JP2006017753A - Device and method for detecting abnormal noise - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To surely and easily detect partial discharges of electrical equipment or abnormal noises from mechanical equipment, irrespective of the presence or absence of a barrier, also by reducing influence due to noises. <P>SOLUTION: Noises are actively silenced by collecting abnormal sound generated from an abnormal sound detecting object by a parabola 11 and converting it into a sound signal through a microphone 12; converting this sound signal into a digital signal through an amplifier 13, a filter 14, and an A/D converter 15; performing FFT operation by a DSP 16, based on this digital signal; identifying the abnormal sound and also obtaining control sound that has the same amplitude as the periodic noises identified by the DSP and is in opposite phase therewith; converting this control sound into a sound signal by an A/D converter 17, a filter 18, and an amplifier 19; emitting it from a loudspeaker 20 toward the abnormal sound detecting object; and making the sound signal interfere with the noises existing in the periphery or the inside of the abnormal sound detecting object. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an abnormal sound detection apparatus and a detection method for detecting an abnormal sound caused by partial discharge of an electric device and an abnormal sound caused by operation of a mechanism device by an acoustic method.

  Various electrical devices such as transformers, circuit breakers, switches, and rotating machines have minute void defects in their insulators, and the electric field concentrates on those parts, causing weak discharge called partial discharge. appear. Further, partial discharge occurs due to the influence of contamination on the insulator surface. Due to these partial discharges, the insulator of the substation equipment is deteriorated, which may lead to dielectric breakdown after a long period of time.

  As a method for detecting the above discharge phenomenon (partial discharge: PD), (1) an electrical method in which a measuring instrument is directly connected to the main circuit of the object to be detected and (2) elasticity associated with the occurrence of the partial discharge. An acoustic method for detecting wave vibration and discharge sound can be mentioned.

  In the former electrical method, the amount of charge due to partial discharge in the detection object is regarded as a voltage generated in the detection impedance Z, as shown in an equivalent calibration / measurement circuit in FIG. In the figure, Q is a generated partial discharge charge amount, Z is a detection impedance, Vd is a voltage generated in the detection impedance by partial discharge, Ck is a coupling capacitor, Qcal is a known charge amount (injection during calibration), and Ca is a detection target. Capacitance, Cb is the capacitance inserted in series in the detection target defect portion, Cc is the capacitance of the detection target defect portion, and when the partial discharge of the charge amount Q occurs in the detection target Further, the voltage generated at both ends of the detection impedance Z cannot be obtained by calculation based on the detection object capacitance Ca, the coupling capacitor capacitance Ck, the detection impedance Z and its frequency characteristics. Therefore, the sensitivity adjustment (calibration) of the measuring instrument is performed by injecting a known charge amount into the object to be detected.

  As a representative example of the latter acoustic method, there is an Acoustic Emission (hereinafter referred to as AE) method. The AE method is a method for estimating a partial discharge occurrence value by detecting an elastic wave mainly transmitted to a metal surface or the like by a sensor and performing signal processing.

In addition, there is a method of capturing a partial discharge by sound (for example, see Non-Patent Document 1). In this document, as shown in FIG. 5, a discharge sound from a discharge source is collected by a parabola 1 and converted into a sound signal by a microphone 2, the sound signal is amplified by an amplifier circuit 3, and a high-pass filter 4 is used for a high frequency band. After extracting only, detecting by the envelope detection 5 and shaping the waveform by the low-pass filter 6, FFT (fast Fourier transform) is performed by a DSP (Digital Signal Processor) 7 to identify the discharge sound.
IEA-98-35, "Development of acoustic diagnostic equipment in substation facilities", Suecho et al.,

(1) Problems with an electrical method in which a measuring device is directly connected to the main circuit of the detection object for detection For measurement, it is necessary to connect the measuring device directly to the detection object. Furthermore, a power supply facility for applying a test voltage is also required, and unless this is the case, field products cannot be diagnosed by this method.

(2) Problems with acoustic diagnostic methods that detect acoustic wave vibration and discharge sound due to the occurrence of partial discharges The AE method directly or indirectly detects the elastic skin due to the discharge phenomenon during operation, but indirectly with switchboards, etc. As an automatic detection method, measurement is performed by attaching an AE sensor to the side panel of the switchboard. In this case, while it is possible to detect an elastic wave associated with a discharge phenomenon inside the switchboard, it is difficult to specify in which part the discharge phenomenon occurs. Furthermore, the influence of noise (white noise) is large, and it is difficult to extract a signal accompanying a small discharge phenomenon from the noise. In addition, when a single device such as an instrument transformer (VT) inside the switchboard is targeted, insulation of the AE sensor becomes a problem.

