JP5803261B2 - Thermal displacement correction method and thermal displacement correction apparatus for machine tool - Google Patents

Description

  The present invention relates to a thermal displacement correction method and a thermal displacement correction apparatus for a machine tool.

  A machine tool processes a workpiece by controlling the position of each drive shaft by a control device. In this machine tool, the structure may be thermally deformed due to heat generation due to internal factors such as machining by tools and rotation of the motor, and heat transfer due to external factors such as room temperature fluctuations in the installation environment. Since the thermal deformation of the structure affects the processing position, the processing accuracy may be reduced.

  Thus, for example, Patent Document 1 discloses a method of correcting a thermal displacement in a machine tool with high accuracy using a learning function of a neural network. This method corresponds to the output of each temperature measuring means based on the neural network theory from the output of a plurality of temperature measuring means arranged in each part of the machine tool and the output of the position measuring means arranged on the table. Create a teacher data table that predicts the thermal displacement of the machine tool. Then, when machining the workpiece, the thermal displacement amount of the machine tool is predicted based on the data table from the output of each temperature measuring means. The movement of the machine tool is corrected based on the prediction result.

  Further, for example, Patent Document 2 discloses a machine tool that can ensure high machining accuracy by correcting the amount of thermal displacement using a neural network. This machine tool stores the temperature information from the temperature sensor arranged in each element part of the machine part in the temperature storage means, and also stores the thermal displacement amount of each axis at this time in the thermal displacement amount storage means. Then, the temperature information from the temperature storage means is input, the thermal displacement amount is input from the thermal displacement amount storage means as teacher data, and learning is performed by a neural network to identify the function. At the time of actual control, a thermal displacement amount inference value of the mechanical unit is obtained based on the function and temperature information from the temperature sensor, and a correction value of the thermal displacement amount of each axis is obtained based on the inference value. Based on this correction value, the movement amount of each axis is corrected.

  Further, for example, Patent Document 3 discloses a thermal displacement correction device that can perform thermal displacement correction with high accuracy without using a neural network to reduce the productivity of a workpiece. This thermal displacement correction device receives temperature information from temperature sensors in each part of the machine tool and calculates a correction amount of thermal displacement between the tool and the workpiece using a neural network constant value. The neural network constant value is adjusted by learning the thermal displacement and temperature information read from the teacher data storage unit as teacher data. When machining the workpiece, the temperature information is received from each temperature sensor, and the thermal displacement amount of the machine tool is calculated based on the neural network constant value. Based on the calculation result, the movement of the machine tool is corrected.

JP-A-6-8107 JP 7-75937 A Japanese Patent Application Laid-Open No. 11-114776

  In the thermal displacement correction using the neural network described in each of the above patent documents, it is not possible to mathematically guarantee the accuracy of the thermal displacement correction for unlearned teacher data. Moreover, learning know-how is necessary, and stable learning results are not always obtained by over-learning, local solutions, or the like. In addition, there is a problem that physical design information such as dimensions and physical property values of the thermal displacement portion is not included, and is merely a numerical alignment with the measurement data.

  The present invention has been made in view of the above circumstances, and provides a thermal displacement correction method and a thermal displacement correction apparatus for a machine tool capable of performing highly reliable thermal displacement correction with respect to thermal deformation of the machine tool. With the goal.

(Machine tool thermal displacement compensation method)
Thermal displacement correction method for a machine tool (claim 1) the present invention is based on the temperature information from the temperature sensor disposed at a predetermined position of the machine tool, differ estimate possible of the machine tool thermal displacement 2 A method of correcting the thermal displacement of the machine tool generated during machining of a workpiece using a type of thermal displacement estimation process, wherein the heat of which processing time is short among the two types of thermal displacement estimation processes. A first thermal displacement estimation process executing step for continuously executing the displacement estimation process from the start of machining the workpiece, and a thermal displacement estimation process having a long processing time among the two types of thermal displacement estimation processes. A second thermal displacement estimation process execution step executed every predetermined time from the start of processing of the object, a first thermal displacement estimation value obtained in the first thermal displacement estimation process execution step, and the second thermal displacement estimation process execution The difference from the second thermal displacement estimated value obtained in the process is The difference between the thermal displacement estimated value comparison step for comparison with the stored threshold value which is the allowable range of thermal displacement of the machine tool and the thermal displacement estimated value obtained in the thermal displacement estimated value comparison step is the range of the threshold value. A correction value calculation step for obtaining a correction value for the command position of the moving body of the machine tool based on the NC program based on the first thermal displacement estimated value, and the correction value obtained in the correction value calculation step A correction step of correcting the command position and a machining stop step of stopping machining of the workpiece when the difference between the thermal displacement estimation values obtained in the thermal displacement estimation value comparison step is outside the threshold range; Is provided.

