of tools for various processing, and a tool magazine or a turret is used as a tool storage space for storing and storing a plurality of tools.

These machine tools use a plurality of tools for various processing, and a tool magazine is used in the form of a tool storage area for storing and storing a plurality of tools.

Also, in general, a machine tool includes an automatic tool changer (ATC) for withdrawing or re-accommodating a specific tool from a tool magazine according to a command of a numerical controller to improve productivity of the machine tool.

In addition, in general, machine tools are provided with an automatic pallet changer (APC) to minimize non-processing time. An Automatic Pallet Changer (APC) automatically exchanges pallets between the workpiece processing area and the workpiece installation area. A workpiece may be mounted on the pallet.

In general, numerical control (NC) is rapidly progressing in various types of machine tools currently being used.

Recently, as machine tools are automatically controlled using a computer equipped with numerical control (NC), machine tools applied with computerized numerical control (CNC) technology, which further develops numerical control (NC), are rapidly increasing. It is a trend that is spreading.

A numerically controlled (NC) or computer numerically controlled (CNC) machine tool has an operating panel. Such a control panel has various function switches or buttons and a monitor.

Unlike general computers, many home appliances or machine tools with special purposes have built-in embedded systems or edge devices that perform given tasks.

Recent machine tools are equipped with a numerical control system called CNC (Computer Numerical Control), and such CNC can control the operation of the machine tool in a desired form in various ways.

In general, a turning center refers to a machine tool equipped with an automatic tool changer and performing various machining by exchanging various types of tools. It is divided into horizontal turning centers.

In general, a machining center refers to a machine tool that is equipped with an automatic tool changer and exchanges various types of tools to perform a wide range of machining that can be performed on a lathe, milling, drilling, boring machine, etc. It is divided into a vertical machining center and a horizontal machining center in which the main axis is mounted vertically.

In general, machine tools such as turning centers and machining centers operate at low speeds of hundreds of millimeters per minute (mm) during processing such as tapping, drilling, boring, and cutting to improve machining precision and prevent safety accidents. It performs machining while moving at the cutting speed. On the other hand, machine tools such as turning centers and machining centers rapidly move at the maximum speed of several tens of meters per minute (m) during non-processing, thereby shortening the machining time and maximizing productivity by reducing the overall cycle time, reduce production costs.

However, in conventional machine tools such as turning centers and machining centers, thermal displacement may occur due to heat generated in the process of processing a workpiece, so that the processing dimensions of the workpiece may be changed.

Accordingly, conventional thermal displacement compensating devices and correction methods of machine tools such as turning centers and machining centers perform correction by directly calculating the correction amount according to thermal displacement with a temperature sensor or by indirectly calculating the correction value through a program, etc. performed.

However, in the case of directly calculating the thermal displacement correction amount with only such a temperature sensor, the precision changes according to the number of attachments of the temperature sensor, and the installation cost increases as a plurality of temperature sensors must be installed. As the correction value is calculated according to the measured temperature, the procedure is complicated and time delay occurs, the temperature change is not reflected in the structure where the temperature sensor is not installed, so the accuracy is reduced, and the miniaturization of the equipment cannot be promoted. there was

In addition, in the case of performing correction only with a temperature sensor in the conventional thermal displacement compensation device and correction method of a machine tool, the temperature change of the structure received from the temperature sensor is not a real-time temperature change, but a past where thermal displacement has already occurred according to the temperature conversion in the structure. There was a problem in that accuracy and reliability decreased because it was not possible to perform precise, accurate and real-time correction for thermal displacement through post-correction by measuring the temperature of the temperature.

Moreover, even when the compensation value is indirectly calculated through a program, etc., accuracy is guaranteed to some extent when the machine tool performs only simple structural processing as before, but milling, turning, and complex processing are performed. In the case of machine tools such as complex turning centers or machining centers equipped with various axis systems to perform complex machining, it is almost impossible to calculate the compensation amount indirectly, or even when calculated, accuracy and precision decrease depending on the program and model, resulting in reliability and reliability. and stability are reduced, and a lot of cost and time are consumed due to the need to install a separate program or equipment, and there are problems that cause user inconvenience.

Accordingly, thermal displacement due to heat generation of the structure is performed by using a temperature sensor in a machine tool such as a complex turning center or machining center for performing milling, turning, and complex processing, equipped with various shaft systems for this purpose, and performing complex processing. Calculate the direct correction amount firstly through direct factors, and secondarily calculate the indirect correction amounts accurately in real time through machine learning based on machine learning through indirect factors according to the processing state or process mode of the workpiece other than the temperature sensor, The final correction amount is calculated based on the calculated direct correction amount and indirect correction amount, and immediately reflected in the processing of the workpiece to minimize time delay by reflecting the existing non-reflected part, maximize accuracy and reliability, and install There is an urgent need to develop a thermal displacement compensating device and compensating method for machine tools capable of reducing costs and maintenance costs and improving user convenience and satisfaction.

The present invention is to solve the above problems, and an object of the present invention is to calculate a direct correction amount through a direct calculation method using a temperature sensor, and calculate an indirect correction amount through machine learning by machine learning for an indirect factor. By calculating the final correction amount by the non-reflected part differently from the conventional method by the direct correction amount and the indirect correction amount, the correction amount according to the thermal displacement of the structure is automatically calculated in real time, and the calculated correction amount is immediately reflected to process the workpiece. It relates to a thermal displacement compensating device and a compensating method of a machine.

More specifically, thermal displacement in a machine tool such as a complex turning center or machining center that includes at least one of milling, turning, composite, or designated machining, and has various axis systems for this purpose and performs composite machining. In the correction unit, the correction amount is directly calculated through a direct factor, which is the temperature of the structure sensed by the temperature sensor, and in addition to direct factors such as the temperature of the structure, the spindle RPM, motor load, feed shaft load, and feed axis are changed in real time. The indirect correction amount is calculated by machine learning indirect factors such as speed through machine learning, and finally, the final correction amount due to the thermal displacement of the structure is calculated automatically in real time with the direct correction amount and the indirect correction amount, and it is reflected immediately to improve the processing of the workpiece. As a result, time delay due to temperature deformation of existing structures and time delay due to the number and location of temperature sensors installed are prevented, and thermal displacement of structures that is more accurate and reliable than inaccurate corrected temperature estimation simply performed through machine learning. It relates to a thermal displacement compensating device and compensating method of a machine tool capable of automatically calculating a final compensating value for correcting a machining origin in real time.

In addition, the present invention stores and manages accumulated data for machine learning for a certain period of time to prevent overfitting of the amount of indirect correction by indirect factors, and is manufactured in the form of an edge device mounted on a machine tool separately as needed to improve compatibility of equipment It is easily and quickly installed on existing machine tools to promote miniaturization of machine tools, reduce space utilization, improve processing precision by calculating accurately and precisely in real time, increase productivity, and To provide a machine learning-based thermal displacement compensation device and compensation method for machine tools through machine learning that can reduce maintenance costs and time due to damage or breakage of workpieces and increase customer satisfaction.

However, the technical problems to be achieved by the present invention and the embodiments of the present invention are not limited to the technical problems described above, and other technical problems may exist.

In order to achieve the object of the present invention, a thermal displacement compensating device for a machine tool according to the present invention is a thermal displacement compensating device for a machine tool for compensating thermal displacement during processing of a workpiece, wherein the workpiece and a tool are provided for processing the workpiece. a structure in which a structure capable of relative movement is installed; a sensing unit to measure the temperature of the structural unit; A thermal displacement correcting unit that automatically calculates a correction amount according to the thermal displacement of the structure unit in real time by machine learning through machine learning a direct factor and an indirect factor that changes in real time in addition to the direct factor; and a control unit that immediately reflects the correction amount calculated by the thermal displacement compensating unit and performs processing of the workpiece.

