JPH0295544A - Detection device for breakdown of tool - Google Patents

Abstract

PURPOSE:To perform a tool detection of high reliability by subjecting to fuzzy inference the output value discriminating the frequency component of an AE (acoustic emission) signal and the output value of an abruptly rising AE signal and providing a fuzzy inference mechanism for outputting a tool breakdown detecting signal. CONSTITUTION:An AE sensor 3 is provided near the body W to be worked of a machine tool main body 1 for machining the body W to be worked by the tool 2 of a drill, etc. An abnormal cutting or drill breakdown is inferred by integrally judging by the fuzzy inference mechanism 17 in a tool breakdown detection means 10 with the discrimination signal and rise signal of the frequency components of the detection signal from this AE sensor 3 as the input value. As a result, the drill breakdown or cutting abnormality can be detected with high reliability even in case of any disturbance and it is prevented to stop a machine tool with erroneous detection.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of the Invention) The present invention relates to an apparatus for detecting tool breakage in a machine tool, and more particularly to an apparatus for detecting tool breakage using fuzzy inference.

(Summary of the Invention) In the present invention, an acoustic emission (hereinafter referred to as rAEJ) signal of a broken tool such as a drill is judged by fuzzy inference, so that the breakage can be detected with high reliability.

(Prior Art and Problems) Conventionally, damage to a tool such as a broken drill of a machine tool has been detected by monitoring the AE multiplier signal. This AE multiplied signal is the ultrasonic frequency vibration generated by a solid object (in this case, a tool such as a drill), and tool breakage detection using this AE multiplied signal detects the unique ultrasonic frequency vibration generated when the tool breaks. It is supposed to be done.

That is, an example of the AE multiplied signal waveform when the drill breaks is shown in Fig. 6, and it rises rapidly at the time of breakage,
After that, it appears to gradually decrease. In addition, when broken 2
FIG. 7 shows the power spectrum of the AE multiplied signal caused by the electrical noise of the device (noise associated with opening and closing of the solenoid) during normal operation (normal operation). As is clear from this figure, when the drill breaks, the maximum frequency is approximately 300 KHz, and under normal conditions, the maximum frequency is approximately 50 KHz. Therefore, in the conventional tool breakage detection device using an AE multiplied signal, a pseudo AE signal generator that can generate an AE multiplied signal containing the frequency component at the time of breakage is installed at the tool position of the workpiece before the machining operation.
The AE double signal generated from this workpiece installation is received by the AE sensor installed on the stand and the received wave level is set, and after the start of work, the AE double signal frequency component output at the time of breakage detected by this AE sensor suddenly rises. The tool breakage output signal is configured to be detected based on the AND with the AE multiplied signal rising signal output value (for example, Japanese Patent Laid-Open No. 61-82161).
Publication No.).

However, in the conventional tool breakage detection device described above, the AE multiplied signal frequency identification means and rise detection means at the time of breakage are simply binary level comparisons and comparisons with a reference value, respectively, so other causes may occur. The resulting AE multiplied signal can accurately separate signals generated from chips on the workpiece, electrical noise associated with the opening and closing of a solenoid, impact sound when an object contacts the workpiece, etc., and perform highly reliable tool breakage detection. The problem was that it was difficult.

Additionally, the installation of the AE sensor has the problem that it is difficult to accurately detect tool breakage and compensate for changes in cutting conditions such as tool type and rotational speed due to differences in position, drill diameter, etc. .

(Object of the Invention) The present invention has been made by focusing on the above-mentioned conventional problems, and the installation of the AE sensor can result in AE multiplied signal variations due to differences in position or type of tool. It is an object of the present invention to provide a highly reliable tool breakage detection device that can reliably detect tool breakage without being affected by cutting conditions or the like.

(Structure and Effects of the Invention) In order to achieve the above object, the present invention has a structure including a machine tool main body for machining a workpiece with a tool such as a drill, and a machine tool provided near the workpiece on the machine tool main body. In a tool breakage detection device comprising an acoustic emission sensor and a detection means for detecting breakage of the tool based on a detection signal of the acoustic emission sensor, the detection means includes an output value that identifies a frequency component of the acoustic emission signal. The present invention is characterized by being provided with a fuzzy logic mechanism that performs fuzzy inference on the output value of the acoustic emission signal that rises rapidly and outputs a tool breakage detection signal.

According to the present invention, tool breakage is detected by comprehensively determining the identification signal and the rising signal of the AE multiplied signal frequency component using fuzzy inference with the above configuration, so that highly reliable tool detection can be performed. be. Therefore, it is possible to effectively prevent inconveniences in which tool breakage is erroneously detected and the machine tool stops abnormally even though the tool is not broken.

(Explanation of Examples) Hereinafter, the present invention will be explained based on illustrated embodiments.

