JP2008087093A - Abnormality detecting device for machine tool - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an abnormality detecting device for a machine tool, which detects the presence/absence of abnormality even when acceleration/deceleration is carried out during operation, by detecting a changing tendency of a waveform. <P>SOLUTION: A vibration sensor 1 is fixed to a driving device X of the machine tool, and detects vibrations generated from the driving device X. A signal cut-out portion 2 cuts out a signal in a specific period during the operation of the machine tool of signals outputted from the vibration sensor 1 as a target signal. An evaluation value computing portion 3 obtains time series data made from an effective value per unit period about the cut-out target signal as an evaluation value. A determining portion 4 compares a normal range set in advance and the evaluation value during the operation of the machine tool by using the evaluation value during the normal operation of the machine tool, and determines that an abnormality is generated in the machine tool when the evaluation value during the operation deviates from the normal range. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an abnormality detection device for a machine tool that detects deterioration due to wear or the like of a drive part in a machine tool.

2. Description of the Related Art Conventionally, a technique for monitoring vibration of equipment has been proposed for early detection of equipment abnormal signs. For example, a vibration detector is provided to detect shaft vibration in a rotating device, and the standard deviation of the amplitude obtained by sampling the output of the vibration detector is obtained from a plurality from the present time to a predetermined time in the past. A technique for determining that there is an abnormal sign when the standard deviation is equal to or greater than a prescribed threshold value is known (see, for example, Patent Document 1).
JP-A-2-232529

  By the way, since the technique described in Patent Document 1 determines that an abnormality occurs when the standard deviation obtained for the amplitude is equal to or greater than a threshold value, the presence or absence of an abnormality is determined during a period in which the device is operating steadily. However, in a machine with acceleration / deceleration such as a machine tool, the standard deviation changes with time even in normal operation, so the technique described in Patent Document 1 cannot detect the presence or absence of an abnormality. There is a case.

  The present invention has been made in view of the above-described reasons, and its purpose is to detect abnormalities in machine tools that enable detection of the presence or absence of abnormalities even when acceleration / deceleration is involved during operation by detecting the tendency of waveform changes. It is to provide a detection device.

  According to the first aspect of the present invention, there is provided a signal detection unit for detecting a signal reflecting the operation of the machine tool, and a signal cut-out for cutting out a signal of a specific period during operation of the machine tool as a target signal from the signals output from the signal detection unit. An output value, an evaluation value calculation unit for obtaining an evaluation value obtained by quantifying the change in the waveform of the target signal, a normal range set in advance using an evaluation value during normal operation of the machine tool, and an evaluation value during operation of the machine tool And a determination unit that determines that the machine tool is abnormal when the evaluation value during operation deviates from the normal range.

  According to a second aspect of the present invention, in the first aspect of the invention, the signal cut-out unit obtains a representative value for each fixed unit period in the specific period, and the evaluation value calculation unit calculates the representative value for each unit period in the specific period. The time series data of representative values is used as an evaluation value, and the determination unit uses a plurality of evaluation values obtained during normal operation of the machine tool based on the distribution of representative values for each unit period. Set the normal range, compare the representative value with the normal range for each unit period when the machine tool is operating, and determine that the machine tool is abnormal when the representative value deviates from the normal range for the specified number of unit periods It is characterized by doing.

  According to a third aspect of the present invention, in the first or second aspect of the invention, the signal cutout section defines a start time of the specific period with an external signal that instructs the operation of the machine tool as a trigger. It is characterized by.

  According to a fourth aspect of the present invention, in the first or second aspect of the present invention, the signal cutout unit obtains a representative value for each fixed unit period, and at the end of the unit period, a plurality of front and rear parts on the time axis are obtained. The representative value in each unit period is compared with a prescribed start threshold, and the representative value in at least one unit period of the first prescribed number of unit periods on the backward side of the time axis is less than or equal to the start threshold, and the time axis Each of the second prescribed number of unit periods on the forward side is set to be a start condition, and a plurality of unit periods after satisfying the start condition are set as the specific period.