(Claim 2) Further, the thermal displacement estimation process executed in the first thermal displacement estimation process execution step is a process based on artificial intelligence or a process based on statistics, and is executed in the second thermal displacement estimation process execution step. The thermal displacement estimation process to be performed may be physical structure analysis.

(Machine tool thermal displacement compensation device)
Temperature compensation apparatus of a machine tool of a third aspect of the present invention uses different two putative possible thermal displacement estimation process thermal displacement, thermal displacement of a machine tool that occurs during processing of the workpiece And a temperature sensor disposed at a predetermined part of the machine tool, and based on temperature information from the temperature sensor, the processing time is short among the two types of thermal displacement estimation processes. Based on the temperature information from the first thermal displacement estimation processing execution means and temperature information from the temperature sensor, the thermal displacement estimation processing is performed out of the two types of thermal displacement estimation processing. Second thermal displacement estimation processing execution means for executing thermal displacement estimation processing for a long time every predetermined time from the start of machining of the workpiece, and threshold storage for storing a threshold that is an allowable range of thermal displacement of the machine tool And a first thermal displacement estimation process The difference between the first thermal displacement estimated value obtained by the execution means and the second thermal displacement estimated value obtained by the second thermal displacement estimation processing execution means is obtained, and the threshold value stored in the threshold storage means When the difference between the thermal displacement estimated value comparing means to be compared with the thermal displacement estimated value obtained by the thermal displacement estimated value comparing means is within the threshold value range, the NC program is based on the first thermal displacement estimated value. a correction value calculating means for obtaining a correction value for the commanded position of the moving body of the machine tool according to a correction means for correcting the commanded position by the pre-Symbol correction values Ru obtained by the correction value calculating means, the thermal displacement estimation Machining stop means for stopping the machining when the difference between the thermal displacement estimated values obtained by the value comparison means is outside the threshold range.

  (Claim 1) According to the present invention, the first thermal displacement estimation process has a shorter processing time than the second thermal displacement estimation process, and the thermal displacement estimation value has real-time characteristics. To be executed. On the other hand, since the second thermal displacement estimation process needs to process a lot of information, the processing time is longer than that of the first thermal displacement estimation process. Therefore, it is periodically executed from the start of machining the workpiece W. Thereby, the thermal displacement estimated value by the 1st thermal displacement estimation process can be monitored by the thermal displacement estimated value by the 2nd thermal displacement estimation process, and it can mutually complement and improve the reliability of a thermal displacement estimated value. It becomes possible.

  (Claim 2) By performing artificial intelligence or statistical processing as the first thermal displacement estimation processing, the thermal displacement estimation value can be obtained in real time. In addition, a highly reliable thermal displacement estimated value can be obtained by performing a physical structure analysis as the second thermal displacement estimating process.

  (Claim 3) According to the thermal displacement correction apparatus for a machine tool of the present invention, the same effects as those in the above-described thermal displacement correction method for a machine tool can be obtained.

It is a perspective view showing the whole machine tool composition concerning an embodiment of the invention. It is a figure which shows schematic structure of the machine tool of FIG. 1, and a thermal displacement correction apparatus. It is a flowchart for demonstrating operation | movement of the thermal displacement correction apparatus of FIG. It is a figure which shows the relationship between the 1st and 2nd thermal displacement estimated value and a threshold value.

(1. Machine configuration of machine tool)
As an example of the machine tool 1, a horizontal machining center will be described as an example and will be described with reference to FIGS. That is, the machine tool is a machine tool having three rectilinear axes (X, Y, Z axes) orthogonal to each other and a vertical rotation axis (B axis) as drive axes.