FIG. 1 is a block diagram showing a schematic configuration of the apparatus of the present invention, in which a drilling machine 1 is shown as the main body of the machine tool. A workpiece W is placed on the bed 1a of the drilling machine 1, and a drill 2 as a tool is provided above the workpiece W. Further, an AE sensor 3 is provided near the workpiece W on the bed 1a.

The detection means 10 includes an analog switch 11 that connects and disconnects the analog detection signal of the AE sensor 3 based on an output from a fuzzy logic mechanism (described later), an amplifier circuit 12 that amplifies the signal from the analog switch 11, and this amplifier 12. As shown in FIG. 7, the amplified detection signal is passed through a first band pass filter (B P
F) 14a and the normal peak value of 50KHz
A second band pass filter (BPF) 14b for passing the frequency band, and a first band pass filter (BPF) 14b for detecting the output values of both of these BPFIs 4a and 14b. Second detector 15a, 15
b, a differentiation circuit 16 for differentiating the output value of the first detector 15a to obtain an AE multiplied signal that rises rapidly, the output value (V, ) of the first detector 15a, and the output of the second detector 15b. value(
v2) and the output value (V3) of the differentiating circuit 16 as input values, and a fuzzy inference mechanism 17 for outputting the breakage of the drill 2 by fuzzy inference, and this fuzzy inference mechanism 1
A first output circuit 18a that sends an abnormal signal to an alarm (not shown) based on the output value from 7, a second output circuit 18b that reports breakage of the drill 2, and the first detector 15.
It is comprised of a breakage abnormality level indicator 19 that displays the breakage abnormality level from a, and a cutting level indicator 20 that displays the cutting level (normal level) from the amplifier 12.

Among these, the fuzzy inference mechanism 17 is a characteristic component of the present invention, and here performs a three-man power one-output type inference of variable 3 in the condition part and variable 1 in the conclusion part. That is, the membership function corresponding to the condition part is formed from the trapezoidal and triangular models shown in FIGS. 2-4.

These membership functions are
A pseudo AE signal generator 30 is placed at the drill position of the workpiece W, and the pseudo AE signal generator 30 is driven via the drive circuit 32 with the value set by the level setting device 31, so that it can be used when the drill 2 breaks or during cutting. The same power spectra are generated and can be easily created based on these.

Here, based on the AE multiplied signal power spectrum in the 300KHz and 50KHz bands, the value obtained by converting the power into voltage (
A membership function whose input value is Vl) (V2) has been obtained (Fig. 2-3), and a member whose input value is the rising value (V3) obtained by differentiating the AE multiplied signal in the 300 KHz band. The Ship function has been obtained (Figure 4).

FIG. 5 shows the membership functions of the conclusion section created based on predetermined production rules. As an example of inference using these membership functions, if AE
Double signal vl is small (PS).

If V2 is medium (PM) and V3 is medium (PM), there is no drill breakage, but there is a high possibility that there is an abnormality (clogged chips, etc.). Alternatively, it has been concluded that if the AE multiplication signal Vl is medium (PM), V2 is medium (PM), and V3 is large (PL), there is a high possibility of drill breakage.

Therefore, a value based on this conclusion is sent from the fuzzy inference mechanism 17 to the first output circuit 18a or the second output circuit 18b, and abnormal cutting or breakage of the drill is notified to the outside.

As described above, in this embodiment, the value that identifies the frequency component of the AE multiplied signal and the detected value of the rapidly rising signal are used as input values to make a comprehensive judgment using fuzzy inference, and abnormal cutting or drill breakage is inferred. As a result, even if there is a disturbance component, drill breakage or cutting abnormality can be detected with high reliability, and inconveniences such as stopping the machine tool due to erroneous detection can be effectively prevented.

In the above-described embodiments, a drill is used as an example of the tool, but it goes without saying that a cutter such as a cutter may also be used.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing a schematic configuration of the apparatus of the present invention, FIGS. 2 to 5 are membership functions, FIG. 6 is an AE waveform diagram when a drill breaks, and FIG. 7 is an AE multiplied signal power spectrum diagram. 1... Machine tool body 2... Drill (tool) 3... AE sensor 10... Detection means 11... Analog switch 12... Amplifiers 14a, 14b... Bandpass filters 15a, 15
b-・1st. Second detector 16...Differential circuit 17...Fuzzy inference mechanism

Claims (1)

[Claims] 1. A machine tool body for machining a workpiece with a tool such as a drill, an acoustic emission sensor provided in the vicinity of the workpiece on the machine tool body, and a detection signal of the acoustic emission sensor. In the tool breakage detection device, the detection means includes an output value that identifies the frequency component of the acoustic emission signal and an output value of the acoustic emission signal that rises rapidly. A tool breakage detection device comprising a fuzzy logic mechanism that performs fuzzy reasoning and outputs a tool breakage detection signal.