  When applying the method of capturing the discharge phenomenon by sound to a switchboard, etc., if there is a barrier in the panel, if there is a gap in the barrier, diagnosis can be made with sound leaking from that gap, but it is completely sealed. If so, it is necessary to open the barrier. In addition, taking out a small discharge sound from noise noise (an effective technique for identifying that the sound is a discharge sound) is also an issue, and it is considered that there is a limit to application to the diagnosis of the switchboard.

  The object of the present invention is to reliably and easily detect partial discharges by an acoustic method, regardless of the presence or absence of a barrier and to reduce the influence of noise and noise, and to reliably and easily detect abnormal noise generated in a rotating machine, etc. An object of the present invention is to provide an abnormal sound detection device and a detection method for detection.

  In order to solve the above-mentioned problem, the present invention generates an abnormal control sound having the same amplitude as that of periodic noise and emits the sound toward the abnormal sound detection target, thereby detecting the abnormal sound detection target. It is designed to interfere with noise noise existing around or inside, thereby actively silencing noise noise and detecting abnormal sounds due to partial discharge or abnormal sounds due to the operation of mechanical devices. And features method.

(1) An abnormal sound detection device that acoustically detects abnormal sound due to partial discharge of electrical equipment or abnormal noise due to operation of a mechanical device in an environment where noise noise exists,
Abnormal sound generated from an abnormal sound detection object is collected by a parabola and converted into a sound signal by a microphone. The sound signal is amplified and signal processed to be converted into a digital signal. Based on the digital signal, FFT is performed by a DSP. Abnormal sound detection means for performing calculation and identifying abnormal sound;
A control sound having the same amplitude and opposite phase as that of the periodic noise noise identified by the DSP is emitted toward the abnormal sound detection target and interferes with the noise noise existing around or inside the abnormal sound detection target. An abnormal sound detecting apparatus comprising a silencing means that actively silences noise and noise.

(2) An abnormal sound detection method for acoustically detecting abnormal noise due to partial discharge of electrical equipment or abnormal noise due to operation of a mechanical device in an environment where noise noise exists,
Abnormal sound generated from an abnormal sound detection object is collected by a parabola and converted into a sound signal by a microphone. The sound signal is amplified and signal processed to be converted into a digital signal. Based on the digital signal, FFT is performed by a DSP. In the abnormal sound detection process to calculate and identify abnormal sound,
A control sound having the same amplitude and opposite phase as that of the periodic noise noise identified by the DSP is emitted toward the abnormal sound detection target object to interfere with noise noise existing around or inside the abnormal sound detection target object. An abnormal sound detection method characterized by having a silencing process of actively silencing noise noise by means of the above.

  As described above, according to the present invention, the control sound having the same amplitude as that of the periodic noise noise is generated in the opposite phase, and is emitted toward the abnormal sound detection target. Because it interferes with noise noise existing inside, thereby actively silencing noise noise and detecting abnormal sound due to partial discharge or abnormal sound due to the operation of the mechanical device, regardless of whether there is a barrier, In addition, it is possible to reliably and easily detect partial discharges of electrical equipment by reducing the influence of noise and noise, and to reliably and easily detect abnormal sounds generated in rotating machines and the like.

  Conventionally, the discharge sound caused by the occurrence of partial discharge is collected by a parabola, converted into a sound signal by a microphone, and the sound signal is identified by FFT operation through an amplifier circuit and various filters. The problem was to extract a small discharge sound from the noise. In the present embodiment, the conventional problem is solved by configuring the partial discharge detection device shown in FIG. Hereinafter, the case of partial discharge detection will be described.

  In FIG. 1, for partial discharge detection, partial discharge sound generated from a detection target (discharge source) is collected by a parabola 11 and converted into a sound signal by a microphone 12, and this sound signal is amplified by an amplifier circuit 13. A signal whose waveform is shaped by a signal processing circuit (high-pass filter, envelope detector, low-pass filter, etc.) 14 is obtained, converted into a digital signal by an A / D converter (analog / digital converter) 15, and a DSP based on the digital signal (Digital Signal Processor) 16 performs the FFT calculation and identifies the discharge sound, which is equivalent to the conventional method.

  Here, in the present embodiment, during the partial discharge detection, a silencer that actively silences noise generated from inside or around the detection target is provided. This silencer generates a control sound having the same amplitude and opposite phase as the periodic noise noise identified by the DSP 16 and causes the noise to interfere. As a result, noise noise is actively silenced, and the influence of noise noise on the collection and identification of the discharge sound accompanying the occurrence of partial discharge at the detection target is reduced.

