JPS5911418B2 - Post process process - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method of controlling the depth of cut of a machine tool using a post process gauge, and particularly to a depth of cut control device that is effective in the case of a long delay time.

A. In machine tools that continuously process workpieces, such as double-head surface grinders and centerless grinders, the workpiece during or immediately after processing is in continuous motion and has slight vibrations, so the measurement surface is unstable and the workpiece is damaged due to heat generation from grinding. The true dimensions cannot be obtained because of thermal expansion, dust and grinding chips from processing adhere to the measurement surface, insufficient time is required for measurement, space for installing the measuring instrument is limited, etc. Therefore, in-process gauging or post-process gauging with short delay times cannot provide satisfactory accuracy.

Therefore, a post-process gauge is required with a delay time long enough to eliminate the adverse measurement conditions described above. However, when the delay time becomes longer in a post-process gauge, many processed workpieces, so-called dead workpieces, accumulate between the processing position and the measurement position, and the workpiece dimensions as a control result may fall short of the target median value. A hunting phenomenon in which the value largely moves up and down at the center and a drift phenomenon in which the value is biased away from the target median value occur.

In order to solve the above-mentioned phenomenon, as a conventional technique, once the depth of cut of the machine tool is corrected, the dead workpieces that have accumulated since then are not measured, and the workpieces where the effect of the correction has been produced reach the measurement position. A correction method has been considered in which the control is interrupted for the duration of the delay time.

However, with this method, the amount of dimensional change of the machine tool within the delay time interval must be corrected at once, so it is necessary to increase the amount of correction at one time according to the length of the delay time. If the delay time is long compared to , the control will have large vertical fluctuations. SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to perform control with higher precision and less variation in cutting depth control of a machine tool using a post process gauge with a long delay time. To summarize the present invention, the method of the present invention uses a first control limit range shift that is approximately the same size as the depth of cut correction amount and shifts the control limit range for the workpiece dimension in the opposite direction every time the depth of cut is corrected. The first control limit range shift accumulated within the delay time interval is reset for each delay time interval, and the total correction amount for the delay time interval averaged over an appropriate number of delay time intervals is calculated. It consists of carrying out a second control limit range shift in which the control limit range is shifted every delay time interval equally in magnitude and the same in direction.

The details of the present invention will be described below with reference to figures and tables.

) FIG. 1 is a model diagram showing how the hunting phenomenon in dimensional control occurs when the delay time is long, that is, when there are many stuck workpieces (dead workpieces) between machining and measurement.

The dimensions of each workpiece are indicated by black circles, and in Fig. 1, a shows the state of the workpiece at the machining hole, and b shows the state of the workpiece at the measurement position.After the delay time Td, the workpiece at the machining hole appears at the measurement position. ing. In this figure, there are five dead works between Td. Dimensional changes due to thermal displacement of machine tools, tool wear, etc. are assumed to have a constant positive slope, the target median value is set at the 0 position, and the upper and lower control limits are set at U. C. L=+1μ, L. C. L
= 1μ. Here, control limit + UCL range =
The target median value is ゜゜.

Figure a. C. When a workpiece with L is machined, it is measured after Td in figure b, and when it exceeds the control limit range as shown in P, a correction command is issued and the depth of cut of the machine tool is corrected, resulting in a. It is corrected as indicated by y in the figure.

The magnitude and direction of the correction and its point in time are indicated by arrows.
However, since ν after this correction in figure a appears after Td in figure b, even if the correction is made, the effect of the sleeve correction is not reflected at all between the sleeve correction and Td, so in actual machine tools, The correction command continues to be issued even though the correction has already been received. As described above, excessive correction caused by the fact that the effect of correction is not reflected in the workpiece dimensions to be measured next causes a phenomenon in which the control result greatly hunts. In order to prevent such a hunting phenomenon, it has been considered in the past to stop the control consisting of the next measurement and the correction thereof until the effect of the correction appears in the measurement, and this is shown in FIG.

The settings on the drawing are the same as in FIG. When the control limit range is exceeded, as indicated by P in figure b, a correction command is issued, and the machining dimensions are corrected, as indicated by P in figure b. Once the correction is made, the control operation is stopped until it is reflected in the correction effect, that is, during Td, which is the shaded area on the time axis in figure b, and the control operation is resumed after Td, and then the control operation is resumed. When the measured value exceeds the control limit range as shown in Q, correction is made as shown in q. However, in this method, in order to keep the correction value within the control limit range, the amount of one-time dimensional correction must be set larger than the dimensional change of the machine tool within the delay time. Therefore, as the delay time Td becomes longer, the amount of correction at one time becomes larger, and as a result, such a control method has the disadvantage that the dimensional change becomes large and sawtooth-like. Furthermore, since the dimensional change rate of the machine tool itself changes over time, it is necessary to correct the amount of correction to be appropriate at each point in time. Although the above-mentioned drawbacks have been solved by the cutting control device for a machine tool according to the invention, which is shown in a block diagram as an exemplary embodiment in FIG. This will be explained with reference to the figure.