In addition, in another preferred embodiment of the thermal displacement compensation device for a machine tool according to the present invention, the direct factor of the thermal displacement compensation device for a machine tool is the temperature of the structure measured by the sensing unit, and the indirect factor is the temperature of the workpiece It may be a processing state and a process mode that change in real time during the processing of the process.

In addition, in another preferred embodiment of the thermal displacement compensation device for a machine tool according to the present invention, the process mode of the thermal displacement compensation device for a machine tool includes a turning mode for performing only turning of the workpiece; a milling mode in which only milling of the workpiece is performed; a combined mode for performing both turning and milling of the workpiece; or a designation mode for performing selective processing on the workpiece according to a user's designation; It may include at least any one or more of them.

In addition, in another preferred embodiment of the thermal displacement compensation device for a machine tool according to the present invention, the structural part of the thermal displacement compensation device for a machine tool includes a milling spindle, a turret, and a processing unit having a spindle and directly processing the workpiece; a transfer unit for transferring the processing unit; and a support portion installed so that the processing portion and the transfer portion are movable.

In addition, in another preferred embodiment of the thermal displacement compensation device for a machine tool according to the present invention, the indirect factor of the thermal displacement compensation device for a machine tool is the machining part RPM and the machining part motor load according to the processing state and process mode of the workpiece , the conveying speed of the conveying part, and the change in motor load of the conveying part.

In addition, in another preferred embodiment of the thermal displacement compensation device for a machine tool according to the present invention, the sensing unit of the thermal displacement compensation device for a machine tool may be installed in at least one of the processing unit, the transfer unit, and the support unit. .

In addition, in another preferred embodiment of the thermal displacement compensating device for a machine tool according to the present invention, the control unit of the thermal displacement compensating device for a machine tool controls a processing program, a driving program, and a process mode for processing the workpiece. a memory unit for storing information; a processing unit configured to process the workpiece according to the information stored in the memory unit and the correction amount received from the thermal displacement correcting unit; an information collection unit that collects operation information of the processing unit and the transfer unit operated by the processing unit; and a transmission/reception unit configured to transmit the driving information collected by the information collection unit to the thermal displacement correcting unit or to receive the correction amount calculated by the thermal displacement correcting unit.

In addition, in another preferred embodiment of the thermal displacement compensation device for a machine tool according to the present invention, the control unit of the thermal displacement compensation device for a machine tool includes the memory unit, the processing information of the processing unit, and the operation information collected by the information collection unit. , and a display unit for displaying information transmitted and received through the transceiver.

In addition, in another preferred embodiment of the thermal displacement compensating device for a machine tool according to the present invention, the thermal displacement compensating unit of the thermal displacement compensating device for a machine tool machine learns the direct factor and the indirect factor through machine learning to determine the structural part. A database unit for storing information for automatically calculating a correction amount according to thermal displacement of the in real time; a confirmation unit for confirming whether or not processing has started in the processing unit according to the information stored in the database unit and processing information and operation information received from the control unit; a determination unit for determining a processing state and a process mode of the workpiece when processing is started in the processing unit according to a confirmation result of the confirmation unit; a calculation unit for calculating a final correction amount according to the thermal displacement of the structure unit according to information stored in the database unit, processing information and operation information received from the control unit, a confirmation result of the confirmation unit, and a determination result of the determination unit; and a correction unit configured to transmit the information stored in the database unit and the final correction amount calculated by the operation unit to the control unit.

In addition, in another preferred embodiment of the thermal displacement compensation device for a machine tool according to the present invention, the database unit of the thermal displacement compensation unit of the thermal displacement compensation device for a machine tool stores data on direct coefficients, direct weights, and switching Basic data storage unit; a received data storage unit for storing the data received from the transceiver; a machine learning data storage unit for storing a machine learning program and accumulation data for performing machine learning through machine learning; And a real-time data storage unit for storing the data sensed by the sensing unit, the processing state of the workpiece, and the amount of change in processing RPM, motor load of the processing unit, feed speed of the transfer unit, and motor load of the transfer unit according to the process mode in real time. can include

In addition, in another preferred embodiment of the thermal displacement compensating device for a machine tool according to the present invention, the calculation unit of the thermal displacement compensating unit of the thermal displacement compensating device for a machine tool, the confirmation result of the confirmation unit, the determination result of the determination unit, the basic a direct correction amount calculator for directly calculating a correction amount according to the data stored in the data storage unit, the data stored in the real-time data storage unit, and the temperature of the structure measured by the sensing unit; The confirmation result of the confirmation unit, the determination result of the determination unit, data stored in the basic data storage unit, data stored in the received data storage unit, data stored in the real-time data storage unit, data stored in the machine learning data storage unit, and An indirect correction amount calculation unit that automatically calculates an indirect correction amount in real time through machine learning using machine learning according to changes in processing part RPM, processing part motor load, conveying speed of the conveying part, and motor load of the conveying part; and a final correction amount calculation unit that calculates a final correction amount according to the direct correction amount calculated by the direct correction amount calculation unit and the indirect correction amount calculated by the indirect correction amount calculation unit.

In addition, in another preferred embodiment of the thermal displacement compensating device for a machine tool according to the present invention, the calculation unit of the thermal displacement compensating unit of the thermal displacement compensating device for a machine tool performs machine learning using machine learning in the indirect correction amount calculation unit. When the indirect correction amount is automatically calculated in real time, whether or not overfitting is automatically analyzed in real time according to the confirmation result of the confirmation unit and the judgment result of the judgment unit, and if overfitting occurs as a result of the analysis, the machine learning model selected by the indirect correction amount calculation unit It may further include an analysis unit that selects a machine learning model different from the above, the indirect correction amount calculation unit automatically recalculates the indirect correction amount in real time, and transfers the recalculated indirect correction amount to the final correction amount calculation unit.

In addition, in another preferred embodiment of the thermal displacement compensating device for a machine tool according to the present invention, the thermal displacement compensating unit of the thermal displacement compensating device for a machine tool before transferring the final correction amount to the controller, the processing information, If the operation information, the confirmation result of the confirmation unit, the determination result of the determination unit, and the data stored in the machine learning data storage unit are overcorrected, the final correction amount of the correction unit is automatically corrected by a low-pass filter and the corrected final correction amount It may further include a correction unit for transmitting to the control unit.

In addition, in another preferred embodiment of the thermal displacement compensating device for a machine tool according to the present invention, the stored data of the thermal displacement compensating device for a machine tool is controlled by the control unit, the confirmation result of the confirmation unit, and the determination result of the determination unit. Data of up to 30 minutes can be stored at intervals of 10 seconds from the moment the processing unit operates.

In addition, in another preferred embodiment of the thermal displacement compensation device for a machine tool according to the present invention, the machine learning model selected in the analysis unit of the thermal displacement compensation device for a machine tool is at least one of a neural network algorithm, LSTM, or a multilayer perceptron can be

In order to achieve another object of the present invention, a method for correcting thermal displacement of a machine tool according to the present invention calculates a correction amount according to thermal displacement of a structural part by machine learning a direct factor and an indirect factor that changes in real time in addition to the direct factor through machine learning. automatically calculating in real time and immediately reflecting the calculated correction amount to store information data for performing processing of a workpiece; Checking whether or not processing of the workpiece has started according to pre-stored information data and processing information and operation information received from the control unit; Measuring the temperature of the structural part when processing is started as a result of the confirmation; Collecting processing information, operation information, processing information, and temperature information when processing is started as a result of the confirmation; Storing the collected information as real-time data and accumulated data; determining a current processing state and process mode of the workpiece according to pre-stored information data, real-time data, and accumulated data; Directly calculating a correction amount according to the temperature of the structure measured in real time according to pre-stored information data and a judgment result; Automatically calculating an indirect correction amount in real time through machine learning using machine learning according to various changes in processing conditions and process modes that change in real time in the processing process of the workpiece according to pre-stored information data and judgment results; automatically calculating a final correction amount according to the calculated direct correction amount and indirect correction amount; and correcting the processing origin of the workpiece according to the calculated final correction amount.

