JPS59133595A - Active sound attenuator - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION This invention relates generally to acoustic attenuation, and more particularly, to an apparatus for attenuating the entire low frequency acoustic band propagating from a particular source.

Patent application No. 1186, filed on September 4, 1981
No. 77 describes a number of exemplary active sound attenuation devices, particularly with respect to physical systems involving the propagation of sound within ducts. The apparatus and method disclosed in that patent application are also applicable to other acoustic cancellation environments, such as partitioned rooms and private spaces. This invention is
This invention is an improvement and extension of the invention of the second patent application and provides an active sound attenuation device in a free space environment.

Significant reduction of the sound pressure level of sound propagating inside confined enclosures such as ducts has been an unresolved problem for many years. for example.

In factories, noise generated by machinery and various manufacturing operations is transferred from heating and ventilation ducts in such areas to ducts connecting to offices and other parts of the plant where low levels of noise are desired. transported. This is particularly a problem with low frequency noise in the inaudible to 800 hertz range. That is, passive means of attenuating such frequencies are expensive, relatively inefficient, and physically large in size, making them impractical in most low frequency applications.

It began in 1925 and has continued on a very rapid basis to the present day. Developments in electronics are noise
The concept of "active" attenuation can be compared to low-frequency passive (
We have not only made it a possible alternative to (passi, ve) attenuation, but also made it attractive. The so-called active attenuation principle is based on the fact that the speed of sound in air is much lower than the speed of electrical signals. Take a sample of the propagating wave during the time it takes for the sound wave to propagate from the location where it is detected to the remote location where it is attenuated.

This information is processed in an electronic circuit, and 180
There is sufficient time to generate a signal to drive the loudspeaker that introduces a canceling sound having a degree out-of-phase and equal amplitude to the propagating sound. Although the process of active acoustic attenuation described above appears conceptually simple, when one considers all the prior art in this field.

The complexity of this problem and the difficulty of obtaining good g reduction over a relatively wide band of low frequencies can be seen.

One of the first efforts in the field of active damping was disclosed in Loeg US Pat. No. 2.0113.1I16. TJueg's system is a monopole consisting of a microphone, amplifier and speaker. The microphone detects the sound and converts it into an electrical signal, which is sent to an amplifier. A loudspeaker driven by an amplifier.

It is placed downstream of the microphone in a location that provides the time delay necessary to achieve an 18° phase reversal from the sound. The loudspeaker injects a mirror image of the sound into the duct so that when the sound passes through the loudspeaker location,
High- or low-pressure air is introduced that is 180 out of phase with the corresponding high-pressure and low-pressure air of the sound. When the loudspeaker is perfectly synchronized with the passage of the source noise, the pressure of the source noise (17) and the pressure of the loudspeaker are on average zero (
ambient static pressure), the noise is completely cancelled.

As is clear from a review of the Lueg system, attenuation occurs because the distance between the microphone (where the sound sample is taken) and the loudspeaker (where the canceling sound is introduced) is This is the case when the time delay of the electrical signal transmitted is equal to 180 degrees or an odd multiple of 1800. However, this condition will only occur for certain stationary acoustic signals that do not vary in time. Therefore, as a practical matter, since the Lueg system is not provided with means to accommodate two phase changes.

Valid only for a single frequency. What this limitation of the Lueg system shows is that there are two parameters that must be met for good attenuation. That is, it includes a delay time to allow the sound wave to travel from the detection point to the attenuation point, and a phase to allow the introduction point of the sound wave to occur where the attenuation cancels out.

Lueg system associated with phase detection and adaptation (18
In addition to the limitations of ), there is the problem of standing waves generated by the loudspeaker in the upstream direction towards the microphone.

Because of the standing wave pattern, the pressure in the sound field at the microphone is artificially unequal.

This means that the microphone may be placed at the node or antinode of the standing wave at a particular frequency. Therefore, the signal generated by the speaker may be too loud or too small. As a result, the sound field is amplified by the standing waves and the propagation occurring downstream of the loudspeaker can even be louder than the sound generated at the source. Additionally, the standing wave field can enhance the sound pressure within the duct, and more sound can penetrate the walls of the duct and generate secondary waves.

Several active damping systems have been developed following the Lu8g system to avoid the standing wave problem mentioned above and to extend the frequency range of damping. One prior art system utilizes a combination of monopoles and dipoles, where the dipole is placed on one wall of the duct;
The monopole is placed on the opposite side of the duct. This system was developed by M. Jessel and A. Mangiant.
First introduced by e, Act, j, ve 5o
und Absorbers in an Air D
uct”.

, TSV (1,923) 23 (3,683-
390), described in a paper entitled. Jes
The dipoles and monopoles of the sel system are phased so that they add in the downstream direction and subtract in the upstream direction, creating unidirectional propagation when both sources are balanced.

However, the results obtained with the Jessel system were found to be frequency dependent and related to the half wavelength of the dipole separation. moreover.

The complexity of this system makes it unsuitable for use in many practical applications.

As a means of simplifying and providing improved performance over the Jessel system, another system has been developed. One system is U.S. Pat.
, No. 0144,205. SWl n
1) Ank S removed the monopole found in Jessel and changed the phase characteristics of the dipole so that the propagation of one and both sources added in the downstream direction and canceled in the upstream direction towards the microphone. Another active attenuator disclosed in US Pat. No. 1+, 109,108 to Coxon et al. places a microphone between two loudspeakers;
Proper phasing is introduced between the speakers, resulting in the lowest level at the microphone location. This system is reflective and generates a standing wave pattern upstream of the dipole, but the detection system (microphone) is
Unlike the nbanks system, it is not affected because it is placed between the speakers.

In examining the performance of dipole and dipole-monopole systems, it has been found that each of these multi-source systems appears to have geometrically related limitations. The physical spacing of the loudspeaker and microphone is subject to tuning effects that set the frequency and bandwidth for best performance.
ing effect)'. Especially '5w1n
Although high levels of attenuation are possible in the banlcs system, such performance can only be obtained over relatively small bandwidths, on the order of up to about 23 A octaves. Recent approaches to active noise attenuation in ducts have therefore focused on improving the monopole system first introduced by 51113g as mentioned above.

Chaplin et al., US Pat. No. 14.122.306, is an example of such a monopole system, in which Lu
Relatively sophisticated electronic circuitry is incorporated into the basic microphone and loudspeaker configuration found in 8g systems. In Chaplin et al., a secondary canceling sound wave is generated by a first electrical signal representing a primary sound wave detected by a microphone.

The first electrical signal is convolutioned with a second signal induced from the impulse response of the system as part of a program of operating steps.

This process represents the standard operation of an adaptive filter. Regarding this, 18 dap
t:L'Ve Fll terS", WldrO
W-B, + (1971); “Pr1nciples
and Applications of Adapt
ive Filters, A Tutorial
Review''.