FIG. 3 shows a control state based on a first shift that shifts the control limit range for depth of cut correction to prevent hunting, and a first shift reset, and FIG. The control state is shown with a second shift in the content to shift the control limit range to correct for drift. In Fig. 3, a shows the time-series dimensional state of the workpiece at the cutting edge, and b shows the time-series dimensional state of the workpiece at the measurement position, and the delay time T
After d, the workpiece shown in figure a appears at the measurement position in figure b,
In this embodiment, the number of retained works (dead works) is five. It is assumed that the dimensional change of the machine tool is a constant positive slope, the target median value is set at the O position, and the upper and lower control limits are set at U. C. Llμ,L. C. It is set as L-1μ, and the control limit +U. C. L limit range = target median value ゛ ゜
It is.

Figure b shows one L. C. When the control limit range is exceeded in the positive direction, as shown by P in L, a correction command is issued to correct the machine tool's depth of cut by 1't in the negative direction, and the workpiece is corrected, as shown in V in figure a. .

By the way, when the depth of cut correction is carried out in the negative direction by 1μ, the feature of the present invention is that the control limit value range is changed in the same direction as the depth of cut correction amount in the opposite direction, that is, in the positive direction by 1μ V6 in Figure b. Shift it like that. Shifting the control limit value range each time the depth of cut is corrected is referred to as a first control limit value range shift. In this example, the target median value is shifted as a shift in the control limit value range, and the shifted target median value is shown as a dashed line in figure b, and the upper and lower control limits follow the shift and are shown as dotted lines. It is shown in The measurement value of the next workpiece is determined based on the control limit value range shifted by the depth of cut correction amount in this manner, and a correction command is issued based on the determination. This reflects the effect of the correction in the measurement of the next workpiece, and as a result, it is possible to prevent the hunting phenomenon caused by dead workpieces. The first control limit range shift is performed and accumulated for each correction, and is added in the same direction and subtracted in the opposite direction. This first shift is performed in consideration of dead workpieces for which the correction is not reflected, that is, the delay time Td. The first shift amount is reset to return to the initially set control limit value range. Although the resetting time point is shown in Fig. c, synchronization with respect to the start of machining is not necessarily required if the resetting is performed at an interval equal to the length of the delay time Td. However, control using only the first control limit range shift causes a drift phenomenon in which the control result is biased as a whole with respect to the target median value, as is clear from a in FIG.

The magnitude of this drift amount corresponds to the amount of dimensional change of the machine tool within the delay time Td, that is, (drift amount) = (delay time) x (speed of dimensional change of the machine tool). Therefore, in order to overcome such a drift phenomenon, the control limit value, in this embodiment, the target median value, is shifted by an amount corresponding to the above-mentioned drift amount. Drift phenomenon simply means a state in which the control results as a whole are biased from the target median value, so it is a state in which dimensional changes in the machine tool are canceled out by depth of cut correction. The above (drift amount) - (
delay time) x (dimensional change speed of the machine tool) - (total amount of depth of cut correction within the delay time interval). That is, in order to correct the drift phenomenon, a second control limit range shift is performed in which the control limit range is shifted equally in the same direction in the magnitude of the total depth of cut correction within the delay time interval. FIG. 4 shows the control results obtained by performing a second control limit range shift to correct the drift in the control state of FIG. 3.