In addition, in another preferred embodiment of the method for correcting thermal displacement of a machine tool according to the present invention, the method for correcting thermal displacement of a machine tool calculates an indirect adjustment amount through machine learning using machine learning after the step of calculating the indirect correction amount. When calculating automatically in real time, whether overfitting is automatically analyzed in real time according to the confirmation result and the judgment result, and if overfitting occurs as a result of the analysis, select a machine learning model different from the machine learning model selected when calculating the indirect correction amount The method may further include automatically recalculating the indirect correction amount in real time and calculating a final correction amount according to the recalculated indirect correction amount.

In addition, in another preferred embodiment of the method for compensating thermal displacement of a machine tool according to the present invention, the method for compensating thermal displacement of a machine tool after the correction step, before transferring the calculated final correction amount to the control unit, the processing information, the operation Further comprising a step of automatically correcting the final correction amount with a low-pass filter and correcting the machining origin of the workpiece according to the corrected final correction amount when the information, the confirmation result, the judgment result, and the accumulated data are overcorrected due to a small amount. can do.

In addition, in another preferred embodiment of the method for compensating thermal displacement of a machine tool according to the present invention, in the method for compensating thermal displacement of a machine tool, the direct factor is the temperature of the structure measured by the sensing unit, and the indirect factor is the temperature of the workpiece. It may be a change in the processing state of the workpiece and the process mode, which change in real time during the machining process, in the processing part RPM, the processing part motor load, the conveying speed of the conveying part, and the change in the motor load of the conveying part.

An apparatus and method for compensating for thermal displacement of a machine tool according to the present invention perform machining including at least one of milling, turning, composite, or designated machining, and for this purpose, various shaft systems are provided and for performing composite machining In machine tools such as complex turning centers and machining centers, the amount of compensation is calculated through direct calculation through a direct factor, which is the temperature of the structural part sensed by the temperature sensor in the thermal displacement compensator, and changes in real time in addition to direct factors such as the temperature of the structural part. Indirect factors such as spindle RPM, motor load, feed shaft load, feed shaft speed, etc. are machine learning through machine learning to calculate the indirect correction amount, and finally, the final correction amount due to thermal displacement of the structure is automatically calculated with the direct correction amount and the indirect correction amount. As the processing of the workpiece is performed by calculating and immediately reflecting it in real time, the existing non-reflected part is reflected to minimize the occurrence of time delay with the actual thermal displacement, and correction is performed in real time. It has the effect of maximizing reliability.

In addition, the apparatus and method for compensating thermal displacement of a machine tool according to the present invention stores and manages accumulated data for machine learning for a certain period of time to prevent overfitting of the indirect correction amount, and edge mounted on the machine tool separately as needed Manufactured in the form of a device, it is easily and quickly installed on an existing machine tool through compatibility of equipment, thereby miniaturizing the machine tool and reducing space utilization, thereby reducing manufacturing cost, installation cost, and maintenance cost.

Moreover, the apparatus and method for compensating for thermal displacement of a machine tool according to the present invention are direct factors according to the temperature of a structure and the processing state and process mode that change in real time during the processing of a workpiece, processing part RPM, processing part motor load, By reflecting the change in the conveying speed of the conveying part and the motor load of the conveying part in real time, calculating in real time through machine learning based on machine learning, automatically reflecting the final correction amount to process the workpiece, improving machining precision and productivity, There is an effect of conserving resources by preventing unnecessary defective products and increasing the convenience of users and workers.

In addition, the apparatus and method for correcting thermal displacement of a machine tool according to the present invention stores and manages accumulated data for machine learning for a certain period of time so that overfitting does not occur, and, if necessary, separate machine tool It is manufactured in the form of an edge device mounted on the machine and can be easily and quickly installed on existing machine tools through compatibility of equipment to improve compatibility and usability, improve consumer satisfaction, and increase exports.

1 shows a perspective view of a machine tool to which a thermal displacement compensating device and a correction method for a machine tool according to the present invention are applied, and a conceptual diagram of thermal displacement generation.

2 shows a configuration diagram of a thermal displacement compensating device for a machine tool according to the present invention.

3 shows a block diagram of a control unit of a thermal displacement compensating device for a machine tool according to the present invention.

4 shows a configuration diagram of a thermal displacement compensating unit of a thermal displacement compensating device for a machine tool according to the present invention.

5 shows a conceptual diagram of a thermal displacement compensating device for a machine tool according to the present invention.

6 shows a procedure diagram of a thermal displacement compensation method of a machine tool according to the present invention.

7 is a graph of an error amount over time when machine learning by machine learning is performed with LSTM for an indirect correction amount of a thermal displacement correction device and correction method of a machine tool according to the present invention.

8 is a graph of an error amount over time when machine learning by machine learning is performed with MLP for an indirect correction amount of a thermal displacement compensation device and correction method of a machine tool according to the present invention.

9 is a graph of a case where overfitting occurs in the thermal displacement compensating device and compensating method for a machine tool according to the present invention.

10 is a graph for the case where overcorrection occurs in the thermal displacement compensating device and correction method of a machine tool according to the present invention.

Hereinafter, a thermal displacement compensating device and compensating method of a machine tool according to an embodiment of the present invention will be described in detail with reference to drawings. The embodiments introduced below are provided as examples to sufficiently convey the spirit of the present invention to those skilled in the art. Accordingly, the present invention may be embodied in other forms without being limited to the embodiments described below. And, in the drawings, the size and thickness of the device may be exaggerated for convenience. Like reference numbers indicate like elements throughout the specification.

Since the present invention can apply various transformations and have various embodiments, specific embodiments will be illustrated in the drawings and described in detail in the detailed description. Effects and features of the present invention, and methods for achieving them will become clear with reference to the embodiments described later in detail together with the drawings. However, the present invention is not limited to the embodiments disclosed below and may be implemented in various forms. In the following embodiments, terms such as first and second are used for the purpose of distinguishing one component from another component without limiting meaning. Also, expressions in the singular number include plural expressions unless the context clearly dictates otherwise. In addition, terms such as include or have mean that features or elements described in the specification exist, and do not preclude the possibility that one or more other features or elements may be added. In addition, in the drawings, the size of components may be exaggerated or reduced for convenience of description. For example, since the size and thickness of each component shown in the drawings are arbitrarily shown for convenience of description, the present invention is not necessarily limited to the illustrated bar.

Advantages and features of the present invention, and methods for achieving them, will become clear with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various different forms, and only these embodiments make the disclosure of the present invention complete, and the common knowledge in the art to which the present invention belongs It is provided to fully inform the holder of the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numbers designate like elements throughout the specification. The sizes and relative sizes of layers and regions in the drawings may be exaggerated for clarity of explanation.

Terms used in this specification are for describing embodiments, and therefore are not intended to limit the present invention. In this specification, singular forms also include plural forms unless specifically stated otherwise in a phrase. As used herein, "comprise" and/or "comprising" means that a stated component, step, operation, and/or element is the presence of one or more other components, steps, operations, and/or elements. or do not rule out additions.

1 shows a perspective view of a machine tool to which a thermal displacement compensating device and correction method for a machine tool according to the present invention are applied and a conceptual diagram of thermal displacement generation, and FIG. 2 is a configuration diagram of a thermal displacement compensating device for a machine tool according to the present invention. indicates 3 shows a configuration diagram of a control unit of a thermal displacement compensation device for a machine tool according to the present invention, and FIG. 4 shows a configuration diagram of a thermal displacement compensation unit of a thermal displacement compensation device for a machine tool according to the present invention. 5 shows a conceptual diagram of a thermal displacement compensating device for a machine tool according to the present invention. 6 shows a procedure diagram of a thermal displacement compensation method of a machine tool according to the present invention. 7 is a graph of an error amount over time when machine learning by machine learning is performed with LSTM for an indirect correction amount of a thermal displacement correction device and correction method of a machine tool according to the present invention. 8 is a graph of an error amount over time when machine learning by machine learning is performed with MLP for an indirect correction amount of a thermal displacement compensation device and correction method of a machine tool according to the present invention. 9 is a graph of a case where overfitting occurs in the thermal displacement compensating device and compensating method for a machine tool according to the present invention. 10 is a graph for the case where overcorrection occurs in the thermal displacement compensating device and correction method of a machine tool according to the present invention.