(22) McCool, J.M.; Widrow,
B-+ Tnstitut, eof Elect
Rical Engineers, (1976). To avoid the problem of standing waves generated by the loudspeaker and propagated upstream toward the microphone, CbapLjn et al. I suggest lining it. Furthermore, Chaplin et al. propose the possibility of incorporating a circuit that provides a second convolution capable of compensating the feedback signal generated by the upstream standing wave. However, as detailed below, the so-called adaptation process occurs over a period of 5 to 30 minutes and is therefore dependent on the source signal remaining essentially constant over that time.

Moreover, since this control system works extremely KV'b, in many practical cases while the first convolution process is still running.

One system hunt between better and worse results by manually shutting down the system.
ng) ” We still have to avoid vibration.

It should be noted that each of the systems described above has a common problem that reduces, if not eliminates, their potential usefulness in many practical applications. In each case, the speaker portion of the system is placed within the duct environment to introduce canceling sound.

In many applications, existing loudspeakers operate at temperatures 1
It is expected that it will not be able to withstand particles and other foreign substances for long periods of time.

In ductwork environments, active acoustic attenuators, which are modified monopole active acoustic attenuators, are used, which include the physical system, the electronic circuit, and the coupling components between the physical system and the electronic circuit. It can be considered as consisting of separate components. The physical system consists of a duct through which sound waves propagate from a particular source, an acoustic mixer connected in line with the duct, a speaker located within a protective enclosure at a distance from the duct, and the speaker and acoustic mixer. and a waveguide connecting the two. The electronic circuitry of the active attenuator consists of a separate adaptive filter that utilizes a modified version of the Wldrow-Hoff LMS algorithm in a true adaptive acoustic cancellation configuration. A microphone array placed in the duct at a location upstream from the acoustic mixer and a microphone placed in the acoustic mixer downstream from the waveguide form a coupling means between the physical system and the electronic circuit. do.

As discussed in more detail below, the microphone array detects the noise within the duct and converts it into an electrical signal that is sent to a modified adaptive filter in the electronic circuitry. This adaptive filter generates a signal to drive a speaker, which produces a "cancelling" sound that is 185 degrees out of phase with the sound, and this canceling sound propagates through the waveguide to the acoustic mixer. , where a microphone placed downstream from the waveguide in the acoustic mixer, which is combined with the brisk sound, detects the sound resulting from the combination of the brisk sound and the canceling sound (25), and which is achieved in the acoustic mixer. Generates a signal representing a 1 error '!f difference between the #back and the desired #attenuation based on a preset level. This error signal is introduced into the adaptive filter, which is adjusted to drive the loudspeaker so that the canceling sound propagating into the sound mixer more closely resembles the mirror image of the sound. Thus, a high level of attenuation is achieved within the sound mixer and everywhere in the far area.

In one preferred form of the active acoustic attenuator, a compensating adaptive filter was provided to ensure stable operation of the modified LMS algorithm in generating the cancellation sound. This adaptive filter is used in place of a phase correction filter that provides phase correction and requires manual calibration. In a free space environment, it is advantageous to place the cancellation sound generator as close as possible to the source of the sound to be canceled. To help achieve this desired placement of canceling sound sources, the compensating adaptive filter should be properly positioned (26) in the active acoustic device. In accordance with one aspect of the invention, a compensating adaptive filter is relocated from the cancellation loop of the attenuator, thereby reducing the delay in the cancellation loop. This allows the canceling sound source to be placed closer to the source of the sound to be canceled.

According to another ifi of the invention, in the use of the above active sound attenuator in a free space environment, a plurality of canceling sound generators are arranged at a distance from the source of the sound to be canceled and in different propagation directions of the sound.

Other objects and advantages of the invention and the manner in which they may be carried out will become apparent from the following detailed description taken in conjunction with the drawings.

Referring now to the drawings and particularly to FIG. 1, there is shown a sound attenuator in a duct environment, the active sound attenuator being designated by the reference numeral 11. Active attenuator 11 may be considered as a separate component, and the description herein will be directed primarily to the physical system component, with reference to other components as appropriate. The physical system consists of a duct through which sound from a particular source propagates, an acoustic mixer 15 connected to the duct 13, a speaker 17 that is the source of the canceling sound, and a conductor connecting the speaker 17 to the acoustic mixer 15. Contains one wave tube.

Although acoustic mixer 15 is shown to have a slightly larger diameter than one of the ducts, this geometric relationship is not necessary for proper operation of active attenuator 15 and is shown for purposes of explanation and illustration. It's nothing more than that.

One of the immediately obvious advantages of the active attenuator 11 over prior art systems is that the loudspeaker 17 is located remotely from the duct 13 and is connected to the acoustic mixer 15 by a single elongated waveguide.

That is, the waveguide 19 may be equipped with a pulp mechanism (not shown) to prevent duct particles, corrosive materials or other debris flowing through the duct 13 from damaging the speaker 17. Prior art active attenuators place the speakers directly within the duct, and the internal and external conditions encountered in many applications can quickly damage or destroy them. The loudspeaker 17 is not only protected from the internal environment of the duct 1 and the sound mixer 15 by the waveguide 19, but is also housed in an enclosure 21 that protects the loudspeaker 17 from the external environment of the duct 13. The enclosure 21 must have high transmission losses and may also be lined with sound absorbing material, which further prevents the output of the speaker 17 from propagating in directions other than one of the waveguides. It should be understood that waveguide 19 is completely separate from duct 13 and acoustic mixer 15. That is, one waveguide is not a branch duct and therefore the flow from the main duct 15 is not carried through one waveguide. One waveguide simply carries the canceling noise from the loudspeaker 17 to the acoustic mixer 15 and connects the loudspeaker 17 to the duct 15.
It is established for the purpose of isolating and protecting from the environment.

The electronic components of active attenuator 11 are shown in their simplest form in FIG. 1 for purposes of illustration.

A detailed description of the electronics of the invention is provided below (in two) along with a description of the complete circuit used in the invention. A simplified form of the electronic components of the active attenuator 11 includes an adaptive filter 23, an amplifier 25, two phase correction filters and a D generally designated by the reference numeral 51.
Contains C loop. The coupling component of the present invention, which couples the physical system with the electronic system, includes a microphone array 33 placed in the duct at the beginning or upstream side of one of the waveguides for detecting the sound waves, and as will become clear later on. It consists of a microphone 35 disposed in the acoustic mixer 15 downstream from one waveguide for the purpose of

Generally, in the cancellation mode of operation, active attenuator 11 operates as follows. Broadband noise propagates along duct 13 and is detected by microphone array 33, which generates a signal that is sent to adaptive filter 23. Adaptive filter 23 provides an output to drive speaker 17, which introduces canceling noise into acoustic mixer 15 through a single waveguide. Sound waves undergo compression (maximum pressure) and rarefaction (3) of a specific phase and frequency.
0) (pressure minimum part), the pressure of such a wave is equal to the -order sound wave by generating a second-order sound wave having an amplitude equal to the second-order sound wave and a reaction with the combination 1/tension which is 180 degrees out of phase. , the dimensions can be reduced or offset. A microphone 5 disposed in the acoustic mixer 15 downstream from one of the waveguides detects the degree of attenuation or cancellation of the sound after the sound waves generated by the speaker 17 are combined with the pure tone. The signal from the microphone 35 is sent to the adaptive filter 2 as an error signal.

