EP1929465A1

EP1929465A1 - Method for the active reduction of noise, and device for carrying out said method

Info

Publication number: EP1929465A1
Application number: EP06778439A
Authority: EP
Inventors: Harry Bachmann
Original assignee: Anocsys AG
Current assignee: Anocsys AG
Priority date: 2005-09-27
Filing date: 2006-09-15
Publication date: 2008-06-11
Also published as: WO2007036443A1; AU2006296615A1; JP2009510503A; US20090220101A1

Abstract

Disclosed is a method in which a digital system is used for actively reducing noise. According to said method, a digital user signal can be transmitted at a different sampling rate, said sampling rate being adjusted with the aid of adequate converters.

Description

Method for active Geräuschveπninderung and an apparatus for performing the method

The present invention relates to a method for active noise reduction according to the preamble of claim 1 and an apparatus for carrying out the method.

Noise sources are increasingly perceived as an environmental impact and are considered to reduce the quality of life. However, as noise sources are often unavoidable, noise reduction methods based on the principle of wave cancellation have already been proposed.

The principle of Active Noise Canceling (ANC) is based on the cancellation of sound waves due to interference. These interferences are from one or more electro-acoustic

Transducers, such as speakers produced. The signal radiated by the electro-acoustic transducers is calculated by means of a suitable algorithm and continuously corrected. The basis for the calculation of the signal to be radiated by the electro-acoustic transducers is the information supplied by one or more sensors. These are on the one hand information about the nature of the signal to be minimized. For this purpose, for example, a microphone can be used which detects the noise to be minimized. To the others, however, also need information about the remaining residual signal. Again, microphones can be used.

One known application in which noise is actively reduced is, for example, headphones used by helicopter pilots. Thus, noise that comes in headphones from helicopter pilots, actively attenuated by knowledge of the noise generated by the drive of the rotors are used. The signal processing in these known headphones is realized by means of analogue technology, i. the acoustic signals and their processing are purely analog.

It is possible with the known headphones

Useful signal, such as radio traffic to mix, as both signals are analog.

All analog applications have in common that they are optimized for a specific situation. A transfer to other applications is usually not easy to accomplish. It is therefore always a new design of the hardware required when solutions for new applications must be provided.

The present invention is therefore based on the object of specifying a method for noise reduction, which does not have the above disadvantages. This object is achieved by the features stated in the characterizing part of claim 1. Advantageous embodiments and a device for carrying out the method are specified in further claims.

The inventive method is used for active noise reduction in an input signal, which is supplied to an unknown transfer function. The method consists in that the unknown transfer function or its actual

Output is estimated using an adaptive process using the input signal and an error signal. The error signal corresponds to the difference between the estimated output signal and the actual output signal. Furthermore, the estimated

Subtracts output signal from the actual output signal to form a noise reduced output signal. The invention is characterized in that a useful signal is superimposed on the noise-reduced output signal, wherein the useful signal does not influence the error signal, and that a calculation cycle of the adaptive process is longer than one clock interval of the useful signal.

This creates a flexible process that can be quickly transferred to a new application.

In particular, the hardware used for this purpose does not have to be changed, but an adaptation of the required algorithms that control the adaptive process is sufficient. In addition, the fact is taken into account that an active noise reduction high processing power requires. Meanwhile, the demands on the computing power are so high that the input signals to be processed must have a relatively low sampling rate. Input signals with too high sampling rates are therefore not processable. High sampling rates are also not appropriate because an active noise reduction works reliably, especially at relatively low frequencies.

In an embodiment of the inventive

Method is provided that a clock interval of the input signal is adapted to the calculation cycle of the adaptive process and that the clock interval of the estimated output signal is adapted to the clock interval of the useful signal.

This further takes into account the fact that, due to the available computing power, a reduced processing cycle is required in the adaptive process. In the context of the terms "clock interval" and "calculation cycle", it is pointed out that their reciprocal values correspond to the respective sampling rates.

A further embodiment variant of the method according to the invention consists in that the adjustment of the clock interval of the input signal takes place with the aid of a decimation algorithm, and, according to a still further embodiment variant, that the adaptation of the Clock interval of the estimated output signal using an interpolation algorithm.

Both decimation algorithms and interpolation algorithms for changing the sampling rates of digital signals are described, for example, by Ronald E. Crochiere and Lawrence R. Rabiner in the publication titled "Multirate Digital Signal Processing" (Prentice Hall, Inc., Englewood Cliffs, New Jersey, 1983 ).