Machine tools to which the thermal displacement compensating device and compensation method of machine tools according to the present invention are applied include all machine tools, but preferably perform milling, turning, and complex processing, and for this purpose, various axis systems are provided and complex processing is performed. It may be a machine tool such as a complex turning center or machining center to perform. In addition, although not necessarily limited thereto, the machine tool to which the thermal displacement compensating device and the compensating method of the machine tool according to the present invention are applied is general machine tool equipment such as a simple lathe in addition to complex equipment for performing complex processing such as turning or milling. As thermal displacement occurs during machining of a workpiece, processing must be performed by correcting the origin of processing to improve processing precision and increase productivity.

A thermal displacement compensating device and compensating method of a machine tool according to the present invention will be described with reference to FIGS. 1 to 10 . 1 to 5, the thermal displacement compensating device 10 for a machine tool according to the present invention includes a structural unit 100, a sensing unit 200, a thermal displacement compensating unit 300, and a control unit 400. include

In the structure unit 100, various structures for processing a workpiece are installed. That is, the structural part means a processing part, a transfer part, a support part, etc. that can relatively move the workpiece and the tool for machining the workpiece.

As shown in FIG. 1 , according to one embodiment of the present invention, the structural part 100 includes a processing part 110 , a transfer part 120 , and a support part 130 .

The processing unit 110 includes a milling spindle 111, a turret 112, and a spindle 113 and directly processes a workpiece. That is, the milling spindle 111 of the processing unit performs milling of the workpiece clamped on the spindle under the control of the controller, and the turret 112 performs turning processing under the control of the controller, and the spindle 113 clamps the workpiece. In one state, it rotates at various revolutions per minute (RPM) according to the control of the controller. If necessary, it is also possible to perform complex machining in which turning and milling are simultaneously performed.

The transfer unit 120 performs a function of transferring the processing unit according to the control of the control unit. The transfer unit 120 is formed of a ball screw and a servo motor or an LM guide and a servo motor to move the processing unit along the bed or column of the support unit.

The support part 130 is installed so that the processing part and the conveying part are movable. That is, the support unit 130 includes a bed 131 and a column 132, and a milling spindle, a turret, a spindle, a transfer unit, and the like are installed on the bed and the column.

The sensing unit 200 measures the temperature according to the thermal displacement of the structure. This sensing unit 200 may be formed as a temperature sensor. In addition, the sensing unit may be installed in at least one of a processing unit, a transfer unit, and a support unit. Specifically, as shown in FIG. 1, one or a plurality of them may be installed in various positions in the structure, or only one may be installed in a milling spindle, spindle, turret, bed, or column. Accordingly, the temperature change according to the thermal displacement of at least one part of the structure is measured by the temperature sensor of the sensing unit, and the result is transmitted to the thermal displacement correcting unit, so that the direct correction amount can be quickly and accurately calculated through the direct factor.

The thermal displacement correction unit 300 automatically calculates a correction amount according to the thermal displacement of the structural part in real time by machine learning through machine learning the indirect factor that changes in real time in addition to the direct factor and the direct factor.

In addition, the direct factor of the thermal displacement compensation device of the machine tool according to the present invention is the temperature according to the thermal displacement of the structure measured by the sensing unit, and the indirect factor may be the processing state and process mode that change in real time during the processing of the workpiece. . Specifically, the indirect factor may be a change in processing RPM, a motor load in the processing unit, a feed speed in the transfer unit, a motor load in the transfer unit, and a change in coolant temperature according to the processing state and process mode of the workpiece.

The control unit 400 immediately reflects the correction amount calculated by the thermal displacement correction unit to process the workpiece.

As such, the apparatus for compensating thermal displacement of a machine tool according to the present invention, as shown in FIG. 5, directly calculates the correction amount through a direct factor, which is the temperature according to the thermal displacement of the structural unit sensed by the temperature sensor of the sensing unit, and In addition to direct factors such as temperature due to thermal displacement of the machine, indirect factors such as spindle RPM, motor load, feed shaft load, and feed shaft speed that change in real time are changed through machine learning through machine learning according to the real-time changes in importance. The indirect correction amount is automatically calculated in real time, and the final correction amount due to the thermal displacement of the structure is automatically calculated in real time with the direct correction amount and the indirect correction amount. Therefore, it is possible to maximize the accuracy and reliability of processing and correction by performing correction in real time by minimizing the occurrence of time delay with the actual occurrence of thermal displacement.

1 to 3, the control unit 400 of the thermal displacement compensating device 10 for a machine tool according to the present invention includes a memory unit 410, a processing unit 420, an information collection unit 430, and a transmission and reception unit. 440, and a display unit 450. In addition, the control unit includes a numerical control (NC) or a computerized numerical control (CNC), and various numerical control programs are embedded therein. In addition, although not shown in the drawings, according to a preferred embodiment of the present invention, the control unit includes a main operation unit, and this main operation unit includes a screen display program and a data input program according to screen display selection, and a screen display program According to the output of the software switch, it displays the software switch on the display screen, recognizes the on/off of the software switch, and performs the function of giving input/output commands for machine operation.

The memory unit 410 stores information about a processing program, a driving program, and a process mode for performing machining of a workpiece.

The processing unit 420 performs processing of the workpiece according to the information stored in the memory unit and the correction amount received from the thermal displacement compensation unit.

The information collection unit 430 collects operation information of the processing unit and the transfer unit operated by the processing unit. That is, the information collection unit collects and stores operation information of the processing unit and the transfer unit that operate when the processing unit performs machining of the workpiece.

The transceiver 440 transmits the operation information collected by the information collection unit to the thermal displacement corrector or receives the correction amount calculated by the thermal displacement corrector.

The display unit 450 displays processing information of the memory unit, the processing unit, operation information collected by the information collection unit, and information transmitted and received through the transceiver. All of these display units include a liquid crystal display (LCD), a thin film transistor-liquid crystal display (TFT LCD), an organic light-emitting diode (OLED), and a flexible display. ), a 3D display, and an e-ink display.

As shown in FIGS. 1 to 2 and 4 to 5, the thermal displacement compensating unit 300 of the thermal displacement compensating device 10 for a machine tool according to the present invention includes a database unit 310 and a confirmation unit 320 , includes a determination unit 330, a calculation unit 340, a correction unit 350, and/or a correction unit 360.

The database unit 310 stores information for automatically calculating the correction amount according to the thermal displacement of the structure part in real time by machine learning the direct factor and the indirect factor through machine learning.

The confirmation unit 320 checks whether or not processing has started in the processing unit according to the information stored in the database unit and processing information and operation information received from the control unit.

The determination unit 330 determines the processing state and the process mode of the workpiece when processing is started in the processing unit according to the confirmation result of the confirmation unit. Although not necessarily limited thereto, these process modes include a turning mode that performs only turning of a workpiece, a milling mode that performs only milling of a workpiece, a combined mode that performs both turning and milling of a workpiece, or a user's designation. It may be configured to include at least any one or more of designation modes for performing selective processing on a workpiece according to the present invention.

The calculation unit 340 calculates the final correction amount according to the thermal displacement of the structural part according to the information stored in the database unit, the processing information and operation information received from the control unit, the confirmation result of the confirmation unit, and the determination result of the determination unit.

The correction unit 350 transmits the information stored in the database unit and the final correction amount calculated in the calculation unit to the control unit.