This error signal is effectively an indication of the attenuation achieved within the acoustic mixer 15. The adaptive filter operates to adjust its output depending on the nature of the error signal, so that the speaker 17 is driven to produce a canceling tone that is closer in amplitude and 180 out of phase to the pure tone.

The second major component of the active acoustic attenuator 11 of the present invention is the electronic implementation or control of the physical system described above.

A brief discussion of the prior art will be helpful in understanding the advances made to electronic control circuits by the present invention. Linear addition of two equal and opposite signals has long been recognized as one approach to generating electronic or acoustic zeros. Some of the acoustic cancelers developed to date are based on the equal and opposite principle, in which canceling sound waves generated by a loudspeaker are introduced into a confined space, such as a duct, and are transmitted from a specific source through the duct. Reduce pressure fluctuations caused by propagating sound waves. In the simplest model, the signal generated at the source is considered a pure tone represented by a single rotation vector. The cancellation signal must track the source within a certain maximum tolerance to achieve the desired attenuation. This problem has been recognized and solved by the present invention. The key issue with this approach is.

The problem is that the permissible amplitude and phase errors must be kept within tolerances found only in the best acoustic equipment. For example, to achieve an attenuation of 20 dB, the microphone of such a system.

Errors in the loudspeaker and electronics must be less than 1 dB in amplitude and i in phase. As detailed below, one invention overcomes this problem using a tolerance relaxation technique based on feedback principles. Instead of requiring accuracy from all components, the feedback system concentrates performance requirements into a small number of easily controlled devices.

One of the latest so-called active damping systems.

U.S. Pat. No. 11,122,50 to ChapLin et al.
No. '5, this system ostensibly uses an adaptive process to generate a canceling sound wave capable of producing a zero when combined with a sound wave from a source. However, as a review of this system shows, its electronics do not rely on a true adaptive process, but require a trial-and-error approach, and the error is the difference between the desired and actual attenuation. A series of successive guesses or decisions are made from the signal. Eventually, the error signal will become increasingly close to approximating the desired error by the preset value. The trial and error method is not only unduly time consuming (in the range of 5-30 minutes), but also leaves one system hunting between better and worse results after the process is completed. !!, the system must be stopped manually to avoid high vibrations.Furthermore, once the trial and error process begins in the system of Chaplin et al. Changes also result in one system hunting one.

One major aspect of the electronic circuit of the present invention is that the electronic circuit
It is a deterministic system that utilizes true adaptive filters as opposed to the trial and error approach taught by Haplin et al.

This means that a single electronic circuit automatically adjusts its own parameters and seeks to optimize its performance according to specific criteria. Furthermore, by using the feedback principles described above, one electronic circuit of the present invention is less dependent on the tolerance requirements of acoustic equipment used in many known prior art systems. In fact, if the permissible amplitude and phase errors of the acoustic devices used in prior art systems are of the order of 1 dB in amplitude and 6° in phase.

The electronic circuit of the present invention has a noise level of at least 10 decibels.
11) and) Amplitude and phase errors in the range of 1 Da can be tolerated. By relaxing tolerance requirements, installation and maintenance of this system can be performed by technicians of ordinary skill, making the active sound M,l+ff4i attenuator 11 commercially viable in a variety of applications. Make it.

The electronic circuitry of the active acoustic attenuator 11 is a true adaptive acoustic cancellation form of Wldrow-Hoff LIVI.
Utilizes a modified version of the S-adaptive algorithm.

The LMS algorithm was designed for use in signal enhancement systems where signal-to-noise ratio modification or noise reduction is accomplished solely in electronic circuitry.

This algorithm is notable in one inventive circuit to enable operation when vibrational vibration or noise bass and/or attenuation is to be achieved in physical systems with inherent delays, such as in the acoustic field. It was changed to . Wi,
A modified version of the draw-Hoff LMS algorithm retains the signal processing advantages inherent in the original algorithm and allows these advantages to be applied to active acoustic cancellation problems. Sen's adaptive algorithms exist for solving quadratic error functions, and many of these can be modified to work satisfactorily in acoustic cancellers. The purpose of the adaptive algorithm is as detailed below.

The goal is to find an optimal or near-optimal solution to the cancellation filter problem. Such other algorithms, modified in accordance with the present invention, may accomplish this function and should be considered within the scope of the present invention.

Before describing the modification of the LMS algorithm in the adaptive filter of the present invention, an adaptive filter with a standard LMS algorithm governing its operation will be described. Referring to FIG. 2, the mark i'f7. i
An adaptive filter with Widrow-Hoff LMS algorithm is shown. The basic elements of an adaptive filter (transversal)
The filter 5 is shown in dotted lines in FIG. The Traless Versal Filter 5 can be viewed as a series of delay elements, with the filter output derived from a weighted sum of the delay outputs. In Figure 2, a set of n measurements S(↑,) is taken as a sample and n
form sample measurements 5 (jl, where J is the time index of the sample. Each point labeled 57 in Figure 2 can be considered to constitute a sample input value; , Z-1 factor represents the delay.
The # value 37 is the corresponding weighting factor W(j
This weighted constant value is inputted into an adder 111 to form an output YJf, which is the output of the transversal filter 53. This output Yj is compared with the desired response d in an adder 4 and outputs an error signal e,
form.

T-M that controls the operation of the transversal filter 53
The purpose of the S algorithm is to deterministically obtain the weighting factors so as to minimize the error signal e3 and find the weighted sum of the input signals that best matches the desired response. To achieve this goal, the weight vector is changed using the saddle point method on the quadratic error surface.
st, eepest cessent) along the direction of the estimated gradient vector. A detailed treatment of this subject is Rud olf'E
, K,a 1man and N1choLas DeC
Aspect, s of Filter edited by Laris
s”, Prof. Bernard Wid.
Found in the paper by Rowe. A block diagram representation of a digital implementation of the LMS algorithm is shown in the right-hand portion of FIG. 2, contained within the dotted line and designated generally at 55. It should be noted that although shown in digital form, an analog implementation of the LMS algorithm and transversal filter could also be used.

For purposes of explanation, only two sample input values and their corresponding weighting functions are shown in the figure. input sJ
l is sent to multiplier u5. The error signal ej is fed into a 1 multiplier 7 with a scale factor μ which controls the convergence rate and stability of the algorithm. The error signal multiplied by the scale factor is then multiplied by the signal 5fjl in the multiplier 115, and the product is added to the adder +19 and the unit delay 51 or CT(A M (Coeffi
ci, ent Random (38) Ac Ce S S Me l no ry). The weight setting tJ(j+1) is the input signal S(
j+1) and the product then forms part of the output of the transversal filter 55. The same operation is performed in way l' W2 (
This is performed for the human input signal 5 (j-1,) corresponding to j).

