JPH04347559A - Active silencer - Google Patents

Abstract

PURPOSE:To make unnecessary a rotation detecting means, improve generation of a synchronous signal, simplify a structure and improve adaptability in an active silencing control for cyclic noise of a rotary machine. CONSTITUTION:In a control system for obtaining a silencing signal Y through calculation of frequency domain with a signal processing circuit 3, a timing for outputting a time domain data obtained by inverse Fourier transformation of the silencing signal Y to a speaker 4 is controlled by a synchronous signal S. The synchronous signal S is obtained by processing a noise signal N calculated from a silencing error signal E and a silencing signal Y with a band-pass filter 10. The pass-band frequency of the band-pass filter 10 is controlled so as to cause the desired frequency element of a noise signal N to pass with an input frequency F of a compressor 7 as a noise source.

Description

[Detailed description of the invention]

0001

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an improvement in an active muffling device for muffling the noise of rotating machines such as blowers, compressors, engines, etc.

0002

2. Description of the Related Art Rotating machines such as blowers and compressors often generate periodic noise depending on their rotational speed. In order to eliminate this noise, silencing sound waves with the opposite phase and the same amplitude as the periodic noise of these rotating machines are emitted from a separately provided sound wave generating means to cause sonic interference, thereby eliminating the rotating machine noise. Counteracting active sound deadening systems have been used for some time.

FIG. 3 shows the principle. Periodic noise is emitted from the rotating machine 1, and is detected by a microphone 2, which is a sound wave detection means. The noise signal detected by the microphone 2 is subjected to predetermined processing by the signal processing circuit 3, and a sound wave with the opposite phase and the same amplitude is output from the speaker 4, which is a sound wave generating means for sound muffling, and the sound wave is muffled at the position of point A. be done. Here, in the signal processing circuit 3, a transmission characteristic H1 is obtained when the sound wave emitted from the rotating machine 1 propagates to the position of the microphone 2, and a transmission characteristic H1 is obtained when the sound wave emitted from the rotating machine 1 propagates to the position of the speaker 4. Transfer characteristic H2 and speaker 4
An anti-phase waveform is created by calculating on the time axis by convolving impulse response data with a predetermined transfer characteristic and the output of microphone 2, taking into account the transfer characteristic H3 of acoustic feedback from which the sound wave output from is generated. Generally, the sound that reaches the microphone 2 is processed and a signal for silencing is output from the speaker 4. Therefore, in order to satisfy the law of causality, the distance from the rotating machine 1 to the speaker 4 must be It needs to be longer than the distance to microphone 2. However, if the noise of the rotating machine 1 is periodic, this is not the case because the noise after a certain period is predicted.

FIG. 4 shows an example where the noise of the rotating machine 1 is periodic. In the configuration example of this active noise reduction system, the distance from the rotating machine 1 to the speaker 4 is shorter than the distance from the rotating machine 1 to the microphone 2. The fundamental frequency of the rotating machine noise depends on the rotation speed of the rotating machine 1. In the signal processing circuit 3, the output of the microphone 2 is Fourier-transformed, multiplied by a predetermined transfer characteristic, and then inversely Fourier-transformed into a time waveform, which is stored in the memory in the signal processing circuit 3 and applied to the rotation of the rotating machine 1. Output from speaker 4 in synchronization. To create this synchronization signal S, a rotation detector 5 provided close to the rotating machine 1 is used.

Active silencing devices for periodic noise are
To summarize, data of anti-phase waveforms for silencing, which are sequentially calculated by a signal processing circuit, are stored in memory for the periodic noise of rotating machines, which consists of the fundamental frequency component and its harmonic components. The sound is sequentially output from the memory to the silencing sound source at a constant cycle according to the fundamental frequency.

[0006]

Since the noise of rotating machines is generally periodic, a method as shown in FIG. 4 is often used. In such an active noise reduction system, a rotation detector such as a rotary encoder that detects the rotation of a rotating machine or a non-contact reflective detector is used to obtain the synchronization signal. However, these have problems in terms of how to attach them to ready-made rotating machines, installation space, and price. Furthermore, when a synchronization signal is created based on the frequency of the rotating machine input, fluctuations in slippage due to the rotating machine load are not taken into consideration, so it is not necessarily synchronized with the fundamental frequency of the noise. Therefore, the silencing accuracy deteriorates. Furthermore, in the case of a DC machine, a synchronization signal cannot be created. SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems in creating a synchronization signal, improve the silencing accuracy of an active silencing device, and widen its range of use.

[0007]

[Means for Solving the Problems] The present invention includes means for detecting a silencing error signal, a signal processing circuit that calculates a silencing signal using the error signal obtained thereby, stores it in a memory, and outputs it using a synchronization signal. A circuit that calculates a noise signal from the difference between the silencing signal component and the error signal, a bandpass filter that depends on the characteristics of the input of the noise source, and a synchronization circuit that processes the noise signal with the bandpass filter and controls the output of the signal processing circuit. A circuit for obtaining a signal was provided.