The correction unit 360 is a low-pass filter in case of overcorrection because the processing information, operation information, confirmation result of the confirmation unit, judgment result of the determination unit, and data stored in the machine learning data storage unit are small before the correction unit delivers the final correction amount to the control unit. (low pass filter) automatically corrects the final correction amount of the correction unit and transmits the corrected final correction amount to the control unit. Accordingly, it is possible to maximize processing precision by improving accuracy and reliability of the final correction amount. Such overcorrection may occur when there is too little information about accumulation data, operation information, processing information, direct factors, and indirect factors in the initial stage of machining a workpiece of a machine tool.

Although not necessarily limited to this, the memory unit 410 and the database unit 310 may be various storage devices such as ROM, RAM, EPROM, flash drive, hard drive, etc., and the storage function of the data storage unit on the Internet It can also be a web storage that performs.

In addition, both the control unit 400 and the thermal displacement compensation unit 300 are application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), and field programmable circuits (FPGAs). gate arrays), controllers, micro-controllers, microprocessors, and other electrical units for performing functions, but is not limited thereto.

In addition, the transmission/reception unit 440, the thermal displacement correction unit 300, and the sensing unit 200 of the control unit all comply with technical standards or communication methods for mobile communication (eg, Global System for Mobile communication (GSM), CDMA). (Code Division Multi Access), HSDPA (High Speed Downlink Packet Access), HSUPA (High Speed Uplink Packet Access), LTE (Long Term Evolution), LTE-A (Long Term Evolution-Advanced), etc.), Bluetooth™ A radio signal may be transmitted and received with at least one of the transmission/reception unit 440 of the control unit, the thermal displacement correction unit 300, the sensing unit 200, and an arbitrary server on the mobile communication network constructed according to the above. In addition, the transmission/reception unit 440, the thermal displacement compensation unit 300, and the sensing unit 200 of the control unit are all WLAN (Wireless LAN), Wi-Fi (Wireless-Fidelity), Wi-Fi (Wireless Fidelity) Direct, DLNA (Digital Living Network Alliance), wireless broadband (WiBro), wireless communication methods such as WiMAX (World Interoperability for Microwave Access) may transmit and receive wireless signals, but are not limited thereto. In addition, the transmission/reception unit 440 of the control unit, the thermal displacement correction unit 300, and the sensing unit 200 may all transmit and receive wireless signals in a short-distance wireless communication method. For example, the transmission/reception unit 440, the thermal displacement correction unit 300, and the sensing unit 200 of the control unit all use Bluetooth™, Radio Frequency Identification (RFID), Infrared Data Association (IrDA), Short-range communication may be supported using at least one of UWB (Ultra Wideband), ZigBee, NFC (Near Field Communication), Wi-Fi (Wireless-Fidelity), Wi-Fi Direct, and Wireless USB (Wireless Universal Serial Bus) technologies, , but not limited thereto.

In addition, as shown in FIG. 5, the database unit 310 of the thermal displacement compensating unit 300 of the thermal displacement compensating device 10 for a machine tool according to the present invention includes a basic data storage unit 311 and a received data storage unit. 312, a machine learning data storage unit 313, and a real-time data storage unit 314.

The basic data storage unit 311 stores data on direct coefficients, direct weights, and switching. The basic data storage unit may be manually input through a touch screen, keypad, or keyboard in the main operation unit of the thermal displacement correction unit.

The received data storage unit 312 stores data received from the transceiver.

The machine learning data storage unit 313 stores a machine learning program and accumulated data for performing machine learning through machine learning. A machine learning program for performing machine learning through machine learning stored in such a machine learning data storage unit is a neural network algorithm (convolution neural networks, CNN or recurrent neural networks, RNN), LSTM (long short term memory), or a multilayer perceptron ( Multilayer Perceptrons, MLP).

In addition, the accumulated data stored in the machine learning data storage unit stores data for up to 30 minutes at intervals of 10 seconds from the moment the processing unit operates according to the control of the control unit, the confirmation result of the confirmation unit, and the judgment result of the determination unit. . That is, when too much accumulated data, processing information, and operation information are stored, problems due to overfitting occur as shown in FIG. As it is installed, it is possible to promote miniaturization of the machine tool and reduce installation cost, maintenance cost, and manufacturing cost.

The real-time data storage unit 314 stores the data sensed from the sensing unit, the processing state of the workpiece, and the processing unit RPM according to the process mode, the motor load of the processing unit, the transfer speed of the transfer unit, the motor load of the transfer unit, and the amount of change in the coolant temperature. save in real time

In addition, as shown in FIG. 5, the calculation unit 340 of the thermal displacement correction unit 300 of the thermal displacement correction device 10 of the machine tool according to the present invention includes a direct correction amount calculation unit 341 and an indirect correction amount calculation unit ( 342), a final correction amount calculation unit 343, and an analysis unit 344.

The direct correction amount calculation unit 341 directly calculates the correction amount according to the confirmation result of the confirmation unit, the judgment result of the determination unit, the data stored in the basic data storage unit and the data stored in the real-time data storage unit, and the temperature according to the thermal displacement of the structure measured by the sensing unit. Calculate directly. In other words, the direct correction amount calculation unit is a direct factor according to the temperature according to the thermal displacement of the structural part, which is a direct factor in the same way as the existing technology according to the confirmation result, the judgment result, the basic data, the real-time data, and the temperature according to the thermal displacement of the structural part measured by the sensing unit. directly calculate Specifically, through a method such as multivariable linear regression, a direct correction amount is mechanically calculated through a direct coefficient, a direct weight, and a temperature according to the measured thermal displacement of the structural part.

The indirect correction amount calculation unit 342 includes the confirmation result of the confirmation unit, the judgment result of the determination unit, data stored in the basic data storage unit, data stored in the received data storage unit, data stored in the real-time data storage unit, and data stored in the machine learning data storage unit. , and the indirect correction amount is automatically calculated in real time through machine learning using machine learning according to the change of processing part RPM, processing part motor load, conveying speed of conveying part, motor load of conveying part, and coolant temperature. That is, indirect correction amount The calculation unit is based on accumulated data, received data, real-time data, driving information, processing information, direct and indirect factors through machine learning models such as multi-layer perceptrons with multi-layer input layers, hidden layers and output layers, RNN and CNN through neural network algorithms. By performing Supervised Learning, Unsupervised Learning, and Reinforcement Learning, the amount of indirect correction is calculated automatically, accurately, predictably, and reliably in real time.

The final correction amount calculation unit 343 calculates the final correction amount according to the direct correction amount calculated by the direct correction amount calculation unit and the indirect correction amount calculated by the indirect correction amount calculation unit. That is, the final correction amount calculation unit sums the calculated direct correction amount and indirect correction amount, calculates the final correction amount quickly, accurately, and in real time and transmits it to the control unit, and the control unit corrects the machining origin by the transmitted final correction amount, The machining of the workpiece is performed through calibration.

When the indirect correction amount calculation unit automatically calculates the indirect correction amount in real time through machine learning using machine learning in the indirect correction amount calculation unit, the analysis unit 344 automatically determines whether overfitting is in real time according to the confirmation result of the confirmation unit and the judgment result of the judgment unit. analysis, and if overfitting occurs as a result of the analysis, by selecting a machine learning model different from the machine learning model selected in the indirect correction amount calculation unit, the indirect correction amount calculation unit automatically recalculates the indirect correction amount in real time, and the recalculated indirect correction amount is the final correction amount forwarded to the calculator. In addition, the machine learning model selected by the analysis unit may be at least one of a neural network algorithm, LSTM, and a multi-layer perceptron.

Specifically, the analysis unit automatically, quickly and accurately solves and prevents overfitting problems arising from various data such as too much accumulated data, processing information, operation information, direct factors and indirect factors, as shown in FIG. Safety and reliability can be improved. In other words, as the accumulated data is stored in the machine learning data for up to 30 minutes at intervals of 10 seconds from the moment the processing unit operates according to the control of the control unit, the confirmation result of the confirmation unit, and the judgment result of the judgment unit, the overfitting problem is primarily processing precision and correction accuracy and reliability by performing correction in real time by minimizing the occurrence of time delay with the actual thermal displacement occurrence by reflecting the existing non-reflected part by preventing secondary recalibration through the analysis unit. can maximize

In addition, the thermal displacement compensation device for machine tools according to the present invention stores and manages accumulated data for machine learning for a certain period of time to prevent overfitting of indirect correction amounts by indirect factors. Manufactured in the form of an edge device, it can be easily and quickly installed on existing machine tools through compatibility of equipment to improve compatibility and usability, improve consumer satisfaction, and increase exports.