” As a continuation, the LMS algorithm provides a method in which the weight settings determined from the error signal can be directly fed back to the adjustable weights corresponding to the input signals, which are determined essentially without delay. It has been demonstrated to be effective in many adaptive signal processing applications.

In the configuration shown in FIG. 2, the LMS algorithm and transversal filter are not suitable for use in acoustic cancellation problems. Cancellation of the acoustic wave propagating along the duct requires an equal and opposite signal to interact with it and reduce the highest pressure fluctuations generated in the acoustic mixer 15. The interaction of two waves requires a finite travel length and corresponding amount of time. Next, refer to Figure 4.

It can be observed that in the physical system of the articulating attenuator 11, a finite length of time is required for the sound waves to propagate from the microphone array 153 where the input signal is detected. Additionally, there is a delay in the time required for the canceling sound waves generated by speaker 17 to propagate through waveguide 19 to acoustic mixer 15. These delays are estimated as the distance in feet across the duct +-1s, acoustic mixer 5, and waveguide 19 divided by the speed of sound, 1100 feet per second. Expected delays for most systems range from a few milliseconds to a few tenths of a second. 20 per second
Using a sampling rate of 00 or more, the delay expressed in sampling intervals can be from a few to hundreds.

Expressed in terms of sampling interval, the delay is given as follows:

Delay - K (1/sampling rate) where K - integer constant Widrow-Hoff LMS algorithm is modified to take into account the inherent delay in the physical system of active attenuator 11, and the weighting factor determined by the colgorithm is , the weighting coefficients can be matched with the corresponding signal in the transversal filter for which the weighting coefficients have been determined.

For a new input signal sample S(,1), the corresponding weighting coefficient is expressed as follows.

% (j 't=% (j -1) 10μS (:j-
(K+1) le (j-1) (11 Here, the next value of μ=niscale factor weight can be written in the following format.

Wi Cj +1 )−Wt(j)+−μS (j −
K) efil (21 where 1-tap discrimination S (j-K) = each sample value of the input signal is passed to the input sample at an interval of the transversal filter 5
The filter operates to generate the product of the first weighting factor and the final input sample, this product is added to the product of the second weight and the sample closest to the final input sample, and so on. Finally, the (111) product of the final weight and the oldest sample is accumulated. The total integration of these terms is the output y(jl) of the transversal filter 53. The weighting coefficients are updated by the corresponding human samples and error signals to account for the inherent delays in the physical system. Digital implementation of the modified LMS algorithm according to the above (all cases K = 2), the above-mentioned delays are such that the corresponding weighting coefficients, input samples and error signals are combined in an adaptive filter to generate an output signal y(j)'i has been adapted to.

Next, with full reference to Figure 1.4, the modified Widrow-Hof
The electronic circuitry of active attenuator 11 including an adaptive filter that utilizes the fLMS algorithm will now be described. In its simplest form, as shown in FIG. 1, the basic operation of the electronic circuit of active attenuator 11 is explained by the following sequence.

L Take a sample of the sound wave propagating along the duct 1 (signal) 2° Delay, filter and scale the signal (42) U Drive the speaker 17 with the output of step 2 above to properly adjust the signal 4. Detect the acoustic output of the acoustic mixer 15 (error) 5. Adjust stage 2 using a modified LMS algorithm 6. Return to stage 1 The adaptive filter 23 shown in FIG. The port configuration consists of two inputs and one output. The adaptive filter 2 receives the input, which is an electrical signal corresponding to the sound in the duct 15, from the microphone array 33.
, and generates an output signal to drive the speaker 17;
Speaker 17 generates a mirror image replica of the sound, which is introduced into acoustic mixer 15 through waveguide 19.

The output of the acoustic mixer 15 detected by the microphone 5 is introduced into the adaptive filter 2 as an error signal or the sum of the noise obtained by the acoustic mixer and a replica.

The acoustic propagation delay inherent in the physical system of active attenuator 11 must appear as a delay over the entire frequency range of interest. The phase tolerance of this delay is approximately ±115 degrees at each frequency. The block diagram of the electronic circuit shown in FIG. 1 includes two secondary phase correction filters to compensate for acoustic resonance in the acoustic mixer 15. The characteristics of the two filters must be determined manually using appropriate instrumentation based on the duct characteristics of the particular application. As discussed below, this functionality can be accomplished with an adaptive filter, thus eliminating manual calibration of the two filters altogether.

The block diagram of the active attenuator 11 shown in FIG.
A DC loop is also provided, indicated throughout with the reference numeral 1. The microphone 5 does not have the ability to detect very low frequency components of the output from the acoustic mixer 15, and this component, in addition to the DC component, is detected by the adaptive filter 2.
This is required for stable operation of the LMS algorithm in No. 3. Another DC loop is included to provide a DC component and the LMS algorithm has been modified to accommodate such input.

Referring now to FIGS. 1I-6, an evolution of the electronic components of active attenuator 11 is shown. The electronic components of active attenuator 11 include separate adaptive filters, each adaptive filter performing a separate function. Before explaining the adaptive filter process,
The functions of the remaining circuit elements will be explained. Active attenuator 1
This embodiment of the electronic circuit of 1 uses a digital implementation of an adaptive algorithm. 66 has been added to all of Figures 4-6. An analog-to-digital converter (A/D) provides sample values of the selected input in digital format. The A/D converter 6 is a 12-bit successive approximation converter.

Most of the signal processing within the electronic circuit of the present invention is digital, and the sampling rate of A/D converter 65 represents an upper limit on the allowable bandwidth of the signal and error inputs. The sampling rate limit requires that the highest input frequency must be less than half the sampling rate (115). A low pass filter 65 at the system input is required to limit the highest input frequency to less than half the A/D sampling rate. Digital/Analog (
The output signal from the D/A converter 67 also has significant frequency aliasing problems. The output signal from the adaptive filter can excite resonances in the system at any alias or multiple of the output of D/A converter 67.

Therefore, a low pass filter 65 is required at the output of the D/A converter 67 to limit the highest output frequency to the band of interest.

Multipliers 39, 45, 47, 69 and integrators or adders 111, 113, 119° 71 shown in FIGS.
performs the calculations necessary for the adaptive processing in the present invention. T
RW Model TDC1003J 12 bit
Parallel Multiplier-Accumu
lator or a suitable equivalent has been found to provide the necessary high speed computing power.

The first of the five adaptive filters utilized in the electronic circuitry of the active attenuator 11 may be described as a canceling adaptive filter 25 as shown in (kilo) FIG. 1.11. The transversal adapter 23 includes a modified LMS algorithm to govern the operation of its transversal filter in a dual fashion. A detailed explanation of the above description includes one phase response, amplitude, and delay.