  As shown in FIGS. 1 and 2, the machine tool 1 includes a bed 10, a column 20, a saddle 30, a rotating spindle 40, a table 50, a turntable 60, a plurality of temperature sensors 70, and a control device. 80 and a thermal displacement correction device 90.

  The bed 10 has a substantially rectangular shape and is disposed on the floor. However, the shape of the bed 10 is not limited to a rectangular shape. On the upper surface of the bed 10, a pair of X-axis guide rails 11 a and 11 b on which the column 20 can slide is formed in parallel to each other so as to extend in the X-axis direction (horizontal direction). Further, the bed 10 is provided with an unillustrated X-axis ball screw for driving the column 20 in the X-axis direction between the pair of X-axis guide rails 11a and 11b. The X-axis ball screw is rotated. A driving X-axis motor 11c is disposed.

  Further, on the upper surface of the bed 10, a pair of Z-axis guide rails 12 a and 12 b on which the table 50 can slide extend in the Z-axis direction (horizontal direction) orthogonal to the X-axis direction and are parallel to each other. Is formed. Further, the bed 10 is provided with an unillustrated Z-axis ball screw for driving the table 50 in the Z-axis direction between the pair of Z-axis guide rails 12a and 12b. The Z-axis ball screw is rotated. A Z-axis motor 12c to be driven is disposed.

  A pair of X-axis guide grooves 21a and 21b are formed on the bottom surface of the column 20 so as to extend in the X-axis direction and in parallel to each other. In the column 20, a pair of X-axis guide grooves 21a and 21b are fitted on the pair of X-axis guide rails 11a and 11b via ball guides 22a and 22b so that the column 20 can move in the X-axis direction with respect to the bed 10. The bottom surface of the column 20 is in close contact with the top surface of the bed 10.

  Further, on a side surface (sliding surface) 20a parallel to the X axis of the column 20, a pair of Y axis guide rails 23a and 23b on which the saddle 30 can slide extends in the Y axis direction (vertical direction), and Are formed parallel to each other. Further, the column 20 is provided with a Y-axis ball screw (not shown) for driving the saddle 30 in the Y-axis direction between the pair of Y-axis guide rails 23a and 23b. The Y-axis ball screw is rotated. A Y-axis motor 23c to be driven is disposed.

  A pair of Y-axis guide grooves 31a and 31b are formed on the side surface 30a of the saddle 30 facing the sliding surface 20a of the column 20 so as to extend in the Y-axis direction and in parallel with each other. A pair of Y-axis guide grooves 31 a and 31 b are fitted into the pair of Y-axis guide rails 23 a and 23 b so that the saddle 30 can move in the Y-axis direction with respect to the column 20, and the side surface 30 a of the saddle 30 is aligned with the column 20. Are closely in contact with the sliding surface 20a.

  The rotary spindle 40 is rotatably provided by a spindle motor 41 accommodated in the saddle 30 and supports a tool 42. The tool 42 is fixed to the tip of the rotation main shaft 40 and rotates with the rotation of the rotation main shaft 40. Further, the tool 42 moves in the X-axis direction and the Y-axis direction with respect to the bed 10 as the column 20 and the saddle 30 move. The tool 42 is, for example, a ball end mill, an end mill, a drill, a tap, or the like.

The table 50 is provided on the pair of Z-axis guide rails 12 a and 12 b so as to be movable in the Z-axis direction with respect to the bed 10. A turntable 60 is supported on the upper surface of the table 50 so as to be rotatable about the B axis in the vertical direction.
The turntable 60 is rotatably provided by a B-axis motor 61 accommodated in the bed 10 and fixes the workpiece W by magnetic adsorption or the like.

  The temperature sensor 70 is attached to a predetermined part of the machine tool 1, that is, an arbitrary part of the bed 10, the column 20, the saddle 30, the rotary spindle 40, the table 50, and the turntable 60. As the temperature sensor 70, for example, a thermocouple or a thermistor is used.

  The control device 80 controls the spindle motor 41 to rotate the tool 42 and controls the X-axis motor 11c, the Z-axis motor 12c, the Y-axis motor 23c, and the B-axis motor 61 to control the workpiece W and the tool 42. Are moved relative to each other about the X-axis direction, the Z-axis direction, the Y-axis direction and the B-axis. The control device 80 includes a thermal displacement correction device 90 that performs thermal displacement correction of the machine tool 1. However, the thermal displacement correction device 90 is not limited to the one provided inside the control device 80, and can also be applied as an external device.