  The silencer obtains a control sound digital signal having the same amplitude and opposite phase as the periodic noise noise in the DSP 16, converts this signal into an analog signal by a D / A converter (digital / analog converter) 17, A signal in the same frequency band as the noise noise is extracted from the analog signal by the band-pass filter 18, this signal is amplified by the amplifier 19, and is generated as sound by the speaker 20. The speaker 20 emits a sound toward the sound collecting direction of the parabola 11 and cancels the noise noise because the sound is in a phase opposite to that of the noise noise, thereby reducing the noise noise collected by the parabola 11.

  FIG. 2 shows a mode of partial discharge detection or abnormal rotation sound detection by the partial discharge detection device. (A) of the figure shows an application example to a switchboard storage device. When partial discharge occurs along the bushing surface due to surface contamination of the bushing 21, the parabola 11 is attached to the bushing 21. A partial discharge is detected by pointing.

  FIG. 2 (b) shows an example of application to a switchboard storage device. When a partial discharge is generated in the internal winding of an epoxy resin molded instrument transformer (VT) 22, the instrument transformer is shown. Partial discharge is detected by directing the parabola 11 to 22.

  FIG. 2 (c) shows an example of application to a switchboard storage device. When partial discharge is generated on the surface of the insulation-coated busbar connection part 23, the partial discharge is caused by directing the parabola 11 to the connection part 23. Is detected.

  FIG. 2D shows an example of application to a rotating machine. When a partial discharge is generated in the coil insulation portion of the rotating machine 24, the partial discharge is detected by directing the parabola 11 toward the rotating machine 24.

  FIG. 2E shows an example of application to the detection of the operating sound of the rotating machine. When abnormal sound is generated on the rotor shaft of the rotating machine 24, the abnormal rotation is caused by directing the parabola 11 to the rotating machine 24. The presence or absence and abnormal sound characteristics are identified. It is not limited to detecting abnormal noise of a rotating machine, and obtains the same effect by applying it to abnormal noise detection associated with machine operation in other devices having moving parts (devices having a mechanism such as a reciprocating device). be able to.

  FIG. 3 shows a detection waveform example of the occurrence of partial discharge due to an internal defect of an instrument transformer (VT). 1A shows a case where the circuit operation to the speaker 20 of FIG. 1 is stopped and noise noise is not silenced. A signal due to partial discharge is buried in periodic noise noise, and identification by FFT is performed. I find it difficult. FIG. 3B shows a case where noise noise is silenced, and it is understood that the periodic noise signal is canceled and only the signal due to partial discharge is extracted.

The block diagram of the abnormal sound detection apparatus which shows embodiment of this invention. The abnormal sound detection aspect figure by the abnormal sound detection apparatus of this invention. The example of a detection waveform of the partial discharge generation | occurrence | production by the abnormal sound detection apparatus of this invention. The equivalent circuit diagram of the calibration and measurement of the conventional partial discharge detection. The circuit block diagram of the conventional acoustic discharge detection apparatus.

Explanation of symbols

11 Parabola 12 Microphone 13 Amplifier 14 Filter 15 A / D Converter 16 DSP
17 D / A converter 18 Filter 19 Amplifier 20 Speaker

Claims (2)

An abnormal sound detection device that acoustically detects abnormal sound due to partial discharge of electrical equipment or abnormal sound due to operation of a mechanical device in an environment where noise noise exists,
Abnormal sound generated from an abnormal sound detection object is collected by a parabola and converted into a sound signal by a microphone. The sound signal is amplified and signal processed to be converted into a digital signal. Based on the digital signal, FFT is performed by a DSP. Abnormal sound detection means for performing calculation and identifying abnormal sound;
A control sound having the same amplitude and opposite phase as that of the periodic noise noise identified by the DSP is emitted toward the abnormal sound detection target object to interfere with noise noise existing around or inside the abnormal sound detection target object. An abnormal sound detecting apparatus comprising a silencing means that actively silences noise and noise. An abnormal sound detection method for acoustically detecting abnormal noise due to partial discharge of electrical equipment or abnormal noise due to operation of a mechanical device in an environment where noise noise exists,
Abnormal sound generated from an abnormal sound detection object is collected by a parabola and converted into a sound signal by a microphone. The sound signal is amplified and signal processed to be converted into a digital signal. Based on the digital signal, FFT is performed by a DSP. In the abnormal sound detection process to calculate and identify abnormal sound,
A control sound having the same amplitude and opposite phase as that of the periodic noise noise identified by the DSP is emitted toward the abnormal sound detection target object to interfere with noise noise existing around or inside the abnormal sound detection target object. An abnormal sound detection method characterized by having a silencing process of actively silencing noise noise by means of the above.

Images