The basic filter function of the canceling adaptive filter 2 is performed by a transversal filter, which is a non-recursive, finite impulse response type filter. This filter implementation is digital in nature, with input signal history and tap weights stored in constant speed memory. As mentioned above, the operation of the transversal filter is L M
Generate the product of the first weight obtained from the S algorithm and the final human sample, and this product is added to the product of the sample closest to the final input sample to the second way from the algorithm, and as follows: Finally, it can be described that the product of the electrical fault weight and the oldest sample is accumulated. In equations, this is given as:

here.

S i = sample input signal W corresponding to the i-th tap l = weight 1 of the i-th tap - tap identification ■ = number of taps offset adaptive filter 2 (way) When adjustment is stopped, it is regarded as a transversal filter I can do it. The canceling adaptive filter 25 operates by determining a required value (ie, optimal filter condition) for the weight of a desired filter by an adaptive means, and then using the determined weighter to realize the desired filter function in the transversal filter. The purpose of the modified LMS algorithm is to govern the operation of the transversal filter by matching the filter function of the canceling adaptive filter 25 to the overall desired filter function.

The error signal sensed by microphone 55 and sent to canceling adaptive filter 2 is a source pressure wave delayed by acoustic travel from that point to microphone 35;
It can be considered as the sum of a complementary wave generated by the speaker 17 and delayed in its passage through the waveguide 19 and the acoustic mixer 15 to the microphone 35. Microphone array 33, speaker 17. Waveguide 19. When physically placed relative to each other within the microphone array 15 and the acoustic mixer 15, the total delay or length of the microphone array until the speaker 17
The total delay or length from to the microphone 35 plus the delay associated with the operation of the low pass filter 65 must be greater than the delay associated with the operation of the low pass filter 65. In equations, this relationship is expressed as follows (see Figure 1) where:

λm1n - Shortest wavelength involved (lI9) L = Microphone array 5? Distance between I and microphone 35 C = speed of sound d, = distance from one waveguide to microphone 5 d2 = distance from acoustic mixer 15 (i.e. matched to microphone 35) to loudspeaker 17 Df - low Delayed low pass filter 6 associated with pass filter low 5
The delay Df of 5 can be expressed in terms of cutoff frequency or FmaX.

The fourth order low pass filter 65 has a total delay of 115 degrees for each pole at the cutoff frequency, which in good filter design approximates a constant delay filter. Therefore, the delay of the third-order filter can be approximately %Fcutoff,
Df is 17”Cut in the case of two sets of fourth-order filters in series.
It can be regarded as approximately equal to off.

The second adaptive filter found in the electronic circuitry of active attenuator 11 is an uncoupling filter, as shown in FIG. 5 (50).
It may be named adaptive filter 75. One potential problem associated with many active attenuators is the result of the canceling sound introduced by the loudspeaker propagating toward the microphone or microphone array that detects the original sound (7) due to standing waves or duct mechanical This coupling will adversely affect the signal sensing microphone's estimate of the original sound propagating downstream, reducing the effectiveness of the cancellation system. Although unidirectional microphones have been proposed, they alone are not sufficient to overcome all of this inherent system limitation.

Decoupling adaptive filter 75 solves this problem as follows. As shown in FIG. 5, broadband noise source 76 drives canceling loudspeaker 17 and decoupling adaptive filter 75. As shown in FIG. Before starting the system,
Noise source 76 is automatically adjusted to excite duct 13 at a higher acoustic level than the source level required to cancel the expected source noise. This adaptive process reduces the output of the error adder and matches the transfer function of the transversal filter in the decoupling adaptive filter 75 to the acoustic coupling present in one of the ducts, at which point the noise source 76 is terminated. .

At system start-up, the decoupling adaptive filter 75 acts on one drive of the loudspeaker 17. The loudspeaker drive component appearing in the output from microphone array 53 is removed by subtracting an equal and opposite image in summer 71. This subtraction process is accomplished electronically in the digital domain of decoupling adaptive filter 75 and adder Y1. Although this process does not completely eliminate the effect of the output from speaker 17 on microphone array 35, it can provide sufficient reduction in many applications.

Referring now to Figure 4.6, a third adaptive filter of the electronic circuitry of the active attenuator 11 is shown, which may be referred to as a compensation adaptive filter. 1st
To ensure stable operation of the modified L M S algorithm, as described above in connection with the description of the basic electronic components shown in the figures.

A compensation mechanism must be provided in series between the output of the canceling adaptive filter 25 and the input of the error signal.

Although this compensation mechanism is shown in FIG. 1 as two secondary phase correction filters, manual calibration of the two filters can be accomplished by adaptive filters. Seven compensation adaptive filters perform this function.

As shown in FIG. 6, there are seven compensating adaptive filters, 17 loudspeakers. The waveguide 19° acoustic mixer 15 and the microphone 35 are arranged in series with each other and in parallel with a broadband delay circuit consisting of a delay or memory 78. Prior to system start-up and after completion of the membrane bonding process described above, noise source 76 is activated to fully excite duct 13 and acoustic mixer 15. The sum of the parallel signal paths becomes the error input to seven compensating adaptive filters.

This adaptation process matches the overall transfer function of the series path to the true delay of the noise source 76 by delay 78. In this way, weights for the desired filter are generated such that the error signal received by the canceling adaptive filter 23 is within proper phase tolerance to ensure stable operation.

Delay circuit 78 of FIG. 6 delays the signal from noise source 76 by an integer number of total input sample intervals (K,).

A variant of the L M S fulborism of the adaptive filter 25 of FIG. 4 provides a sample interval delay in the calculation of the next value of Wi. The value of K is loudspeaker 1
The compensation adaptive filter of FIG. 7 through FIG. The value of K can be set large so that one value can be used in many applications. By matching the delay from the output of the canceling adaptive filter 25 of FIG. 4 to the error input to the shift in the sample signal used to calculate the next weight value, the stability of the modified LMS algorithm is ensured.

Next, in conjunction with the use of active acoustic attenuators in a free-space environment, a person emits sound waves from another source in a phase relationship with the original sound piece such that the sound pressure in the area concerned is reduced substantially (51k). The general concept of reducing the sound power generated by a sound source by doing so is not new. for example,
Attempts were made shortly after the Second Ascension to reduce the sound level in the ears of people working in noisy environments. Due to the state of electronics technology at the time, the profits? IH was relatively low and the equipment was bulky. As mentioned above, various approaches have been taken to develop systems that allow reduction of noise propagation within ducts. Techniques in the field of sound attenuation, including free space situations
ring (May-June 1982), GlennE
, by Warnaka, 'to ct, ive A
ttenuationof No.1se-The 5t
See ate of tbe Art'.

Before discussing the use of active acoustic attenuators in free space environments, some theoretical aspects of source interference in free space will be discussed. In the case of a simple monopole sound source in infinite space, the sound intensity Ir in the radiation direction is expressed as follows.

Ir=S ρc/2[k/'↓πr] (5) In this expression, r is the distance from the sound source.

k is the wave number (equal to ω/C), ρ is the density of the medium, C is the speed of sound, and S is the strength of the sound source (average amplitude of the surface vibration velocity multiplied by the surface area of the source). ).