The settings of various quantities in figures A and b are the same as in figure 3. In figure c, the target median value shifted by the second control limit range shift is shown. In Figure b, the workpiece size is determined using the control limit range shifted by combining the second shift and the above-mentioned first shift, which is set at the interval of delay time Td, and the depth of cut correction is controlled. Results are shown. The magnitude of the drift phenomenon in dimensional control cannot be grasped unless you look at a certain degree of time course, and shifting the next control limit value range by the total amount of depth of cut correction within one delay time is an effective way to suppress drift. This may result in a hunting phenomenon. Therefore, the control limit value range is shifted for each delay time interval using the total amount of cutting correction for one delay time, which is averaged over a period consisting of a plurality of delay times. In this embodiment, the drift correction time is synchronized at the time of the first control limit range shift reset, and each shift is calculated using the sum of the depth of cut correction amounts at one reset interval, which are averaged over two reset intervals. A second control limit range shift is carried out to perform drift correction every time the control is set.
For example, the total amount of depth of cut correction between the two settings from O to 2Td is (
Since 1μ x 1 time) = 1μ, the total amount of depth of cut correction between Td and 3Td at 2Td is (1μ x 3 times) = 3μ, using 0.5μ as the average value of 1 reset interval. 1
3Td with 1.5μ as the average value of reset interval
Sometimes a shift of 1.5μ is performed. As a result, the drift phenomenon in FIG. 3 is eliminated, as illustrated in FIG. b. By the way, the characteristics of dimensional changes in machine tools are based on the transient thermal displacement of the mechanical system due to heat generation in the motor, heat generation in the rotating spindle, rise in oil temperature, etc. immediately after the machine tool is started, and as the mechanical system reaches thermal equilibrium, the grinding wheel or There is a gradual transition to a steady dimensional change mainly due to tool wear, and the rate of change changes over time.

However, according to the above-mentioned Trinit correction method of the present invention, even if the dimensional change rate of the mechanical system changes over time, it is automatically corrected to an appropriate value, and stable control becomes possible. 1st~
In the embodiment shown in Fig. 4, correction was made each time the control limit range was exceeded, but in control using a post process gauge, dimensional variations among individual machined workpieces are averaged and controlled. In terms of stability, a method has conventionally been used in which a measurement result is corrected every time it exceeds the control limit range n times in a row.Of course, this method can be adopted in the five inventions of the present invention described so far. As further features of the present invention, the following improvements have been made.

Generally, if the number of times the control limit range is continuously exceeded in the same direction is increased, stability will be improved, but on the other hand, the tracking ability of the control response will be reduced. Therefore, the set value n of the number of times is
It is preferable that the value is automatically set to the optimum value for the dimensional change rate of the machine tool and measurement variations. In the present invention, it is determined whether there is insufficient control or excessive control based on the tendency of changes in the dimensions of the workpiece over time and the amount of dimensional correction that has been carried out, and the number of times the control limit value range is exceeded and the number of under-controls are determined according to the determination result. The number of times to do this is automatically set. In the embodiment of the present invention, determination of insufficient control or excessive control is performed as follows.

The successive average value of the workpiece dimensions that change over time within the delay time interval of the system is determined, and the cumulative value of the correction amounts made during that time is compared with the successive average value of the workpiece dimensions.
[(Successive average value of time-series dimensions) + (Cumulative value of depth of cut correction amount)] × (Direction of depth of cut correction) Target median value If this calculation result is equal to the target median value, the value at that point This means that the dimension correction has been performed so that the deviation of the workpiece dimension from the target median value becomes exactly 0 after the system delay time, so there is no excess or deficiency in the control amount.

If the calculation result is larger than the target median value, it is determined that the control is excessive, and if it is smaller than the target median value, it is determined that the control is insufficient. If it is determined that the control is insufficient, the number of times the control limit range is exceeded or under the control limit range in the next time interval is reduced according to a predetermined method. If it is determined that the control is excessive, the number of times the control limit range is exceeded or under controlled in the next time interval is increased according to a predetermined method. If it is determined that the control is appropriate, the number of times the control limit range is exceeded or exceeded will not be increased or decreased. With this method, stability can be improved without impairing the responsiveness and follow-up ability of dimensional control. Figure 3 shows the specific calculation method for the successive average value of the workpiece dimensions, which shows the measured workpiece dimensions PisPi+1...Pi+6 during the period from NTd to (n+1)Td.
When performing this with the target median value as O, the successive average value of the dimensions at this time is +1.77μ, and the sum of the depth of cut correction amounts is -1.
0μ and the direction of the depth of cut correction are negative.

The above discriminant is [(+1.77)+(-1.0)]×(
-1)=-0.77ku0. Therefore, NTd~(n+1)T
d is determined to be under control. Considering from the phenomenon aspect, +1
．． Correcting the depth of cut by -1μ for an average size of 77μ would result in insufficient control. The cumulative value of the depth of cut correction must be the cumulative value within a period corresponding to the delay time of the control system. The control state cannot be determined based on cumulative values that are too long or too short than the delay time. More generally, the following three ways can be considered to increase or decrease the number of times n. (1)
If there is insufficient control, reduce n, and if there is too much control, reduce n.
to the original setting value.

(2) Start from n-1, and in case of excessive control, increase to n
is increased, and in case of insufficient control, it is returned to n-1.