Referring to FIG. 6, a method for compensating thermal displacement of a machine tool according to the present invention will be described. As shown in FIG. 6, the method for correcting thermal displacement of a machine tool according to the present invention includes a data storage step (S1), a confirmation step (S2), a temperature measurement step (S3) with a sensing unit, an information collection step (S4), and information storage. Step (S5), judgment step (S6), direct correction amount calculation step (S7), indirect correction amount calculation step (S8), analysis step (S9), final correction amount calculation step (S10), correction step (S11), correction step ( S12). In each step, the operating process, operating principle, configuration, and contents of the device are the same as those of the thermal displacement compensation device for a machine tool according to the specification of the present invention. Hereinafter, with reference to FIG. be explained by

Information for processing the workpiece by automatically calculating the correction amount according to the thermal displacement of the structural part in real time by machine learning the direct factor and the indirect factor that changes in real time in addition to the direct factor, and immediately reflecting the calculated correction amount through machine learning save the data

After the data storage step (S1), it is confirmed whether or not to start processing the workpiece according to the previously stored information data and processing information and operation information received from the control unit.

After the confirmation step (S2), the temperature of the structural part is measured when the confirmation result of processing is started.

After the temperature measurement step (S3), if processing is started as a result of the confirmation, processing information, operation information, processing information, and temperature information are collected.

After the information collection step (S4), the collected information is stored as real-time data and accumulated data.

After the information storage step (S5), the current processing state and process mode of the workpiece are determined according to the previously stored information data, real-time data, and accumulated data.

After the determination step (S6), the correction amount is directly calculated according to the temperature according to the thermal displacement of the structure measured in real time according to the pre-stored information data and the determination result.

After the direct correction amount calculation step (S7), the indirect correction amount is measured in real time through machine learning using machine learning according to various changes in the processing mode and processing mode that change in real time during the processing of the workpiece according to the pre-stored information data and the judgment result. automatically calculated as

After the indirect correction amount calculation step (S8), when the indirect correction amount is automatically calculated in real time through machine learning using machine learning, overfitting is automatically analyzed in real time according to the confirmation result and the judgment result, and the analysis result is overfitting When this occurs, a machine learning model different from the selected machine learning model is selected when calculating the indirect correction amount, the indirect correction amount is automatically recalculated in real time, and the final correction amount is recalculated according to the recalculated indirect correction amount.

If overfitting does not occur in the analysis step (S9), the final correction amount is automatically calculated according to the calculated direct correction amount and indirect correction amount after the analysis step.

After the final correction amount calculation step (S10), the processing origin of the workpiece is corrected according to the calculated final correction amount.

After the correction step (S11), before transmitting the calculated final correction amount to the control unit, if the processing information, operation information, confirmation result, judgment result, and accumulated data are small and overcorrected, the final correction amount is automatically corrected with a low-pass filter and corrected Correct the machining origin of the workpiece according to the final correction amount. In other words, the correction step compares whether overcorrection has occurred, and if overcorrection occurs, the final correction amount is automatically corrected by a low-pass filter and transmitted to the control unit. If overcorrection does not occur, the final correction amount is transmitted to the control unit as it is through the existing correction unit. do. Such overcorrection may occur when there is too little information about accumulation data, operation information, processing information, direct factors, and indirect factors in the initial stage of machining a workpiece of a machine tool.

Referring to FIGS. 1 to 10 , the operating principle and operation process of the thermal displacement compensating device and compensating method for a machine tool according to the present invention will be described.

As shown in FIG. 1, the structure of a machine tool 1 such as a complex turning center or machining center for performing milling, turning, composite machining, or user-specified machining, and having various axis systems for this purpose, and performing composite machining. Various thermal displacements occur during the processing of blown workpieces or depending on the ambient temperature. Specifically, in FIG. 1, thermal displacement occurs during milling by the milling spindle, and at this time, X-axis and Z-axis corrections are required (A in FIG. 1). Thermal displacement occurs due to the spindle arranged in , and at this time, correction of the X-axis, Y-axis, and Z-axis is required. At this time, correction of the X, Y, and Z axes is required. (C in FIG. 1) In addition, thermal displacement occurs during turning due to the spindle and turret (D in FIG. 1). As the structure is transported along the transfer unit even when performing or not being processed, thermal displacement occurs due to the transfer speed and transfer load (E in FIG. 1).

In addition, the process mode is a turning mode that performs only turning of the workpiece, a milling mode that performs only milling of the workpiece, a combined mode that performs both turning and milling of the workpiece, or selectively for the workpiece according to user designation. It may be configured to include at least one or more of designated modes for performing processing. For example, if the time of the milling process becomes longer, the temperature sensor attached to the lower turret and the turret feed axis temperature sensor, which are required for the turning process, generate relatively less thermal displacement, and the effect of thermal displacement on the entire machine tool increases over time. become a negative factor. In addition, at this time, the motor load of the milling spindle and the motor load of the column relatively have more influence on the thermal displacement than the lower turret or turret feed axis, and thus become more central factors among indirect factors. In this way, in the process of machining in a machine tool, in addition to the direct factor according to the temperature conversion according to the thermal displacement of the structure, the indirect factor changes variously in real time according to the processing state and process mode of the workpiece, and the importance and weight are too many variables. It cannot be calculated through a simple calculation like the direct correction amount according to the direct factor as it changes as .

In this way, as described above, machining is performed including at least one of milling, turning, composite, or designated machining, and for this purpose, a complex turning center or machining center equipped with various axis systems and performing complex machining is a workpiece. In the case of machines, it is almost impossible to calculate the correction amount indirectly, or the reliability, accuracy, and predictability of the correction amount are remarkably low depending on the program model to be calculated, and problems may occur.

Accordingly, as shown in FIG. 5, the apparatus and method for compensating thermal displacement of a machine tool according to the present invention depend on the processing state and processing mode of the workpiece in addition to the direct correction amount reflecting the direct factor by temperature according to the thermal displacement of the structural part and the direct factor. The indirect correction amount is calculated by reflecting the change in processing part RPM, processing part motor load, conveying speed of the conveying part, motor load of the conveying part, and coolant temperature as indirect factors, and the final correction amount is calculated to perform processing. It is possible to maximize the precision of processing and the accuracy and reliability of correction by performing correction in real time by minimizing the occurrence of time delay with the actual occurrence of thermal displacement by reflecting the unreflected part of the .

Specifically, in order to solve the above-described problems, the present invention primarily arranges one or more sensing units in a structure unit as shown in FIG. 1, and as shown in FIGS. 2 to 4 and 6, the sensing unit is a structural unit. The temperature according to the thermal displacement of is measured and transmitted to the thermal displacement correction unit, and the direct correction amount calculation unit of the calculation unit calculates the correction amount primarily without time delay through direct coefficients and direct weights.

In addition, as shown in FIGS. 2 to 4 and 6, in the case of a multi-axis machine tool that performs milling, turning, and complex processes as described above, various thermal displacements of the structural part occur during processing of the workpiece, and thermal change In addition to the direct factor by the temperature according to the thermal displacement of the structural part, the correction part above is the processing part RPM that varies in real time according to the processing state and process mode of the workpiece, the motor load of the processing part, the conveying speed of the conveying part, the motor load of the conveying part, and the coolant Considering the change in temperature as an indirect factor, supervised learning based on accumulated data, received data, real-time data, operation information, processing information, direct factors and indirect factors that change with various variables and combinations at the moment in the indirect correction calculation unit of the calculation unit Supervised Learning), Unsupervised Learning, and Reinforcement Learning are performed to automatically, accurately, predictably, and reliably calculate the amount of indirect correction in real time.