゛ If there are two sound sources S, , S2, and each sound source has an amplitude equal to S but opposite sign, the sound intensity is approximately expressed by the following equation.

I r=s2ρc/2 [k/2πr]2sin2(
kd c6sθ) (6) As shown in Figure 7,
The two sources are separated by a distance 2d, and the angle θ is measured from the line joining the two sources. In the derived representation of sound intensity for two monopoles, r is much larger than d (i.e., this representation is for remote regions). As can be seen from the two monopole representations, the sound cancels when θ equals 90°.

Valid along the axis of symmetry between the two sound sources.

CO8θ=nλ/2d, n=o, 1.2...
It can be shown that zero (cancellation) is generated when (7). In this expression,
λ is the wavelength of each frequency of sound, and CO8θ is of course limited to 1 or less. Since cos θ is less than 1, n
Values of n such that λ/2d is less than or equal to 1 are allowed.

Therefore, if d is large compared to λ (λ/4(1
), the maximum allowed value of n can be high, which means that the radiation pattern consists of a large number of ropes.

The zero for θ-00 is relatively narrow and the area of reduced sound pressure level is small.

Therefore, even though radiation into the sidelobes is reduced by absorption, this source arrangement is fairly ineffective and impractical. The closer the sound source is, the more θ−π/2
The zero (cancellation) near is widened. If d/λ or kd becomes small, the two monopoles are treated as dipoles. In the case of a dipole, the approximate sound intensity is expressed by the following formula:

I r-[s2ρc/2':] [2dk/11πr]
2cO82θ (8) As can be seen from equation (8), in order to achieve an effective noise reduction by an additional source that radiates the entire sound of opposite phase (57), the position of this sound source should be as close as possible to the original sound source. Must be nearby. The distance 2d between the sound sources is the dipole condition k(2
a) = 2π(2d/λ) (1 must be satisfied. The higher the frequency.

The distance must be small.

Meeting these conditions at high frequencies is more difficult due to the physical size of the sound source.

However, in this higher frequency range, an enclosure surrounding the free space environment usually provides good insertion loss. On the other hand, the low frequency range is usually more difficult to attenuate passively, and in this case active noise control systems are better applied.

The total radiated sound power is Ir ds over the spherical surface surrounding the sound source.
can be determined as the double integral of Relative to the sound power radiating from a single monopole, the sound power of a double monopole is twice as large when they are far apart. However, in the case of a dipole, the total radiated sound power is smaller than the sound (58) power in the case of two monopoles. There are methods to calculate far-field radiation characteristics using all near-field techniques, the exact amount of total radiated sound power depending of course on the U1 spacing of the sound source and the frequency of the sound produced by the source. For example, J, A, cost.

f3oc, Am,, 6+6111. l October 197
9. Robert D.

By Marciniak. 'Near Field
d. Underwater Measurement
Marciniak
paper describes a near-field acoustic measurement technique for accurately calculating the far-field radiation curve of acoustic transducers. The analysis is based on the Green
)'s Sekifi' (Helmholtz using r)
ltz) is based on the evaluation of the form of the integral, whereby only knowledge of the near-field acoustic pressure is required for the total calculation of the far-field radiation.

From the theoretical consideration of remote domain calculation, and for example, regarding the sound source array in FIG.
r image)'' can be a canceling sound source. The second sound source is seen to generate a sound intensity pattern in a distant area that can be substantially calculated. As previously mentioned, the two As the sound sources are brought closer together, more complete cancellation occurs.With further use of techniques such as those described in the Marci, N1AK paper, the far-field effects of more complex near-field arrays can be calculated. The active acoustic attenuators described above are particularly suitable for use in such close area arrays in order to produce good acoustic cancellation and/or reduction in some or all of the remote areas.Multiple cancellations Appropriate near field positioning for the sound source can be determined experimentally or by utilizing the computational techniques described above.

For example, as shown in FIG. 8, the active acoustic attenuator 11 is a speaker S. 1 and a second cancellation sound generator S. Contains 2.

Both canceling sound generator S. I, SC2 is driven by the common output of the active acoustic attenuator circuit. The attenuator 11 also includes first sensing means, such as a microphone M1, for detecting vibrations from the source S of the sound to be canceled. A second sensing means, such as a microphone M2, detects the sum of the vibrations from the sound source S and the vibrations of the canceling sound generator "CI, S02" to generate an error input signal for the circuit of the attenuator 11. .The single cancellation loop of the attenuator can accommodate the average acoustic delay from the generator to the error detection means.In the above configuration, the microphone M1 is substantially coincident with the source S of the sound to be canceled; The sound source and the microphone M2 are arranged in a plane.

The electrical circuit portion of active acoustic attenuator 11 shown in FIG. 8 may be substantially the same as that shown in FIG. However, it has been determined that certain modifications of the circuit result in improved performance of the attenuator, as shown in FIG. Referring to FIG. 9, compensation adaptive filter 79
' is removed from the cancellation circuit and placed in series with the error signal path from the error microphone to the cancellation adaptive filter 25'. Compensating adaptive filter 79' is initialized as shown in FIG. 6 and then held in position in the circuit during operation of the attenuator circuit. Relocating the compensation adaptive filter from the cancellation loop (61) means that the pickup microphone 35 and the cancellation speaker 1
This eliminates some of the delay introduced between Therefore, the cancellation speaker may be placed closer to the pickup microphone and therefore the sound source.

10-14, a series of far field radiation curves represents a simulation of a double canceling source acting on a single point source of noise. In each case, frequency is in hertz and distance is in feet. The curve was placed in the center of the grid system. It represents a point source of sound that is to be canceled out.

The source strength is approximately +30 dB. Each 10 dB is indicated by a dotted circle concentric around the center of the noise source or grid. The +30 dB circle is the outermost circle of the five and is shown in a slightly more intense manner to indicate the strength of the source. The orientation of the cancellation source and detector in each simulation substantially corresponds to the orientation shown in FIG. The two matched cancellation sources are symmetrically spaced apart from the noise source on the horizontal axis (1411), which is also common to the error sensor. This system is implemented using a single adaptive acoustic cancellation loop as disclosed in (-) for the acoustic attenuator 11.

The solid curve is the far range radiation pattern of the combined sound source and two canceling sound generators. This Prone I represents only distant area noise reduction in 60 sweeps centered on the noise source. There is no distance information on the plot. If the solid line pattern exceeds the thick dotted circle, the combined region strength is greater than the original sound source. In other words, the canceling device is increasing the radiated sound rather than reducing it. The advantage of greatly reducing the radiated sound level, often without a large cone't, is that it overcomes the problem caused by the increase in sound in the other direction.

The comparison of FIG. 10.12 and also the comparison of FIG. 11.15 shows improved performance at lower frequencies for the same source-cancelling generator spacing. As mentioned above, the general suggestion is that the smaller the source spacing or the longer the wavelength (lower frequency), the better the performance in a two-source arrangement. This also holds true for a sound source placed in a straight line in the simulated radiation curve (d
″!.

