JPS59146741A - Abnormality diagnostic device upon cutting - Google Patents

Abstract

PURPOSE:To maintain the accuracy of abnormality diagnosis by a method wherein a comparator, which does not treat a difference due to the difference of comparing time as the abnormality of cutting if there is a model data same as a load datum upon actual cutting in the memorized model data within a predetermined allowable time range, is provided in the device. CONSTITUTION:In case of the actual cutting of a work to be monitored, the magnitude of a cutting load is detected by a detector 2 and is converted into digital by a converter 3 while this signal is inputted into a coincidence detector 22 as the cutting load datum 6. On the other hand, a retrieve time 24, corresponding to a time 17 obtained by a time generator 20, is generated by a retrieve time generator 21. The corresponding model datum 7 among the retrieve times 24 is taken out of a memory 4 and the coincidence between the datum 6 is detected by the coincidence detector 22. In case the signal from the detector 22 is non-coincident, a new retrieve time 24, advanced or delayed by a given time based on the time 17, is generated by the generator 21. The renewed retrieve time 24 is monitored by an allowable time threshold detector 23 and a load abnormality signal 13 is generated only when it has exceeded a predetermined renewal time, thus, the abnormality of cutting is diagnosed. Accordingly, the accuracy of the diagnosis may be secured within the time.

Description

[Detailed Description of the Invention] The invention is filled with an abnormal feeling when performing actual cutting based on seven chill works.

Equipment for diagnosing abnormalities in the cutting tool U system in machining #'
i It has the circuit shown in FIG. In Figure 1, cut t]
Regarding the IJ model work, the cutting load 1 such as the machine tool's spindle motor load, feeding motor load, vibration, etc. is detected by the cutting load detector 2, and the digital converter 3 converts it into a digital signal and calculates the magnitude and time of the load. The current time generated by the generator 20 is stored in the memory 4. After that, regarding the monitored workpiece of "actual cutting n1", the digital converter 3 obtains a digital signal (Ei) indicating the thickness of the cutting load, and the cutting model stored in the memory 4 corresponding to the time at that time. The thickness of the load on the workpiece was compared with the comparator 5, and if there was a difference between the two by a predetermined value or more, a cutting abnormality was diagnosed and an abnormal output δ was output.

However, it is difficult to match the load data regarding the model data and the time of Fi 7, and Fukiri St+
Due to slight changes in the dimensions of the workpiece or variations in speed, there may be some hyperactivity between the two data sets relative to the time axis. (21st Shudani Teru) For this reason, Kiri 131
``Model work cut 1j1'' The load data record 1 indicates that the start time and the monitored workpiece cut 14υ/benefit opening time are different, and the cut 1 recorded in the memory 4 in Fig.
u Model work load Cheetah and supervision vs. unwork cutting load data 6 are considered to be different load data even if they are 1 river - work i'L time 8 and 11.5 When 90, it is abnormal and tool 1 diagnosis I'm about to do it.

In this device, time 8 and ff, 71ii
l Load data 7 of the cutting model workpiece whose time of 9 is stored in the memory 4 or cutting load of the monitored workpiece -7-Evening 6
This change is detected by a sudden change in time, and even if the difference between the two data detected by the comparator 5 during that time of change is larger than normal, it is not considered abnormal. This prevents misdiagnosis. As a concrete circuit, as shown in FIG. Difference with load Hooter '7? Load difference signal @ by load difference detector 11
14 I7 and take out 7 - power, time from time generator 20 and digital R converter 3 j? The unit time load change non-detector No. 10'
Therefore, the change rate between I+5 of the load data becomes 7 and the abnormality T-11-11 constant standard total 15 times and the load difference signal 14 and the abnormality 'l'Ij constant standard li/! Comparing 15 with Daikomukaibetsuki 12, at the time of this 54 change, Himonken 8B + L
The 11th hour was stopped due to a misdiagnosis due to the 4th wide setting. Therefore, the conventional apparatus has the disadvantage that the abnormality diagnosis is 1 to 11 times when no one is present at any given time.

Therefore, the present invention aims to improve the abnormality diagnosis accuracy at the time 6' when the load data changes sharply by looking at the faulty VC ship of the child series.
The purpose is to provide a blood device for diagnosing different shrines at the time of incision 3υ.

In order to achieve this goal, the invention is load data when cutting a model work. As a note, in a device that detects cutting abnormalities by comparing the actual cut 81'' with the M upper tertator from time to time, Within range, already memorized. Is there any model data that is the same as the load data during fruit cutting? Is there a comparator that searches and if the same data is found, the difference between the seven delta data and the load data due to the difference in comparison time is not treated as a cutting abnormality? It is characterized by being equipped.

Here, ヰλ4 diagram? The present invention will be described in detail with reference to the following.