  According to a fifth aspect of the present invention, in the invention according to any one of the first to fourth aspects, the signal detection unit is a vibration sensor that detects vibration of the machine tool.

  According to a sixth aspect of the present invention, in any one of the first to fourth aspects, the machine tool uses a motor as a drive source, and the signal detection unit is a current sensor that detects a load current of the motor. And

  According to the configuration of the invention of claim 1, an evaluation value obtained by quantifying the waveform change of the target signal obtained during a specific period during operation of the machine tool is obtained, and the evaluation value is compared with the normal range. Since it is judged whether there is an abnormality, not only when the machine tool is operated at a constant speed, but also when it is operated with acceleration / deceleration so that the operation speed changes with time, or when the tool is replaced. Even if it is, the presence or absence of abnormality of the machine tool can be determined.

  According to the configuration of the invention of claim 2, since time series data of representative values for a plurality of unit periods is used in a specific period, it is evaluated whether each unit period is within a normal range based on the distribution of representative values. The vibration waveform can be evaluated by the time change of the representative value for each unit period, and the waveform change can be evaluated with a relatively small amount of calculation.

  Here, the effective value, the peak value of the envelope, the level of the specific frequency, and the like can be used as the representative value. However, in the embodiment described below, an example in which an effective value is used as a representative value is shown.

  According to the configuration of the invention of claim 3, since the start point of the specific period is determined by the external signal instructing the operation of the machine tool, the start point of the specific period during operation of the machine tool is set to substantially the same timing. It is possible to compare waveforms by matching the start time not only when there is almost no waveform change as in constant speed operation, but also when the waveform changes over time in a specific period. become.

  According to the configuration of the invention of claim 4, since the timing for extracting the target signal is determined based on the change tendency of the representative value for each unit period, a signal indicating the start time of the specific period cannot be obtained from the machine tool. But it is possible to automatically determine the starting point of a specific period. Moreover, since the start time is determined according to a certain rule, the determination of the start time of the specific period can be formulated and set with objectivity. As a result, even when the waveform changes over time in the specific period, the waveforms can be compared by matching the start time points.

  According to the configuration of the invention of claim 5, it is possible to detect the presence or absence of abnormality by monitoring the vibration of the machine tool. In particular, when the machine tool has a rotating portion, it is possible to detect an increase in vibration due to shaft runout or bearing wear.

  According to the configuration of the sixth aspect of the invention, since the load current of the motor that drives the machine tool is monitored, it is possible to detect an increase in load due to wear of the movable part. Further, since the load current of the motor is monitored, it is possible to detect whether there is an abnormality in the machine tool without using a sensor.

  As shown in FIG. 2, the machine tool described below includes a drive device X that uses a motor as a drive source, and is configured to removably attach a tool Y to the drive device X. It is assumed that the position and the operating speed of the driving device X are controlled according to a program. Examples of this type of machine tool include a single-function machine tool, such as a machining center or turning center, which automatically selects a tool to be used from a plurality of tools held in a tool magazine according to a program. It may be used. Hereinafter, a spindle including the housing 11 is assumed as the driving device X.

  As shown in FIG. 1, the abnormality detection device for a machine tool described in the present embodiment includes a vibration sensor 1 that is fixed to a housing 11 of a drive device X and detects vibration of the drive device X. The vibration sensor 1 includes a sensor. The apparatus main body 13 is connected via the line 12. The sensor wire 12 is attached to the vibration sensor 1 and is detachably connected to the apparatus main body 13 by a connector or the like. Further, it is desirable to use a shielded wire for the sensor line 12 so that noise components are not mixed into the target signal through the sensor line 12.