[0008]

[Operation] A noise signal is obtained from the rotating machine noise waveform calculated from the output of the silencing error signal detection means and the component of the silencing signal, that is, the antiphase waveform data, or the directly detected rotating machine vibration waveform, and the noise signal is Extracts only one component of the fundamental frequency or its harmonics using a bandpass filter that has a band that changes based on signals that correlate with the frequency of rotating machine noise such as the input frequency, input voltage, and input current, and utilizes this. It is possible to obtain a synchronization signal that is synchronized with the rotating machine noise. Therefore, it is possible to create a clock, that is, a synchronization signal, with a constant period according to the fundamental frequency of the rotation noise without performing rotation detection.

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 is a block diagram of an embodiment of an active silencer according to the present invention.

A compressor 7 is installed in the compressor room 6, and a noise-muffling speaker 4 is installed at a distance from the compressor 7 that is sufficiently shorter than the half wavelength of the sound wave having the maximum frequency to be muffled. A silencing error detection microphone 8 is installed in front. The error signal E detected by the silencing error detection microphone 8 is input to the signal processing circuit 3, and a silencing signal Y is calculated. The silencing signal Y is stored in a memory provided in the signal processing circuit 3, and is sequentially output in response to the synchronizing signal S. The muffling signal Y output from the signal processing circuit 3 is input to the speaker 4 in synchronization with the synchronizing signal S, and a muffling sound wave having an opposite phase waveform is emitted. Furthermore, the silencing signal Y is sent to the subtracter 12 via a filter 9 with the same characteristics as the transfer characteristic H4 from the input of the speaker 4 to the output of the silencing error detection microphone 8, and is subtracted from the error signal E here. , becomes the noise signal N. The noise signal N is input to a bandpass filter 10, and a synchronization signal S is output.

Since the fundamental frequency component of the compressor noise is approximately proportional to the input frequency of the compressor 7, the pass center frequency of the bandpass filter 10 is changed by the frequency command F from the compressor input.

In the bandpass filter 10, only one component of the fundamental frequency of the noise or its harmonics is extracted. At this time, the passband width of the bandpass filter 10 is set to have enough margin to avoid detecting harmonic components adjacent to the frequency component to be extracted. Furthermore, by changing the passing center frequency according to the input frequency, voltage, current, etc. of the rotating machine, it is possible to create a synchronization signal that can follow variations in the rotating machine load and rotation speed.

In FIG. 1, the input to the filter 9 is the silencing signal Y, but only the fundamental frequency component of the silencing signal Y may be input. Assuming that the filter 9 is realized by software, the calculation can be shortened if only the fundamental frequency component of the silencing signal Y is input to the filter 9.

The signal processing circuit 3 may calculate the silencing signal Y according to the following equation.

[0015]

[Math 1]

Here, Hsm represents a transfer characteristic from the input of the speaker 4 to the output of the silencing error detection microphone 8. The initial value Y0 of Y is set to 0, and Y is updated sequentially. Hsm can be calculated using the following formula.

[0017]

[Math 2]

Here, if Y0 is set to 0 and the calculation is repeated, Hsm approaches the optimal value. The transfer characteristic H4 of the filter 9 may be replaced with the transfer characteristic Hsm obtained here and updated sequentially.

FIG. 2 shows another embodiment of the present invention, in which the noise source is an axial DC fan 11.

In an axial fan, noise at a fundamental frequency determined by the product of the fan rotation speed and the number of blades and its harmonic components is generated due to turbulence in the suction flow.

The system configuration is almost the same as the embodiment shown in FIG. 1, but since a DC fan is used, the passing center frequency of the bandpass filter 10 is determined by the input voltage command D of the DC motor that drives the fan. It is controlled by. Since the input voltage of the DC motor and the rotational speed of the fan are related, the relationship between the input voltage and the rotational speed of the fan is determined in advance by measurement, and the data is used to determine the relationship between the input voltage and the rotational speed of the fan.
Determine the passing center frequency of.

[0022]

According to the present invention, in active silencing of periodic noise of a rotating machine, a silencing signal calculated in the frequency domain is converted into the time domain, and a synchronization signal that determines the output timing is controlled by the silencing error. Since it is created by extracting the fundamental frequency component of the noise signal calculated from the signal and the silencing signal, a rotation detector is not required and the structural burden can be reduced. Furthermore, the active noise reduction device according to the present invention is less likely to cause howling and enables stable control.

[Brief explanation of drawings]

FIG. 1 is a block diagram of one embodiment of the present invention.

FIG. 2 is a block diagram of another embodiment of the invention.

FIG. 3 is a block diagram for explaining the principle of active silencing.

FIG. 4 is a block diagram of an active silencer for periodic noise.

[Explanation of symbols]

1 Rotating machine 2. Microphone 3 Signal processing circuit 4 Speaker 5 Rotation detector 7 Compressor 8. Silence error detection microphone 9 Filter 10 Bandpass filter 11 Axial DC fan

Claims (1)

[Claims] 1. A silencing error signal detection means, a signal processing circuit that calculates a silencing signal from the error signal obtained by the silencing signal, stores it in a memory and outputs it using a synchronization signal, and detects noise from the difference between the components of the silencing signal and the error signal. It is characterized by having a circuit for obtaining a signal, a bandpass filter that depends on the characteristics of the input of the noise source, and a circuit for processing the noise signal with the bandpass filter to obtain a synchronization signal for controlling the output of the signal processing circuit. active silencer.