In FIG. 9, the upper graph shows the case where there is no correction by generating accumulated data every 10 seconds (before in FIG. 9) and the case of correction through LSTM for 10 hours (lstm in FIG. 9) and the error value (error in FIG. 9) , and the lower graph in FIG. 9 shows the case of correction through LSTM for 10 hours (lstm in FIG. 9) and the error value when there is no correction by generating accumulated data for a total of 1 hour (before in FIG. 9) (Error in FIG. 9) is compared. As shown in FIG. 9 , when too much accumulated data is secured, overfitting occurs, resulting in an increase in error value, resulting in a decrease in accuracy and reliability.

To prevent this, after calculating the indirect correction amount in the indirect correction amount calculation unit, the analysis unit automatically calculates the indirect correction amount in real time through machine learning using machine learning in the indirect correction amount calculation unit. Overfitting is automatically analyzed in real time, and if overfitting occurs as a result of the analysis, a machine learning model different from the machine learning model selected in the indirect correction calculation unit is selected, and the indirect correction calculation unit automatically recalculates the indirect correction amount in real time. Calculate and transfer the recalculated indirect correction amount to the final correction amount calculator. Specifically, the overfitting problem is primarily caused by storing data for up to 30 minutes in machine learning data at intervals of 10 seconds from the moment the processing unit operates according to the control of the control unit, the confirmation result of the confirmation unit, and the judgment result of the judgment unit. processing precision and correction accuracy and reliability by performing correction in real time by minimizing the occurrence of time delay with the actual thermal displacement occurrence by reflecting the existing non-reflected part by preventing secondary recalibration through the analysis unit. can maximize

Then, as shown in FIGS. 2 to 4 and 6 , the final correction amount is calculated quickly, accurately, and in real time by adding the direct correction amount and the indirect correction amount calculated by the final correction amount calculation unit of the operation unit.

Thereafter, the final correction amount is transmitted to the correction unit, and the correction unit transmits it to the control unit again, and the control unit corrects the machining origin by the transmitted final correction amount and performs processing of the workpiece through compensation according to the thermal displacement of the machine tool.

That is, as shown in FIG. 7, an indirect correction amount is calculated by machine learning by machine learning with long short term memory (LSTM) with 360 accumulated data for a total of 20 hours (Hour), and then the final correction amount is performed (FIG. 7 In Lstm), it can be seen that the error value (error in FIG. 7) converges and the final correction amount is easily corrected as compared to the case where correction is not performed (before in FIG. 7) as time elapses.

Similarly, as shown in FIG. 8, with 360 accumulated data for a total of 20 hours (Hour), the indirect correction amount is calculated by machine learning by machine learning with Multilayer Perceptrons (MLP), and then the final correction amount is performed (in FIG. 8, MLP), it can be seen that the error value (error in FIG. 8) converges and the final correction amount is easily corrected as compared to the case where correction is not performed (before in FIG. 8) as time elapses.

As a result, in a state where accumulated data, real-time data, operation information, and processing information are sufficiently secured, the direct correction amount is directly calculated in the same way as the existing method, and the indirect correction amount is calculated in real time through machine learning regardless of the machine learning model. If the final correction amount is automatically calculated and quickly calculated, the time delay and the conventional time delay and While solving the accuracy and reliability problem, it calculates the final compensation amount quickly and precisely in a predictable state, corrects the machining origin, and processes the workpiece to improve machining precision, increase productivity, and promote convenience for users and operators. can

However, as shown in FIG. 10, if the machining is started in a complex machine tool having various axis systems to perform a milling process, a turning process, a complex process, and a designated process, and there is little accumulated data, optimization is not performed. Overcorrection occurs, as in That is, when sufficient accumulation data is not secured, for example, at the beginning of processing, overcorrection may occur in indirect correction amount calculation through machine learning based on machine learning due to lack of data. Accordingly, before the correction unit transfers the final correction amount to the control unit, if the processing information, the operation information, the confirmation result of the confirmation unit, the judgment result of the determination unit, and the data stored in the machine learning data storage unit are too small, a low pass filter (low pass filter) is used. ), the final correction amount of the correction unit is automatically corrected (part G in FIG. 10), the error value is corrected close to 0, and the corrected final correction amount is transmitted to the control unit, and processing is performed with thermal displacement correction with the corrected final correction amount. carry out Accordingly, it is possible to maximize the processing precision by improving the accuracy and reliability of the final correction amount.

As such, the apparatus and method for compensating for thermal displacement of a machine tool according to the present invention perform processing including at least one of milling, turning, composite, and designated processing, and for this purpose, various shaft systems are provided and complex processing is performed. In machine tools such as complex turning centers or machining centers, the thermal displacement compensator calculates the amount of correction through direct calculation through direct factor, which is the temperature according to the thermal displacement of the structural part sensed by the temperature sensor in the thermal displacement compensator, and the temperature according to the thermal displacement of the structural part. In addition to direct factors such as, indirect factors such as spindle RPM, motor load, feed shaft load, and feed shaft speed, which change in real time, are machine learned through machine learning to calculate the indirect correction amount, and finally, the direct correction amount and the indirect correction amount determine the structure of the structure. The final correction amount due to thermal displacement is automatically calculated in real time and reflected immediately to process the workpiece. As a result, time delay due to temperature deformation of existing structures and time delay due to the number and location of temperature sensors installed are prevented, and simply More accurate and reliable than inaccurate correction temperature estimation performed through machine learning. The final correction value for the correction of the processing origin by the thermal displacement of the structure is automatically calculated in real time, and the final correction amount is automatically reflected to process the workpiece. It can improve processing precision and productivity, prevent unnecessary defective products, conserve resources, and increase user and operator convenience.

In addition, the embodiments according to the present invention described above may be implemented in the form of program instructions that can be executed through various computer components and recorded on a computer-readable recording medium. The computer readable recording medium may include program instructions, data files, data structures, etc. alone or in combination. Program instructions recorded on the computer-readable recording medium may be specially designed and configured for the present invention, or may be known and usable to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floptical disks. medium), and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter or the like as well as machine language codes generated by a compiler. A hardware device may be modified with one or more software modules to perform processing according to the present invention and vice versa.

Specific implementations described in the present invention are only examples and do not limit the scope of the present invention in any way. For brevity of the specification, description of conventional electronic components, control systems, software, and other functional aspects of the systems may be omitted. In addition, the connection of lines or connecting members between the components shown in the drawings are examples of functional connections and / or physical or circuit connections, which can be replaced in actual devices or additional various functional connections, physical connection, or circuit connections. In addition, if there is no specific reference such as “essential” or “important”, it may not be a component necessarily required for the application of the present invention.

In addition, although the detailed description of the present invention described has been described with reference to preferred embodiments of the present invention, those skilled in the art or those having ordinary knowledge in the art will find the scope of the present invention described in the claims to be described later. It will be understood that various modifications and changes can be made to the present invention without departing from the spirit and technical scope. Therefore, the technical scope of the present invention is not limited to the contents described in the detailed description of the specification, but should be defined by the claims.