Furthermore, a further embodiment is characterized in that a time difference existing between the useful signal and the noise-reduced output signal is corrected by means of an adaptive delay unit.

In addition to the method according to the invention, a device is also provided with an input signal which is acted upon by an unknown transfer function which has an actual output signal. The device further comprises:

an adaptive processor unit, which is supplied with the input signal and has an estimated output signal,

- Means for generating an error signal from the estimated output signal and the actual output signal and

Means for generating a noise-reduced output signal, wherein the error signal of the adaptive processor unit is applied. Furthermore are available:

- A useful signal source for generating a useful signal, and - Means for superimposing the useful signal to the noise-reduced output signal, wherein a calculation cycle of the adaptive processor unit is longer than a clock signal of the desired signal.

Another embodiment of the device according to the invention comprises:

- Means for adjusting a sampling interval of the input signal to the calculation cycle of the adaptive processor unit and - means for adjusting a sampling interval of the estimated output signal to a sampling interval of the useful signal.

A further embodiment is that the means for generating a noise-reduced output signal is at least one loudspeaker unit, which is acted upon by the actual output signal and the estimated output signal.

Furthermore, it is provided in a further embodiment that the at least one loudspeaker unit additionally receives the useful signal.

Yet another embodiment of the inventive device is that the Input signal via an analog / digital converter unit is applied to the means for adjusting the sampling interval of the input signal to the calculation cycle of the adaptive processing unit and that the estimated output signal via a digital / analog converter unit is applied to the means for generating a noise-reduced output signal.

Finally, another embodiment is that a first filter unit is arranged before the means for adjusting the sampling interval of the input signal to the calculation cycle of the adaptive process unit.

The present invention is based on

Embodiments with reference to drawings further explained below. Show it:

1 is a block diagram with an inventive device in a schematic representation,

Fig. 2 is a block diagram with another

Embodiment, also in a schematic representation, and

Fig. 3 shows a comparison with the block diagrams according to FIGS. 1 and 2 changed part.

1 shows a block diagram with a device according to the invention - including transfer function H - for active noise canceller (ANC), wherein the possibility is given to mix a generated in an external useful signal source 7 useful signal S.

The transfer function H is initially unknown

Great, which is estimated in an adaptive process unit 15. Alternatively, an actual output O of the transfer function H in the adaptive process unit 15 is estimated. The transfer function H is used to explain the device according to the invention or the method according to the invention and is itself not a part of the method according to the invention or of the device according to the invention. The transfer function H describes an actual output signal O, which is due to a voltage applied to the transfer function H.

Input signal I is created. Based on the application in a helicopter mentioned in the introduction to the introduction, the input signal I corresponds to the helicopter noise, as can be found, for example, in the cockpit, and the actual output signal O to the acoustic signal, as is still present in the headphones. The transfer function H therefore describes the change of the input signal I through the headphone shell. Active noise reduction is now achieved by estimating the transfer function H or its output signal. For this purpose, as can be seen from FIG. 1, the input signal I is fed via an analog / digital converter unit 1, via a first filter unit 12 and via a first decimation unit 4 to an adaptive processor unit 15. In the adaptive Processing unit 15 is determined using an adaptive algorithm, an estimated output O *, which is supplied to an interpolation unit 5. In the interpolation unit 5, a clock rate is set, which corresponds to the useful signal S. Thus, the prerequisite has been created that the estimated output signal 0 * and the useful signal S can be added, which happens in the addition unit 8 by this is acted upon by both the estimated output signal 0 * and the useful signal S. The output signal of the addition unit 8 is supplied via a digital / analog converter unit 2 to a superposition unit 14, to which the actual output signal 0 is likewise supplied, whereby the estimated output signal 0 * is inverted before the superposition ie before the superimposition with the actual output signal 0 *. so that upon coincidence of the two signals, a signal cancellation, ie the estimated output signal 0 * clears the actual output signal 0 completely. If there is no coincidence of the two signals, then no signal cancellation takes place but a reduction corresponding to the degree of agreement.

It should be noted that the superposition unit 14 is to be regarded as part of a model which describes the situation - again with reference to the helicopter example - in the space described by the auricle. In fact, the estimated output signal 0 * is applied to one, possibly to a plurality of loudspeaker units (not shown in Fig. 1) for generating an acoustic signal. The signal cancellation (at full Coincidence of actual and estimated output signal) or the signal reduction (with still different signals) takes place in the closed space.

In order that the adaptive processor unit 15 or the calculations carried out in it can be adapted continuously to the possibly changing transfer function H, an error signal ε is fed back to the adaptive processor unit 15. Therein, the estimated output signal O * or the transfer function estimated in the adaptive processor unit 15 is optimized until the error signal ε has reached a minimum.