(2. Explanation of thermal displacement correction)
Next, correction of thermal displacement by the thermal displacement correction device 90 will be described. The thermal displacement correction by the thermal displacement correction device 90 is performed during the processing of the workpiece W, and two different types of thermal displacement that can estimate the thermal displacement of the machine tool 1 based on the temperature information from the temperature sensor 70. An estimation process is used. The two types of thermal displacement estimation processes are a thermal displacement estimation process with a relatively short processing time (first thermal displacement estimation process) and a thermal displacement estimation process with a relatively long processing time (second thermal displacement estimation process). ).

  The first thermal displacement estimation process includes artificial intelligence or statistical processing, and the second thermal displacement estimation process includes physical structure analysis. Examples of artificial intelligence include pattern recognition such as neural network, fuzzy reasoning, and genetic algorithm. Examples of statistical processing include statistical processing such as multiple regression, response surface methodology, and cluster analysis. Examples of the physical structure analysis include a finite element method and a boundary element method.

  Artificial intelligence or statistical processing is executed continuously from the start of machining of the workpiece W because the processing time is shorter than the physical structure analysis and the thermal displacement estimated value has real-time characteristics. On the other hand, since it is necessary to process a large amount of information, the physical structure analysis with a longer processing time than artificial intelligence or statistical processing is reliable in the thermal displacement estimation value. Executed. And by monitoring the thermal displacement estimated value by artificial intelligence or statistical processing with the thermal displacement estimated value by physical structure analysis, it becomes possible to complement each other and improve the reliability of the thermal displacement estimated value.

(3. Configuration of thermal displacement correction device)
Next, the thermal displacement correction device 90 will be described with reference to FIG. The thermal displacement correction device 90 includes a first thermal displacement estimation process execution unit 91, a second thermal displacement estimation process execution unit 92, a threshold storage unit 93, a thermal displacement estimated value comparison unit 94, and a correction value calculation unit 95. The correction unit 96 and the processing stop unit 97 are provided. Here, the 1st thermal displacement estimation process execution part 91, the 2nd thermal displacement estimation process execution part 92, the threshold value memory | storage part 93, the thermal displacement estimated value comparison part 94, the correction value calculating part 95, and the correction | amendment part 96 And the process stop part 97 can also be comprised by each separate hardware, and can also be set as the structure respectively implement | achieved by software.

Based on the temperature information from the temperature sensor 70, the first thermal displacement estimation process execution unit 91 continuously executes the first thermal displacement estimation process, for example, the thermal displacement estimation process using a neural network from the start of machining the workpiece W. To do.
Based on the temperature information from the temperature sensor 70, the second thermal displacement estimation processing execution unit 92 performs a second thermal displacement estimation process, for example, a thermal displacement estimation process by a finite element method, every predetermined time from the machining start W of the workpiece. To run.

The threshold value storage unit 93 stores a threshold value that is an allowable range of the thermal displacement amount of the machine tool 1. This threshold value is an allowable range of a difference between the first thermal displacement estimation value obtained by the first thermal displacement estimation processing execution unit 91 and the second thermal displacement estimation value obtained by the second thermal displacement estimation processing execution unit 92. As shown in FIG. 4, this threshold value is set for each execution time point t1, t2, t3... Of the second thermal displacement estimation process.
The thermal displacement estimated value comparison unit 94 includes a first thermal displacement estimated value obtained by the first thermal displacement estimation processing execution unit 91 and a second thermal displacement estimated value obtained by the second thermal displacement estimation processing execution unit 92. Is compared with the threshold value stored in the threshold value storage unit 93.

The correction value calculation unit 95 obtains a correction value for the machining command position based on the first thermal displacement estimated value when the difference between the thermal displacement estimated values obtained by the thermal displacement estimated value comparing unit 94 is within the threshold range. .
The correction unit 96 corrects the machining command position with the correction value obtained by the correction value calculation unit 95.
The machining stop unit 97 stops machining the workpiece W when the difference between the thermal displacement estimated values obtained by the thermal displacement estimated value comparing unit 94 is outside the threshold range.