  As the vibration sensor 1, an acceleration pickup is used, and the vibration sensor 1 is attached in close contact with the outer surface of the housing 11 of the driving device X as shown in FIG. 2. In attaching the vibration sensor 1 to the drive device X, it is required that the operation of the drive device X is not hindered and the attachment state is not changed by the operation of the drive device X. Therefore, the vibration sensor 1 is fixed to the housing 11 of the driving device X by adhesion or welding. Further, in order to allow the vibration sensor 1 to be attached and detached, a configuration in which a holder that removably couples the vibration sensor 1 to the housing 11 may be employed.

  Since the electrical signal output from the vibration sensor 1 is a signal reflecting the operation of the work opportunity, the vibration sensor 1 functions as a signal detection unit. The signal output from the vibration sensor 1 is input to the signal cutout unit 2.

  When the signal cutting unit 2 can use a signal instructing the operation of the machine tool in order to determine the operation start time of the drive device X, that is, the start time of the specific period, the signal is output as an external signal. As a trigger. That is, the time point at which the operation start signal of the drive device X is given may be set as the start time point of the specific period. Further, in this case, since it is considered that a signal for stopping the operation of the drive device X is also given, the time point may be set as the end point of the specific period.

  On the other hand, when an external signal cannot be used, the fact that the output amplitude of the vibration sensor 1 changes greatly before and during the operation of the drive device X is used to specify the machine tool during operation according to the following procedure. A period signal is extracted as a target signal. Here, it is assumed that the driving device X repeats the operation and the stop every time the work to be processed by the tool Y is input.

  The output amplitude of the vibration sensor 1 is very small before the driving device X is operated, and the output amplitude of the vibration sensor 1 is larger than that before the driving device X is operating. Therefore, a unit period is determined in accordance with an appropriate time resolution necessary for cutting out the specific period, and the output of the vibration sensor 1 for each fixed unit period is monitored by the signal cutting unit 2. Here, as an example, the unit period is set to 10 ms. The signal cutout unit 2 uses the boundary of each unit period as a point of interest, and compares representative values (in this case, effective values) in a plurality of unit periods before and after the time axis with a specified start threshold. . The backward side of the time axis represents the past side, and the forward side of the time axis represents the future side.

  On the backward side of the time axis, the effective value is compared with the start threshold for the first specified number (for example, 3) unit periods, and on the forward side of the time axis, the effective value is calculated for the second specified number (for example, 30) unit periods. Compare the value to the start threshold. Here, at least one effective value in the backward unit period is equal to or less than the start threshold (condition A), and all effective values in the forward unit period are equal to or greater than the start threshold (condition B). Is a start condition, and when the start condition is satisfied, the point of interest is set as a candidate for the start time. In the starting point candidate, the difference between the effective values within each time corresponding to the same number of unit periods before and after the target point (for example, three) is obtained, and the boundary of the next unit period on the forward side on the time axis is obtained. The same processing is performed using as a point of interest. This process is repeated until a certain time (for example, 400 ms) from the point of interest that first satisfies the conditions A and B.

  On the other hand, when even one of the unit periods on the forward side of the target point is less than the start threshold, that is, when the condition B is not satisfied, the end point of the next unit period is set as the target point, and the conditions A and B are Determine if you are satisfied. In addition, when the condition B is satisfied but any of the unit periods on the backward side of the target point exceeds the start threshold value, that is, when the condition A is not satisfied, the second specified number (that is, 30) from the target point. A point of time separated by the unit period is taken as a point of interest, and it is determined whether conditions A and B are satisfied for this point of interest.

  As described above, after determining whether or not the above-described conditions A and B are satisfied for a certain period of time after the first point of interest, among the points of interest (starting point candidates) that satisfy both of the conditions A and B Thus, the candidate having the largest difference in effective value within the time corresponding to the same number of unit periods before and after the target point is used as the start time. If the determination of the start time is simplified, the first candidate may be used as the start time.