It looks like it is. As further seen in Figure 111,
As the distance between the source of the sound to be canceled and the source of the canceled sound becomes significantly narrower, the far field radiation curve becomes more complex. In this case as well, as shown in FIG.
Even in the illustrated embodiment, a cone 1 of canceled or substantially attenuated sound is generated.

In the illustrated radiation plot, the error microphone is spaced apart from the source of the sound to be canceled in the direction of the desired cancellation. In the plot, this is to the right of the source along the horizontal axis. Although the illustrated radiation plot is for a single frequency, the active acoustic attenuator of the present invention provides broadband reduction of acoustic vibrations or noise. The far field radiation pattern produced by the attenuator varies over the effective spectrum of attenuation, depending of course on the spectral content of the noise.

[Brief explanation of the drawing]

FIG. 1 is a diagram of a simplified form of an active acoustic attenuator. FIG. 2 is a partial block diagram of a digital implementation of an adaptive filter with the standard Widrow-Hoff T, MS algorithm. FIG. 5 is a partial block diagram of a digital implementation of the adaptive filter of the present invention using a modified Widrow-Hoff LMSMS algorithm. FIG. 4 is a block diagram of the electrical circuit components of the adaptive filter. FIG. 5 is a block diagram of the membrane-bonded adaptive filter portion of the adaptive filter. FIG. 6 is a block diagram of the compensation adaptive filter portion of the adaptive filter. FIG. 7 is a schematic illustration of the relative orientation of the source of the sound to be canceled and the source of the canceling sound. Figure 8 shows a seventh device with detection means and active acoustic attenuator.
2 is a schematic diagram of the relative orientation of the sound sources in the figure; FIG. FIG. 9 is a diagram of the blower (65) in a modified form of the electric circuit of FIG. 10-14 are far field simulation radiation curves for certain positions of the active acoustic attenuator relative to the sound source. 11--active acoustic attenuator, 15--duct, 15--acoustic mixer, 17--speaker, 19--waveguide, 21--
- enclosure, 25 - adaptive filter, 25 - amplifier, 29 - phase correction filter, 51 - low pass filter, 5 - microphone array, 35 - microphone, 7 - sample value, 3 m - Multiplier, 111, 11
3--adder, 115.117-multiplier, 119--adder, 51--unit delay, 53--) Lanceversal filter, 55--LMSMS algorithm 5
--A/D converter, 65--low-pass filter, 67--D
/A converter, 69--multiplier, 71--adder, 7
5-- Theory coupled adaptive filter, 76-- Noise source,
78--memory, 79--compensating attube filter,
S. ,,So2-One-cancellation sound generator 5S--Sound source,M,,
M2-1 microphone. (66) Fig 10 Ft'g // Fig /2 F79 13 Fig /4 Proceedings Amendment Written on January 19, 1981 Kazuo Wakasugi, Commissioner of the Patent Office L Case indication 1982 Patent Application No. 206067 No. Otsu
The name of the invention (device), the active sound attenuation device of the product of the guide, and the relationship with the person making the amendment Patent applicant 9 (name) Lord Corporation 1 Agent 6 Number of inventions increased by amendment Helicopter
Invention 7, Subject of Amendment Column 8 of Detailed Description of the Invention, Contents of Amendment The following sentence is added between page 27, line 10 and line 1 of the specification, as provided separately. "One object of this invention is to provide an active sound attenuation device suitable for attenuating sound propagating in a plurality of different directions from a source. Related. A further particular object is to provide an active sound attenuation device suitable for attenuating sound propagating in a plurality of different directions from a source. The object of the present invention is to provide an active sound attenuation device suitable for attenuating sound propagating within enclosures as well as within localized ducts or the like.'' (2)

Claims (1)