(4. Processing by thermal displacement correction device)
Next, processing by the thermal displacement correction device 90 will be described with reference to FIG. The processing by the thermal displacement correction device 90 can be executed during machining.

  As shown in FIG. 3, when machining of the workpiece W is started (step S <b> 1), predetermined portions of the machine tool 1, that is, the bed 10, the column 20, and the saddle are continuously detected from the temperature sensor 70 and the displacement sensor (not shown). 30, temperature information and displacement information of any part of the rotation spindle 40, the table 50, and the turntable 60 are input (step S2). And the thermal displacement estimation process by a neural network is performed as a 1st thermal displacement estimation process (step S3), and the obtained 1st thermal displacement estimated value is memorize | stored (step S4). This process is based on a known technique and will be outlined below.

  The initial value of the weighting coefficient of the neural network of the predetermined part of the machine tool 1 described above is set. Then, the amount of thermal displacement is calculated from the input temperature information and the set initial value of the weighting coefficient of the neural network. Then, the input displacement information is compared with the calculation result as teacher data, and an inference error is calculated. When the inference error is large, a partial derivative of the weighting coefficient with respect to the error is obtained, and a new weighting coefficient is calculated.

  Then, the above calculation is repeated until the inference error becomes sufficiently small. When the inference error is sufficiently small, the thermal displacement amount can be inferred and the convergence is achieved. Then, the weighting coefficient at that time is reset as the weighting coefficient of the neural network. Thus, the first thermal displacement estimation process is completed.

  Then, it is determined whether or not a predetermined time for executing the preset second thermal displacement estimation process has elapsed (step S5), and when the predetermined time has not elapsed, the process returns to step S2 and described above. Repeat the process. On the other hand, when the predetermined time has elapsed, thermal displacement estimation processing by the finite element method is executed as second thermal displacement estimation processing (step S6), and the obtained second thermal displacement estimation value is stored (step S7). . This process is based on a known technique and will be outlined below.

  Three-dimensional model data of the entire machine tool 1 and condition data for thermal deformation analysis are set. This three-dimensional model data is obtained by, for example, dividing the three-dimensional shape data of the entire machine tool 1 created by using a three-dimensional CAD system into minute elements and converting this into model data necessary for analysis. The condition data includes the heat transfer coefficient, the linear expansion coefficient, the Young's modulus, the physical property value including the Poisson's ratio and the specific gravity of the material constituting the machine tool 1, and the heat transfer coefficient between the material and the atmosphere. is there.

  Based on the input temperature information and the set three-dimensional model data and condition data, a temperature distribution for each minute element is calculated. Then, based on the calculated temperature distribution, the thermal displacement amount of each microelement is calculated, and further, the thermal displacement amount of the predetermined part of the machine tool 1 described above is calculated. Thus, the second thermal displacement estimation process is completed.

  Then, a difference between the first thermal displacement estimated value and the second thermal displacement estimated value is obtained (step S7), and it is determined whether or not the obtained difference is within a threshold range (step S8). When the obtained difference is within a threshold range (for example, the difference D1, D2 between the first thermal displacement estimated value and the second thermal displacement estimated value at the time points t1, t2 shown in FIG. 4), Based on the estimated thermal displacement value, a correction value for the command position at the tip of the rotary spindle 40 is calculated (step S9).

  And the command position of the front-end | tip of the rotation spindle 40 is correct | amended with the calculated correction value (step S10). That is, the command position output from the control device 80 is corrected to the correction command position based on the correction value. Then, it is determined whether or not the thermal displacement correction is further executed (step S11). When the thermal displacement correction is further executed, the process returns to step S2 and the above processing is repeated, and the thermal displacement correction is not executed. Ends the thermal displacement correction program.

  On the other hand, in step S8, when the obtained difference is outside the range of the threshold value (for example, the difference D3 between the first thermal displacement estimated value and the second thermal displacement estimated value at time t3 shown in FIG. 4), The processing of the object W is stopped, and the thermal displacement correction program is terminated. Thereby, it can prevent that the workpiece W becomes defective. At this time, the processing is resumed after resetting the weighting coefficient of the neural network.

(5. Modifications)
In the above-described embodiment, the thermal displacement correction method uses two different types of thermal displacement estimation processing. However, three or more types of thermal displacement estimation processing may be used, and the reliability of the thermal displacement estimation value is further improved. It becomes possible to improve.