(3) Start with n=set value, and if control is insufficient, n is decreased by 1, and if control is excessive, n is increased by 1.

In the case of control adaptation, n is not increased or decreased. Incidentally, the present invention is effective when implementing the above-mentioned first shift and second shift in control using a post-process gauge with a long delay time.

FIG. 5 is a block diagram of the entire apparatus as an embodiment of the system of the present invention. A workpiece W is continuously machined in a machine tool 1 having a depth of cut correction device 9, and after a delay time Td necessary for accurate measurement after processing, it is moved to the position of a gauge 2 containing a differential transformer, for example, and there - Dimensions are measured using process gauges. A workpiece size measurement value signal consisting of an alternating current modulated according to the workpiece size is amplified by an amplifier 3 and converted to a direct current by a rectifier 4. For convenience of signal processing, the measured value signal is digitized by the AD converter 5, and all subsequent processed signals are considered as digital values. The measured value signal is compared with the control limit value range from the control limit value calculation circuit 12 in the control limit discrimination circuit 6, and when the measured value exceeds the control limit value range, a signal indicating this is sent to the number of times discrimination circuit 7. The number of times determination circuit 7 counts how many times the dimensional measurement value exceeds or falls below the control limit range, and operates the depth of cut correction command circuit 8 when the count reaches a predetermined number n. A depth of cut correction command including information as to whether the control limit value range has been exceeded in the positive or negative direction is sent to the depth of cut correction device 9 of the machine tool 1. The depth of cut correction device 9 performs a certain amount of correction (1 μ in the above-mentioned embodiment) for each command by advancing or retracting the depth of cut depending on the positive or negative information, thereby correcting the dimensions of the workpiece.
As a feature of the present invention, the commands from the depth of cut correction command circuit 8 are simultaneously sent to the first control limit range shift circuit 10. In the circuit 10, a shift amount corresponding to the magnitude and direction of the depth of cut correction amount is generated, and the shift amount is sent to the median value correction circuit 11 to shift the target median value by the amount. The shifted target median value and the upper and lower control limits (U.C.L, L.C.L) set from the outside are combined in the control limit value range calculation circuit 12 to obtain the shifted control limit. A value range is created. The shifted control limit value range is sent to the control limit determination circuit 6, and subsequent determination of workpiece dimension measurements is made by comparison with the shifted control limit value range. As a result, the hunting phenomenon can be prevented by control using the post process gauge. The shift by this feedback system is the above-mentioned first control limit range shift.
In the first control limit range shift amount calculation circuit 10, the first shift amount accumulated for each correction is calculated by a timer 15 that generates a pulse at every delay time Td.
It is reset every d intervals. Further, the correction command issued from the cutting correction command circuit 8 is further transmitted to the correction number storage circuit 13. In the circuit 13, the cumulative value of the number of correction commands within a period consisting of at least one predetermined number of delay times is stored (for example, if the predetermined number is 2 at 4Td in FIG. The cumulative value is 4 between
becomes. ), based on the stored number of times, the second control limit value shift amount calculation circuit 14 calculates
The average value of the total amount of correction for the delay time interval (in the above example, 4÷2−
2) is calculated and determined for each delay time interval based on the calculation timing from the delay time timer 15. Information on the calculated total correction amount (2μ in the above example) is sent to the median correction circuit 1.
1, and similarly to the first shift described above, the target median value is shifted, and as a result, the control limit value range is shifted. Here, the cumulative value of the first control limit range shift amount has already been reset, so this feedback system, which is a shift from O, is the second control limit value shift described above, and is controlled by the post process gauge. This is to correct the drift phenomenon. Further, in another invention of the present application, the predetermined number of times in the number of times determination circuit 7 is automatically set to an optimum value, which provides a method that further improves the above-mentioned method.