According to the invention, a useful signal S is superimposed on a noise-reduced output signal Q. This must be taken into account in the calculation of the error signal ε, this being done by a subtraction unit 9, on the one hand the useful signal S of the useful signal source 7 on the other hand, for example, with a microphone (not shown in Fig. 1) and recorded with a second analog / digital Converter unit 3 converted noise-reduced output signal Q is supplied. As a result of the superposition of the useful signal S and the estimated output signal 0 * to form the actual output signal O, the useful signal S must be subtracted during the generation of the error signal ε, which takes place in the subtraction unit 9 as described above. Since the digital / analog converter unit 2 and the analog / digital converter unit 3 at the same clock rate to be operated, the output signal of the subtraction unit 9 corresponding to the error signal ε must be adapted to its calculation cycle before being passed to the adaptive processor unit 15. For this purpose, a second decimation unit 6 is provided, which carries out the required adaptation in the clock rates or in the clock intervals.

So that so-called antialiasing effects can be prevented, a first and / or a second filter unit 12, 13 is provided in the embodiment according to FIG. 1 before the first decimation unit 4 and / or before the second decimation unit 6.

In Fig. 1, the adaptive processor unit 15 is shown in dashed lines. Within the dashed frame two components of the adaptive processor unit 15 are shown, wherein an adjustable transfer function W and an associated with this error calculation unit LMS are present. The adjustable transfer function W ideally coincides with the transfer function H. Only then does the estimated output 0 * correspond to the actual output O and complete signal cancellation is the result. In case of inequality, a correspondingly reduced signal cancellation or only a signal reduction takes place. The error calculation unit LMS acts on the adjustable transfer function W in such a way that the largest possible signal reduction or even a complete signal cancellation is obtained. For this For example, a so-called LMS (Least Mean Square) algorithm is suitable, although this is one of many possible implementation variants. Basically, the algorithms known from adaptive signal processing for determining the estimated output signal, as described, for example, by Ronald E. Crochiere and Lawrence R. Rabiner in the publication titled "Multirate Digital Signal Processing" (Prentice Hall, Inc., Englewood Cliffs, US Pat. New Jersey, 1983) can be used in the adaptive processor unit 15.

It has already been pointed out that the two analog / digital converter units 1 and 3 convert an analog signal recorded for example with a microphone (not shown in FIG. 1) into corresponding digital signals. Furthermore, a calculated digital signal, namely the estimated output signal 0 *, is converted by the digital / analogue converter unit 2 into an analogue signal which, for example, is applied to a loudspeaker (not shown in FIG. 1). Since the converter units 1, 2 and 3 are part of the same CODEC, they are operated at the same sampling rate.

The CODEC has to work with a high sampling rate as soon as the useful signal S is to satisfy corresponding qualitative requirements, as is the case, for example, with music. For CD quality, the sampling rate is 44.1 kHz. Consequently, the converter units 1 to 3 are also operated with this clock frequency of 44.1 kHz. The Indian By contrast, the algorithm used by adaptive processor unit 15 operates at much lower frequencies, for example at 8 kHz. This conversion is, as mentioned, performed by the decimation units 4 and 6. The interpolation unit 5 converts the adaptive one

Processor unit 15 estimated output signal, which has a sampling rate of 8 kHz, in the 44.1 kHz necessary for the reproduction of music.

Thus, the addition of the unit 8 and the

Subtracting unit 9 applied to an identical sampling rate. As a result, the signals can be easily added or subtracted.

Fig. 1 shows an embodiment of the present invention

Invention in which antialiasing effects are avoided. For this purpose filter units 12 and 13 are provided in front of the respective decimation units 4 and 6, respectively. The two filter units 12 and 13 now ensure that the following decimation units 4 and 6 take into account only relevant signal components by filtering out all signal components above half of the reduced sampling rate, in this case all signal components above 4 kHz.

Fig. 2 shows an embodiment of the present invention, in which no filter units 12 and 13 are provided. Accordingly, a deterioration of the signal processing, in particular in the adaptive processor unit 15, to be expected here, because at this variant must be reckoned with antialiasing effects.

FIG. 3 shows a modified part of the block diagram shown in FIGS. 1 and 2. Thus, an adaptive delay unit 20 is included in the signal path between the addition unit 8 and the subtraction unit 9 before its input in order to be able to compensate for a delay of the useful signal S. The delay of the useful signal S arises in the signal path via the addition unit 8, the digital / analogue

Converter unit 2 and the analog / digital converter unit 3. Accordingly, the useful signal S, which is fed directly to the subtraction unit 9, delayed, so that an accurate calculation of the error signal ε is possible.