1: Machine tool 10: Bed, 11a, 11b: X-axis guide rail, 11c: X-axis motor 12a, 12b: Z-axis guide rail, 12c: Z-axis motor 20: Column, 21a, 21b: X-axis guide groove, 22a , 22b: Ball guide 23a, 23b: Y-axis guide rail, 23c: Y-axis motor 30: Saddle, 31a, 31b: Y-axis guide groove 40: Rotating spindle, 41: Spindle motor, 42: Tool 50: Table 60: Turn Table: 61: B-axis motor 70: Temperature sensor 80: Control device 90: Thermal displacement correction device 91: First thermal displacement estimation processing execution unit 92: Second thermal displacement estimation processing execution unit 93: Threshold storage unit 94 : Thermal displacement estimated value comparison unit 95: Correction value calculation unit 96: Correction unit 97: Processing stop unit

Claims (3)

Based on the temperature information from the temperature sensor disposed at a predetermined position of the machine tool, with the machine tool 2 kinds estimation differ capable of thermal displacement of the thermal displacement estimation process occurs during the machining of the workpiece A method for correcting thermal displacement of the machine tool,
Of the two types of thermal displacement estimation processing, a first thermal displacement estimation processing execution step for continuously executing thermal displacement estimation processing with a short processing time from the start of machining the workpiece;
Of the two types of thermal displacement estimation processes, a second thermal displacement estimation process execution step for executing a thermal displacement estimation process with a long processing time every predetermined time from the start of machining the workpiece;
The difference between the first thermal displacement estimation value obtained in the first thermal displacement estimation processing execution step and the second thermal displacement estimation value obtained in the second thermal displacement estimation processing execution step is stored in advance. A thermal displacement estimated value comparison step for comparing with a threshold value that is an allowable range of thermal displacement of the machine tool;
When the difference between the thermal displacement estimated values obtained in the thermal displacement estimated value comparing step is within the threshold value range, based on the first thermal displacement estimated value, with respect to the command position of the moving body of the machine tool by the NC program A correction value calculation step for obtaining a correction value;
A correction step of correcting the command position by the correction value obtained in the correction value calculation step;
When the difference in thermal displacement estimated value obtained in the thermal displacement estimated value comparison step is outside the threshold range, a machining stop step for stopping machining of the workpiece;
A thermal displacement correction method for a machine tool comprising: In claim 1,
The thermal displacement estimation process executed in the first thermal displacement estimation process execution step is a process by artificial intelligence or a process by statistics, and the thermal displacement estimation process executed in the second thermal displacement estimation process execution step is A thermal displacement correction method for machine tools, which is physical structure analysis. A device that corrects thermal displacement of a machine tool that occurs during machining of a workpiece using thermal displacement estimation processing that can estimate two different types of thermal displacement,
A temperature sensor disposed in a predetermined part of the machine tool;
Based on the temperature information from the temperature sensor, a first thermal displacement estimation that continuously executes a thermal displacement estimation process with a short processing time from the start of machining the workpiece among the two types of thermal displacement estimation processes. Processing execution means;
Based on temperature information from the temperature sensor, among the two types of thermal displacement estimation processes, a second thermal displacement that executes thermal displacement estimation processing with a long processing time at predetermined intervals from the start of machining the workpiece. An estimation processing execution means;
Threshold storage means for storing a threshold that is an allowable range of thermal displacement of the machine tool;
A difference between a first thermal displacement estimated value obtained by the first thermal displacement estimating process executing means and a second thermal displacement estimated value obtained by the second thermal displacement estimating process executing means; Thermal displacement estimated value comparison means for comparing with the threshold value stored in
When the difference between the thermal displacement estimated values obtained by the thermal displacement estimated value comparing means is within the threshold range, based on the first thermal displacement estimated value, with respect to the command position of the moving body of the machine tool by the NC program Correction value calculation means for obtaining a correction value;
And correcting means for correcting the commanded position by the correction value before Symbol correction values Ru obtained by calculation means,
When the difference between the thermal displacement estimated values obtained by the thermal displacement estimated value comparing means is outside the threshold range, a machining stopping means for stopping the machining;
A thermal displacement correction device for a machine tool comprising:

Images