1: machine tool

10: thermal displacement compensation device

100: structural part

200: sensing unit

300: thermal displacement correction unit

400: control unit

Claims

In the thermal displacement compensation device of a machine tool for correcting thermal displacement during processing of a workpiece,

a structural unit in which a structure capable of relatively moving the workpiece and the tool for machining the workpiece is installed;

a sensing unit to measure the temperature of the structural unit;

A thermal displacement correcting unit that automatically calculates a correction amount according to the thermal displacement of the structure unit in real time by machine learning through machine learning a direct factor and an indirect factor that changes in real time in addition to the direct factor; and

The thermal displacement compensating device for a machine tool comprising a; control unit for performing processing of the workpiece by immediately reflecting the correction amount calculated by the thermal displacement compensating unit.
According to claim 1,

The direct factor is the temperature of the structure part measured by the sensing part,

The thermal displacement compensation device for a machine tool, characterized in that the indirect factor is a processing state and a process mode that change in real time during the processing of the workpiece.
According to claim 2,

The process mode is

a turning mode for performing only turning processing of the workpiece;

a milling mode in which only milling of the workpiece is performed;

a combined mode for performing both turning and milling of the workpiece; or

a designation mode for selectively processing the workpiece according to a user's designation; A thermal displacement compensation device for a machine tool comprising at least one of the following.
According to claim 2,

The structural part,

A milling spindle, a turret, and a machining unit provided with a spindle and directly processing the workpiece;

a transfer unit for transferring the processing unit; and

Thermal displacement compensating device for a machine tool comprising a;
According to claim 4,

The thermal displacement compensation device of the machine tool, characterized in that the indirect factor is a change in the machining part RPM, the processing part motor load, the conveying speed of the conveying part, and the motor load of the conveying part according to the processing state and the process mode of the workpiece.
According to claim 4,

The thermal displacement compensation device for a machine tool, characterized in that the sensing unit is installed on at least one of the processing unit, the transfer unit, and the support unit.
According to claim 4,

The control unit,

a memory unit for storing information about a processing program, a driving program, and a process mode for performing machining of the workpiece;

a processing unit configured to process the workpiece according to the information stored in the memory unit and the correction amount received from the thermal displacement correcting unit;

an information collection unit that collects operation information of the processing unit and the transfer unit operated by the processing unit; and

The thermal displacement compensating device for a machine tool, comprising: a transceiver for transmitting the operation information collected by the information collection unit to the thermal displacement compensating unit or receiving a correction amount calculated by the thermal displacement compensating unit.
According to claim 7,

The control unit,

The thermal displacement compensation device for a machine tool further comprising: a display unit for displaying the memory unit, the processing information of the processing unit, the operation information collected by the information collection unit, and the information transmitted and received through the transmission and reception unit.
According to claim 7,

The thermal displacement correction unit,

a database unit for storing information for automatically calculating a correction amount according to the thermal displacement of the structure unit in real time by machine learning the direct factor and the indirect factor through machine learning;

a confirmation unit for confirming whether or not processing has started in the processing unit according to the information stored in the database unit and processing information and operation information received from the control unit;

a determination unit for determining a processing state and a process mode of the workpiece when processing is started in the processing unit according to a confirmation result of the confirmation unit;

a calculation unit for calculating a final correction amount according to the thermal displacement of the structure unit according to information stored in the database unit, processing information and operation information received from the control unit, a confirmation result of the confirmation unit, and a determination result of the determination unit; and

and a correction unit for transmitting the information stored in the database unit and the final correction amount calculated in the operation unit to the control unit.
According to claim 9,

The database unit,

a basic data storage unit for storing data on direct coefficients, direct weights, and switching;

a received data storage unit for storing the data received from the transceiver;

a machine learning data storage unit for storing a machine learning program and accumulation data for performing machine learning through machine learning; and

A real-time data storage unit for storing the data sensed by the sensing unit, the processing state of the workpiece, and the amount of change in processing RPM, motor load of the processing unit, feed speed of the transfer unit, and motor load of the transfer unit according to the process mode in real time. A thermal displacement compensating device for a machine tool, characterized in that for doing.
According to claim 10,

The calculation unit,

A direct correction amount that directly calculates a correction amount according to the confirmation result of the confirmation unit, the judgment result of the determination unit, the data stored in the basic data storage unit, the data stored in the real-time data storage unit, and the temperature of the structure measured by the sensing unit. calculator;

The confirmation result of the confirmation unit, the determination result of the determination unit, data stored in the basic data storage unit, data stored in the received data storage unit, data stored in the real-time data storage unit, data stored in the machine learning data storage unit, and An indirect correction amount calculation unit that automatically calculates an indirect correction amount in real time through machine learning using machine learning according to changes in processing part RPM, processing part motor load, conveying speed of the conveying part, and motor load of the conveying part; and

and a final correction amount calculation unit for calculating a final correction amount according to the direct correction amount calculated by the direct correction amount calculation unit and the indirect correction amount calculated by the indirect correction amount calculation unit.
According to claim 11,

The calculation unit,

When the indirect correction amount calculation unit automatically calculates the indirect correction amount in real time through machine learning using machine learning, the overfitting is automatically analyzed in real time according to the confirmation result of the confirmation unit and the judgment result of the judgment unit, and the analysis result is overloaded When a sum occurs, a machine learning model different from the machine learning model selected by the indirect correction calculation unit is selected, the indirect correction amount calculation unit automatically recalculates the indirect correction amount in real time, and the recalculated indirect correction amount is transmitted to the final correction amount calculation unit. A thermal displacement compensating device for a machine tool, comprising: an analysis unit to do.
According to claim 11,

The thermal displacement correction unit,

Before the correction unit transfers the final correction amount to the control unit, if the processing information, the driving information, the confirmation result of the confirmation unit, the judgment result of the determination unit, and the data stored in the machine learning data storage unit are small, the low-pass signal is overcorrected. A correction unit for automatically correcting the final correction amount of the correction unit with a filter and transmitting the corrected final correction amount to the controller.
According to claim 11,

The accumulated data is the thermal displacement of the machine tool, characterized in that the storage of data for up to 30 minutes in a 10-second period from the moment the processing unit operates according to the control of the control unit, the confirmation result of the confirmation unit, and the judgment result of the determination unit. correction device.
According to claim 12,

The machine learning model selected by the analysis unit is at least one of a neural network algorithm, LSTM, and a multi-layer perceptron.
Information for processing the workpiece by automatically calculating the correction amount according to the thermal displacement of the structural part in real time by machine learning the direct factor and the indirect factor that changes in real time in addition to the direct factor, and immediately reflecting the calculated correction amount through machine learning storing data;

Checking whether or not processing of the workpiece has started according to pre-stored information data and processing information and operation information received from the control unit;

Measuring the temperature of the structural part when processing is started as a result of the check;

Collecting processing information, operation information, processing information, and temperature information when processing is started as a result of the confirmation;

Storing the collected information as real-time data and accumulated data;

determining a current processing state and process mode of the workpiece according to pre-stored information data, real-time data, and accumulated data;

Directly calculating a correction amount according to the temperature of the structure measured in real time according to pre-stored information data and a judgment result;

Automatically calculating an indirect correction amount in real time through machine learning using machine learning according to various changes in processing conditions and process modes that change in real time in the processing process of the workpiece according to pre-stored information data and judgment results;

automatically calculating a final correction amount according to the calculated direct correction amount and indirect correction amount; and

Compensating the machining origin of the workpiece according to the calculated final correction amount; thermal displacement compensation method of a machine tool comprising the.
According to claim 16,

After calculating the indirect correction amount,

When the indirect adjustment amount is automatically calculated in real time through machine learning using machine learning, the overfitting is automatically analyzed in real time according to the confirmation result and the judgment result, and if the analysis result is overfitting, the indirect correction amount calculation Selecting a machine learning model different from the selected machine learning model to automatically recalculate the indirect correction amount in real time and calculating the final correction amount according to the recalculated indirect correction amount; How to correct above.
According to claim 16,

After the correction step,

Before the calculated final correction amount is transmitted to the control unit, if the processing information, the operation information, the confirmation result, the judgment result, and the accumulation data are small and overcorrected, the final correction amount is automatically corrected by a low-pass filter and the corrected final Correcting the machining origin of the workpiece according to the correction amount; Thermal displacement compensation method of the machine tool, characterized in that it further comprises.
According to claim 16,

The direct factor is the temperature of the structure measured by the sensing unit,

The indirect factor is a change in processing RPM, processing motor load, feed speed of the transfer unit, and motor load of the transfer unit according to the processing state and process mode of the workpiece that change in real time during the processing of the workpiece Workpiece, characterized in that Thermal displacement compensation method of the machine.