A flexible adaptation of the hardware of the present invention requires a digital implementation of the active noise reduction unit. Since loudspeakers are already present in such active noise reduction units, integration of other acoustic signals, such as speech or music, is desired.

As has already been pointed out, the signals detected for example with microphones are analog and must be converted into a digital format for further processing with the adaptive processor unit. CODECs are an efficient option for this purpose. They are cheap and optimized for audio applications and also have several channels. A CODEC will open all channels with identical sampling rate. Examples of suitable CODECs are the algorithms with the designations TLV 320 AIC 23 or TLV 320 AIC 25, which have been developed by Texas Instruments Inc. However, the present invention is not limited to the use of these algorithms.

The use of conventional converter units instead of CODECs for each channel is basically conceivable, in which case an individual sampling rate can be set for each channel.

The adaptation of the clock rates or of the clock intervals can be carried out in a digital signal processing unit (DSP - Digital Signal Processor), which is present in any case in a variant of the inventive apparatus for calculating the adaptive processes. This eliminates additional costs that otherwise have to be expended for the decimation units or the interpolation units.

Claims

claims

A method for active noise reduction in an input signal (I) which is fed to an unknown transfer function (H), the method being that the unknown transfer function (H) or its actual output signal (0) by means of an adaptive process Using the input signal (I) and an error signal (ε) is estimated, the error signal

(ε) corresponds to the difference between the estimated output (O *) and the actual output (0) and that the estimated output (0 *) is subtracted from the actual output (0) to form a noise reduced output (Q), characterized in that a useful signal (S) is superimposed on the noise-reduced output signal (Q), wherein the useful signal (S) does not influence the error signal (ε) and that a calculation cycle of the adaptive process is longer than one clock interval of the useful signal (S).

2. The method according to claim 1, characterized in that a clock interval of the input signal (I) is adapted to the calculation cycle of the adaptive process and that the clock interval of the estimated output signal (0 *) to the clock interval of the useful signal (S) is adjusted.

3. The method according to claim 2, characterized in that the adaptation of the clock interval of the input signal (I) by means of a decimation algorithm.

4. The method according to claim 2 or 3, characterized in that the adaptation of the clock interval of the estimated output signal (O *) takes place with the aid of an interpolation algorithm.

5. The method according to any one of claims 1 to 4, characterized in that between the useful signal (S) and the noise-reduced output signal (Q) existing delay difference is corrected by means of a delay unit (20).

6. An apparatus for carrying out the method according to one of claims 1 to 5 with an input signal (I), which is applied to an unknown transfer function (H) having an actual output signal (O), further:

an adaptive processor unit (15), which is supplied with the input signal (I) and has an estimated output signal (0 *),

- Means for generating an error signal (ε) from the estimated output signal (0 *) and the actual

Output signal (0) and

- means (14) for generating a noise-reduced output signal (Q), wherein the error signal (ε) of the adaptive processor unit (15) is acted upon, marked by

- A useful signal source (7) for generating a useful signal (S), and

- Means for superimposing the useful signal (S) the noise-reduced output signal (Q), wherein a calculation cycle of the adaptive process unit (15) is longer than a clock signal of the useful signal (S).

7. Apparatus according to claim 6, further characterized by - means (4) for adjusting a sampling interval of

Input signal (I) to the calculation cycle of the adaptive process unit (15) and

- Means (5) for adjusting a sampling interval of the estimated output signal (O *) to a sampling interval of the useful signal (S).

8. Apparatus according to claim 6 or 7, characterized in that the means (14) for generating a noise-reduced output signal (Q) is at least one speaker unit, the actual

Output signal (O) and the estimated output signal (0 *) are applied.

9. Apparatus according to claim 8, characterized in that the at least one loudspeaker unit in addition to the useful signal (S) is acted upon.

10. Device according to one of claims 6 to 9, characterized in that the input signal (I) via an analog / digital converter unit (1) the means (4) for Adjustment of the sampling interval of the input signal (I) to the calculation cycle of the adaptive processing unit (15) is applied and that the estimated output signal (O *) via a digital / analog converter unit (2) the means (14) for generating a noise-reduced output signal (Q ) is acted upon.

11. The device according to claim 7, characterized in that a first filter unit (12) before the means (4) for adjusting the sampling interval of the input signal (I) to the calculation cycle of the adaptive process unit (15) is arranged.