  On the other hand, when the start time of the specific period is automatically determined from the output of the vibration sensor 1 as described above, the end time can be a certain time after the start time. It is desirable to determine automatically from the output of the vibration sensor 1 as well. As for the end point, similarly to the start point, the boundary of the unit period is set as a point of interest, and the effective values in a plurality of unit periods before and after the time axis with respect to the point of interest are compared with a specified end threshold value. Here, the number of unit periods before and after the point of interest is different from the start time, the backward number is the third specified number (for example, 6), and the forward number is the fourth specified number ( For example, 10). Here, when the start time is obtained, the start time candidate is searched for on the forward side of the time axis, but when the end time is obtained, the start time candidate is searched for on the backward side of the time axis.

  For the detection of the end time, at least one effective value of the fourth specified number of unit periods on the forward side with respect to the point of interest is less than or equal to the end threshold (condition C), and all of the unit periods on the reverse side The effective value is equal to or greater than the end threshold (condition D) as an end condition, and when the end condition is satisfied, the point of interest is set as a candidate for the end time. In the end point candidate, after obtaining the difference between the effective values within each time corresponding to the same number of unit times (for example, 3) before and after the target point, the next unit on the backward side of the time axis The same processing is performed with the boundary of the period as the point of interest. This process is repeated until a certain time (for example, 200 ms) from the point of interest that first satisfies the conditions C and D.

  If the condition D is not satisfied, that is, if there is at least one unit period whose effective value is less than the end threshold among the third specified number of unit periods on the backward side from the point of interest, the above processing is repeated at the next point of interest. Further, when the condition C is not satisfied even if the condition D is satisfied, the time point when the movement is made to the backward side by a time corresponding to the number of unit periods obtained by subtracting 1 from the fourth specified number (for example, 9). Whether or not the conditions C and D are satisfied is determined as a point of interest.

  As described above, after determining whether or not the above conditions C and D are satisfied for a certain period of time from the first point of interest, among the points of interest (candidates for end time) that satisfy both the conditions C and D Thus, the end point candidate where the difference in effective value within the time corresponding to the same number of unit periods before and after the target point is the maximum is used as the start point. If the determination of the start time is simplified, the first candidate may be used as the start time.

  As is clear from the above description, in order to extract the start time point and the end time point from the output of the vibration sensor 1 in the signal cutout unit 2, the output of the vibration sensor 1 during a series of operations is stored. It is necessary to keep. Therefore, the signal extraction unit 2 needs a FIFO (first-in first-out) memory for storing the output of the vibration sensor 1 for a certain time (for example, 2000 ms).

  With the above-described operation, the signal cutout unit 2 can cut out the target signal for a specific period corresponding to the operation of the machine tool from the output of the vibration sensor 1, and the waveform change is evaluated for this target signal. In order to evaluate the change in the waveform of the target signal, that is, the change in the amplitude over time, the target signal is given to the evaluation value calculator 3.

  The evaluation value calculation unit 3 uses, as an evaluation value, time series data in which effective values of unit periods included in a specific period are arranged in the time progression direction. If the two time-series data match, it can be said that the target signals that are the sources of the two time-series data are almost the same. However, there are variations in the target signal, and there is an error in the cut-out position of the specific period by the signal cut-out unit 2, so that the evaluation values rarely match completely. If the time-series data is regarded as a vector, the norm between the target signal obtained when the machine tool is operating normally and the target signal obtained while the machine tool is operating reflects the degree of coincidence between the two. However, even when the machine tool is operating normally, the time series data is not constant, so it is not practical to use the norm as the degree of coincidence.

  Therefore, the evaluation value obtained by the evaluation calculation unit 3 is input to the determination unit 4, and the evaluation unit 4 determines the evaluation value when the machine tool is operating normally as pre-processing for inspecting whether there is an abnormality in the machine tool. A large number is collected, and a normal range of effective values is set based on the distribution of effective values for each unit period for the collected evaluation values.