In addition, in FIG. 414 VC, the same parts as in FIGS. 1 and 3 are given the same reference numerals. In FIG. 4, the cutting load 1 when cutting the model workpiece and the cutting load 1 when cutting the monitored workpiece during actual cutting are shown. The cutting load signal from the detector 2 for C detection is converted into a digital signal by the digital converter 3. Converted into this digital signal! 1ilJth 1'' load data is manually input to the storage device 4 and stored together with the time from the time generator 20. The model data 7 based on the model work of Zuno et al. and the actual cutting load data 6 are input to a coincidence detector 22 to detect coincidence between the data and the cutting. If the data do not match in the match detector 22 of C, the search time generator 2 ], ? Drive 1, then I4,
When searching at a constant time using the time 17 from the clock generator 20 as a reference, 1■124f! : Set 7 and search for ml storage 71
Now. If the search time exceeds the predetermined allowable time (71), that is, within the search time, a load abnormality signal is output from the boundary crossing point detector 23 (8t).

This coincidence detector 22. Search time generator 21, tri
A time/open time crossing point detector 23 and a comparator 5 are constructed.

Cutting of the model workpiece and actual cutting of the monitored workpiece at the same time? Distinction (7 times a little rfl L, <Ij'l bright.

Cutting model work 81" Thickness of load 1? Detector 2
Then, the digital change is digitized by the digital converter 3, and the digital signal M of C is stored in the memory device 4 along with the L:D time from the time generator 20. Repeat with model work cutting redundancy f1. L,h-
What about model data, which is load data, and time? It's true that I only remember C.

In the case of actual cutting of the desired workpiece, the actual cutting F'l
11 Workpiece Cutting 111'' The load thickness formula is detected by the detector 2, and then converted into dizotal by the digital converter 3, and this digital signal is used as the cutting load dichronized signal by the coincidence detector 2.
2. Manpower. On the other hand, the search time generator 21 generates a search time 24 corresponding to the current time 17 obtained by the time generator 20. During this search time 24, the corresponding model Cheetah is taken out from the storage device 4. A coincidence detector 111 22 detects whether the extracted model data 7 and the load data 6 mentioned above match. If the signals from the coincidence detector 22 do not match, a search time generator 2 is used to generate a new search time 24 that is shifted forward or backward by a certain period of time with respect to time 17. Memory device 4 of this model work
Reading from 〃) Detection of match and mismatch 1 is repeated, and the search time is 24? Update. Changing the search time 24 to allowable time crossing point detector 23
However, when the 0 supervisor exceeds the predetermined update time, a load error 1p; G4 M 13 occurs and a cutting error occurs.
Examine.

The cutter del workpiece is cut by C. The load rust was 1 EC kg higher than the fixed standard set by Itomi Publishing (oh)), and it was 1! Rather than judging it as J4N, the time sword 1 etc. are determined in advance and the time is 8.
11M's range [44 (within 6 [: greatness)
Check the force and force to see if there is a rotational load with the supervising home work among the loads of the cutting model work in -7L, and if it is detected that there is the same load, then the cutting load of the cutting model work = t t = ? f open tag 11 time and monitoring z
1 Since the cutting monitoring start time of the unworked workpiece differs from J)', it is determined that the abnormality monitoring standard (F['4k has been exceeded) and is not diagnosed as an abnormality. 9, it is also different from the first part + +i't. According to the present invention, it is possible to prevent IA diagnosis due to different cutting start times of soaks.As described in May, according to the present invention, the time to start storing the cutting load of the cutting model workpiece and the cutting of the monitored workpiece can be prevented. Misdiagnosis can be prevented even if the monitoring start time is different, and the accuracy of cut r4υ abnormality diagnosis of the monitored workpiece can be maintained constant over the entire monitoring area without becoming dizzy like with conventional methods. It can be done in one shift and is very useful for products such as Road Tanabata.

[Brief explanation of drawings]

Figures 1 to 3 show an example of a conventional abnormality diagnosis device. Figure 11 is a schematic block diagram, Figure 2 is a load curve diagram, and Figure 3 is a block diagram showing a portion of Figure 1 in detail. Figure, 4th
The figure is a block diagram showing an embodiment of an abnormality diagnosis device during cutting according to the present invention. In the figure, 21 is a search time generator, 22 is a matching detector, and 23 is a permissible time boundary crossing point detector. m (1 Sen)

Claims (1)

[Claims] →; Write down the model data that is the load data when cutting the del work, and compare the load data during actual cutting with the model data from time to time.
The abnormality detection device detects that there is load data that is the same as the actual cutting load data from among the model data already stored within a predetermined time tolerance from the actual cutting time. Search if there is a difference between the model data and the load data due to the difference in comparison time.
;': indicates cutting abnormality and 11. An abnormality diagnosing device at the time of t'iIJ, characterized in that it is equipped with two comparators.