Claims: An apparatus for canceling vibrations from an L source. comprising first sensing means for detecting said vibrations, said first sensing means generating a first electrical signal representative of amplitude and phase characteristics of said vibrations; said vibrations from said source; Comprising canceling means for generating canceling vibration for coupling. The canceling means includes a first canceling vibration generator spaced from the source in a tenth propagation direction of the vibration and a second canceling vibration generator spaced from the source in a second propagation direction of the vibration. and a second sensing means disposed at a position spaced apart from the first sensing means in the propagation direction of the vibration, the second sensing means detecting the vibration from the source and the canceling means. detecting the sum of the vibrations and generating a second electrical signal representing the amplitude and phase characteristics of the sum; comprising an adaptive filter for driving the canceling means to generate the canceling vibrations; The front F7 adapter filter has a delay inherent in the propagation of the I'Jil vibration from the first detection means to the second detection means, and a delay inherent in the propagation of the I'Jil vibration from the first and second canceling vibration generators to the canceling vibration. and a delay inherent in the propagation of the canceling imaging to a point of coupling with vibrations from the source. The filter sequentially delays, filters and scales the first electrical signal from the first sensing means to produce a corresponding output for driving the canceling means, and then the second sensing means. 9) determining said delaying, filtering and scaling operations based on said second electrical signal received from said means for providing a corresponding corrective output, thereby driving all said canceling means to reduce said delaying, filtering and scaling operations within preset limits; A cancellation signal is generated that has mirror image amplitude and phase characteristics of the vibrations from the source. The apparatus characterized in that it is operable to: 2. The device according to claim 1, wherein the second sensing means is arranged at a position spaced apart from the first sensing means in the direction of propagation of the vibrations in which cancellation is desired. 3. The apparatus of claim 1, wherein the first offset vibration generator, the second offset vibration generator, and the second sensing means are aligned in a linear arrangement. )■. The device according to claim 1, wherein the first sensing means is substantially located at the vibration source. 5. Apparatus according to claim 1, wherein the second sensing means is arranged at a position spaced apart from the first sensing means in the direction of propagation of the vibrations in which cancellation is desired. 6. A device for canceling vibrations from a source, comprising a first detection means for detecting all of the vibrations, the first detection means having a first detection means representing the amplitude and phase characteristics of the vibrations. generating an electrical signal; canceling means for generating a canceling vibration to combine with vibrations from said source; a second sensing located at a location spaced apart from said first sensing means in the direction of propagation of said vibrations; means, the second sensing means detecting the sum of the vibrations from the source and the canceling means and generating a second electrical signal representative of the amplitude and phase characteristics of the sum; an adaptive filter for driving the canceling means to generate the canceling vibration, the adaptive filter having a delay inherent in the propagation of the vibration from the first detecting means to the second detecting means; and L M modified to accommodate the delay inherent in the propagation of the canceling vibration from the canceling means at the coupling point dimension of the canceling vibration and the vibration from the source.
a transversal filter using an S algorithm, said adaptive filter continuously delaying, filtering and scaling said first electrical signal from said first sensing means to drive said canceling means; 1 and then decisively adjusts said delaying, filtering and scaling operations based on said second electrical signal received from said second sensing means to generate a corresponding modified output. A canceling means is actuated to generate a canceling vibration having a mirror amplitude and phase characteristic of the vibration from the source within preset limits. and the adaptive filter is further operable to determine the phase characteristics of the second electrical signal generated by the second sensing means for introducing into the canceling adaptive filter to ensure stable operation thereof. a compensating adaptive filter having means for adjusting the compensation adaptive filter, the compensating adaptive filter being coupled in series between the second sensing means and the input of the canceling adapter (5); The said device which does. 7. Control device for a device for canceling vibrations from a source, modified to accommodate acoustic delay LM
a canceling adaptive filter including a transversal filter using an S algorithm, the canceling adaptive filter having an input coupled to the first sensing means for detecting all vibrations from the source of the vibrations to be canceled; an output coupled to a canceling means for generating a canceling vibration to be coupled with a vibration from a source, and an error input, and wherein the first canceling means is coupled to the canceling means for generating a canceling vibration to be combined with the vibration from the source, and an error input, and a compensating adaptive filter coupled between a second detecting means disposed at a position spaced apart from the first detecting means and an error input of the canceling adaptive filter; Said control device, characterized in that the Y-finger is operable to decisively adjust the phase characteristics of the output (6) of the second sensing means in order to ensure that: 8. A device for canceling vibrations in a physical system. a first detection means for detecting said vibration, said first detection means generating a first electrical signal representing amplitude and phase characteristics of said vibration; said vibration from said physical system; canceling means for generating canceling vibration for coupling; comprising a second detecting means disposed at a position spaced apart from the first detecting means in the propagation direction of the vibration, the second detecting means detecting a sum of vibrations from the physical system and the canceling means and generating a second electrical signal representative of the amplitude and phase characteristics of the sum; an adaptive filter, the adaptive filter comprising a modified L; M S algorithm and all transversal filters; the adaptive filter continuously delays, filters, and scales the first electrical signal from the first sensing means to drive the canceling means; generating a corresponding output of and then determining said delaying, filtering and scaling operations based on said second electrical signal received from said second sensing means to provide a corresponding modified output, thereby said canceling Means is actuated to generate canceling vibrations having mirror amplitude and phase characteristics of said vibrations from said physical system at preset limits. The device characterized in that it is operable to: 9. The modified LMS algorithm is configured to account for the delay inherent in the propagation of the vibration from the first sensing means to the second sensing means and the delay between the canceling vibration from the canceling means and the vibration from the physical system. 9. The apparatus of claim 8, including means for accommodating the delay inherent in the propagation of said canceling vibrations to a point of coupling. 10 A device for canceling vibrations in a physical system. comprising a first sensing means for detecting said vibration, said first sensing means generating a first electrical signal representing amplitude and phase characteristics of said vibration; said vibration from said physical system; canceling means for generating canceling vibration for coupling; comprising a second detecting means disposed at a position spaced apart from the first detecting means in the propagation direction of the vibration, the second detecting means detecting a sum of vibrations from a physical system and the canceling vibration and generating a second electrical signal representing amplitude and phase characteristics of the sum; driving the canceling means to generate the canceling vibration; Equipped with an adaptive filter. The adaptive filter has a delay inherent in the propagation of the vibration from the first sensing means to the second sensing means, and a delay inherent in the propagation of the vibration from the canceling means to a coupling point between the canceling vibration and the vibration from the physical system. a transversal filter and an LMS algorithm modified to accommodate a delay (9) inherent in the propagation of said canceling vibration; 5. successively delaying, filtering and scaling 5 to produce a corresponding output for driving said canceling means.
The delay, filtering and scaling operations are then deterministically adjusted based on the second electrical signal received from the second sensing means to provide a correspondingly modified output, thereby driving the canceling means to preset. The apparatus is operable to generate canceling vibrations having amplitude and phase characteristics that are mirror images of the imaging from the physical system within limits. IL A device for completely canceling vibrations in a physical system, comprising a first detection means for detecting the vibration, the first detection means generating a first electric signal representing the amplitude and phase characteristics of the vibration. (10) A canceling means for generating a canceling vibration to combine with the vibration from the physical system; a canceling means for generating a canceling vibration to combine with the vibration from the physical system; 5. The second detection means detects the sum of the vibration from the physical system and the canceling vibration, and generates a second electrical signal representing the amplitude and phase characteristics of the sum. comprising electronic circuit means including a canceling adaptive filter and a compensating adaptive filter, the compensating adaptive filter deterministically adjusting a phase characteristic of the second electrical signal generated by the second sensing means; is operable to introduce this into said canceling adaptive filter to ensure its stable operation, said canceling adaptive filter having a modified LMS algorithm and a transversal filter, said canceling adaptive filter successively delaying, filtering and scaling said first electrical signal from said means to produce a corresponding output for driving said canceling means; decisively adjusting the delay, filtering and scaling operations based on a second electrical signal to provide a corresponding modified output, thereby driving the canceling means to eliminate the vibrations from the physical system within preset limits. generates canceling vibrations with mirror image amplitude and phase characteristics. The device characterized in that it is operable to: 12. The modified LMS algorithm is configured to reduce the delay inherent in the propagation of the vibration from the first sensing means to the second sensing means and the vibration from the canceling means to the canceling vibration and the vibration from the physical system. 12. Apparatus according to claim 11, including means for accommodating the delay inherent in said canceling oscillations up to the point of coupling with said canceling oscillations. 13. A device for canceling vibrations in a physical system. comprising a first sensing means for detecting said vibration, said first sensing means generating a first electrical signal representing amplitude and phase characteristics of said vibration; said vibration from said physical system; canceling means for generating canceling vibration for coupling; comprising a second detecting means disposed at a position spaced apart from the first detecting means in the propagation direction of the vibration, the second detecting means detecting the sum of the vibration from the physical system and the canceling vibration and generating a second electrical signal representing the amplitude and phase characteristics of the sum; driving the canceling means to generate the canceling imaging; With offset adaptive filters. a compensating adaptive filter operable to decisively adjust the phase characteristics of the second electrical signal generated by the second sensing means for introducing into the canceling adaptive filter to ensure stable operation thereof; , said canceling adaptive filter (15) is configured to reduce the delay inherent in the propagation of said vibrations from said first sensing means to said second sensing means; an LMS algorithm and a transversal filter modified to accommodate a canceling vibration and a delay inherent in the propagation of the canceling vibration to a coupling point with the vibration from the foreground physical system, the canceling adaptive filter sequentially delays, filters and scales the first electrical signal from the first sensing means to produce a corresponding output for driving the canceling means, and then the second sensing means. determining the delay, filtering and scaling operations to provide a corresponding modified output based on the second electrical signal received from the second electrical signal, thereby driving the canceling means to offset the physical system within preset limits. generating a canceling vibration having mirror image amplitude and phase characteristics of said vibration;
The device characterized in that it is operable to: (ill)