  To set the normal range, the effective value distribution for each unit period is regarded as a normal distribution, the average value and the variance of the effective value are determined for each unit time, and the shape of the distribution is determined. On the other hand, a range obtained by multiplying the variance by an appropriate coefficient may be set as the normal range. If the normal range is set in this way, it can be compared with the normal range every time an evaluation value is obtained during operation in order to check the presence or absence of abnormality of the machine tool or the sign of abnormality. It can be detected quickly.

  On the other hand, when detecting signs of machine tool abnormalities, rapid detection is not always required, so for the evaluation values obtained for a plurality of specific periods during operation of the machine tool, the variance of the effective value for each unit period A value obtained by multiplying the variance obtained during normality by a coefficient may be used as a threshold, and a range below the threshold may be used as a normal range. In this case, if the variance obtained during operation of the machine tool exceeds a threshold value, it can be determined that there is an abnormality sign or abnormality.

  As described above, when the evaluation value obtained during operation of the machine tool deviates from the normal range, the determination unit 4 determines that an abnormality has occurred in the machine tool or a sign of abnormality has occurred, and the output unit 5 To inform the user of the abnormality. The output unit 5 may perform notification by characters using a display device in addition to a device that generates an indicator light and a notification sound.

  In the above example, the output of the vibration sensor 1 is used for the target signal. However, since the drive source of the drive device X is a motor, the load current of the motor can be used for the target signal, and the motor is servo controlled. In this case, the output of the encoder provided in the motor can be used as the target signal.

It is a block diagram which shows embodiment of this invention. It is a schematic block diagram same as the above.

Explanation of symbols

1 Vibration sensor (signal input part)
2 Signal Extraction Unit 3 Evaluation Value Calculation Unit 4 Judgment Unit 5 Output Unit 11 Housing 12 Sensor Line 13 Device Main Body X Drive Device Y Tool

Claims (6)

  A signal detection unit that detects a signal reflecting the operation of the machine tool, a signal extraction unit that extracts a signal of a specific period during operation of the machine tool from the signals output from the signal detection unit, and a target signal An evaluation value calculation unit that obtains an evaluation value obtained by quantifying the waveform change, and a normal range set in advance using the evaluation value during normal operation of the machine tool and the evaluation value during operation of the machine tool are compared. An abnormality detection device for a machine tool, comprising: a determination unit that determines that the machine tool is abnormal when the value is out of a normal range.   The signal extraction unit obtains a representative value for each unit period in the specific period, the evaluation value calculation unit uses time series data of the representative value for each unit period in the specific period as an evaluation value, and the determination unit The normal range for each unit period is set based on the distribution of representative values for each unit period using a plurality of evaluation values obtained during normal operation of the machine tool. 2. The machine tool according to claim 1, wherein the representative value is compared with the normal range for each period, and it is determined that the machine tool is abnormal when the representative value deviates from the normal range in a specified number of unit periods. Anomaly detection device.   The abnormality detection of the machine tool according to claim 1 or 2, wherein the signal cut-out unit defines a start time of the specific period with an external signal instructing an operation of the machine tool as a trigger. apparatus.   The signal cutout unit obtains a representative value for each fixed unit period, compares the representative value in a plurality of unit periods before and after the time axis with a specified start threshold for each end point of the unit period, The representative value in at least one unit period of the first specified number of unit periods on the backward side of the second side is less than or equal to the start threshold value, and the representative value of the second specified number of unit periods on the forward side of the time axis is all 3. The machine tool abnormality detection device according to claim 1, wherein a plurality of unit periods after the start condition is satisfied are set as the specific period with a start threshold value or more as a start condition.   5. The machine tool abnormality detection device according to claim 1, wherein the signal detection unit is a vibration sensor that detects vibration of the machine tool. 6.   5. The machine tool abnormality according to claim 1, wherein the machine tool uses a motor as a drive source, and the signal detection unit is a current sensor that detects a load current of the motor. Detection device.

Images