The configuration of the apparatus for carrying out the method is shown below. The work size measurement values from the AD converter 5 are converted from analog values to digital values, and are further sent to the measurement value storage circuit 16 and stored in chronological order. Based on this time-series measurement value, a delay time timer 15
A successive average value of the dimensions for each delay time interval is calculated based on the calculation timing. This calculation method is described in the first specification of the present application.
Details are shown on pages 6 to 17. At the same time, based on the number of corrections within the delay time interval from the correction number storage circuit 13, the cutting correction cumulative value calculation circuit 19 calculates the total amount of cutting correction performed within one delay time interval. Then, in order to determine whether the control state is excessive, suitable, or insufficient, the control state determination circuit 20 compares the output of the dimension successive average calculation circuit Cl7 and the output from the cutting depth correction cumulative value calculation circuit 19. The control state determination circuit 20 calculates [(successive average value of dimensions) + (cumulative value of depth of cut correction amount)] x (direction of depth of cut correction), and the calculation result is the median value of the target z. If it is equal to the target median value, it is determined that the control is appropriate, if it is larger than the target median value, it is determined that the control is excessive, and if it is smaller than the target median value, it is determined that the control is insufficient. The result of this determination is output to the number increase/decrease circuit 18, and the set values of the number of times over and under the limit values set externally in advance are increased when the control state is excessive, and are decreased when the control state is insufficient. Then, control is continued based on the reset number of under and over counts. The block diagram shown in FIG. 5 is one embodiment of the present invention, and the block diagram shown in FIG.
It is also possible to have a computer perform the functions 14, 16, 17, 18, 19, and 20.

The system described so far assumes that the depth of cut correction is a fixed amount.

However, the present invention is not limited to quantitative control;
It is also possible to apply the present invention to proportional control in which correction is performed using a depth of cut correction amount depending on the magnitude of the deviation of the measurement results from the target median value. As an effect of the present invention, as shown in Fig. 4, the dimensional control results for the workpiece are highly accurate control in which hunting phenomena, drift phenomena, etc. are minimized, and dimensional changes are gentle around the target value. is obtained. Another effect of the present invention is that even when the dimensional change rate of the machine tool has a transient phenomenon, appropriate drift correction can be obtained from the transient state to the steady state, and it is possible to automatically correct the drift without impairing the tracking ability of the control system. Control results that are suitable for the current situation can be obtained.

Another advantage of the present invention is that since the post process gauge is highly accurate based on a long delay time, the measurement accuracy of the workpiece dimensions is highly reliable, so the measurement results obtained by the post process gauge can be simply used as dimensions. In addition to being used for control purposes, the same measurement results can also be used to simultaneously and reliably perform tasks such as eliminating defective products and sorting processed workpieces into several size ranges. It is possible.

[Brief explanation of the drawing]

FIG. 1 is a model diagram of dimensional control using a post-process gauge with a long delay time when a hunting phenomenon occurs.

Claims (1)

[Claims] 1. Dimensions of a processed workpiece are measured by a post process gauge at a predetermined delay time, and the measured value by the post process gauge falls within a predetermined control limit range by at least 1
If the depth of cut exceeds a predetermined number of times or more, the depth of cut of the machine tool is corrected by a predetermined correction amount in the direction of exceeding a predetermined control limit value range, or by a correction amount according to the deviation from the target median value. In a method for controlling a depth of cut for a machine tool in which a command is issued and a depth of cut correction is performed, the control limit value range is shifted in a direction substantially the same as the amount of correction and in a direction opposite to the direction of the correction for each depth of cut correction command. performing a first control limit range shift for each delay time interval, resetting the first control limit range shift amount accumulated within the delay time interval, and further delaying at least one or more predetermined number of intervals. A second control limit value that shifts the reset control limit value by approximately the same amount as the amount obtained by averaging the sum of correction amounts performed within the time interval by the number of delay time intervals and in the same direction as the averaged amount.
A control method for a machine tool characterized by performing a control limit value range shift of. 2 Dimensions of the processed workpiece are measured using a post-process gauge at a predetermined delay time, and the measured value by the post-process gauge falls within a predetermined control limit range by at least 1
If the depth of cut exceeds a predetermined number of times or more, the depth of cut of the machine tool is corrected by a predetermined correction amount in the direction of exceeding a predetermined control limit value range, or by a correction amount according to the deviation from the target median value. In a method for controlling a depth of cut for a machine tool in which a command is issued and a depth of cut correction is performed, the control limit value range is shifted in a direction substantially the same as the amount of correction and in a direction opposite to the direction of the correction for each depth of cut correction command. performing a first control limit range shift for each delay time interval, resetting the first control limit range shift amount accumulated within the delay time interval, and further delaying at least one or more predetermined number of intervals. a second control for shifting the reset control limit value to approximately the same size as the amount obtained by averaging the sum of correction amounts performed within the time interval over the number of delay time intervals and in the same direction as the averaged amount; A limit value range shift is performed, and the difference between the sequential average value of the time-series measured dimensions within the plurality of delay time intervals and the target median value, and the amount of cutting correction within the plurality of delay times. 1. A method for controlling depth of cut for a machine tool, characterized in that the control state of the depth of cut correction amount is determined by comparing the total sum of the amount of depth of cut, and the predetermined number of times is changed based on the determination result.