EP3371806A1

EP3371806A1 - Multi-glazed window incorporating an active noise reduction device

Info

Publication number: EP3371806A1
Application number: EP16809474.6A
Authority: EP
Inventors: Christian Carme
Original assignee: Technofirst SA
Current assignee: CARME, CHRISTIAN; Technofirst SA
Priority date: 2015-11-02
Filing date: 2016-11-02
Publication date: 2018-09-12
Anticipated expiration: 2036-11-02
Also published as: FR3043241A1; US10161180B2; US20180313138A1; WO2017077234A1; FR3043241B1; EP3371806B1

Abstract

The invention relates to a multi-glazed window formed by a frame produced from profiles supporting at least two glass panes separated by an air layer, said window incorporating an active noise reduction device for a noise coming from a noise source, which device comprises: - at least one loudspeaker which is in the form of a hollow body installed in one of the profiles of said frame and one face of which is constituted at least partially by a vibrating membrane disposed between the two adjacent glass panes in such a way as to vibrate and generate a counter-noise in the air layer, - an actuator associated with the membrane, which actuator is capable of inducing a vibratory movement of said membrane, - at least one control microphone carried by the frame, said microphone being installed in the air layer in order to sense the acoustic signals in said air layer, - a control electronics suitable for controlling the actuator according to the acoustic signals sensed by the control microphone, characterised by the fact that the control microphone is closer to that glass pane which is further from the noise source than the other glass pane.

Description

MULTI-WINDOW WINDOW INTEGRATING A DSSPOSSTSF OF ACTIVE NOISE REDUCTION

Description

TECHNICAL FIELD OF THE INVENTION

The subject of the invention is a multi-glazed window incorporating an active noise reduction device.

It concerns the technical field of devices for improving the sound insulation of a window.

State of the art

Document US Pat. No. 6,285,773 (Carme) discloses an active noise canceling system, comprising one or more linear loudspeakers arranged at the edge of a double glazing, in the air gap between the two panes and / or inside. a frame profile of this double glazing. In this noise-canceling system, the loudspeaker makes it possible to produce an electro-acoustic system which is practically invisible, and which does not impair either the visual comfort or the light transmission of the glazing, the proposed system making it possible to improve the sound insulation of a loudspeaker. double glazing especially in the low frequencies.

The loudspeaker described in the Carme patent comprises a vibrating membrane disposed between two adjacent panes so as to vibrate and generate a counter-noise in the air space. This membrane is associated with an actuator adapted to induce a vibratory movement to said membrane. An electronics control device controls the actuator based on the acoustic signals picked up by at least one control microphone carried by the window frame. The Carme patent, however, does not dwell on the position that the control microphone must have to optimize noise filtration.

The patent EP 0,710,946 (CENTER SCIENTIFIC AND TECHNICAL BUILDING) also relates to a multi-glazed window incorporating an active noise reduction device. In this document, it is taught to position control microphones in the middle of the air space, at equal distances from the two panes, in the longitudinal median plane of the window. The results achieved in terms of noise attenuation, however, are not optimal. In addition, attenuation is effective only in a narrow band of frequencies corresponding to low frequencies. The patent document CN 201.620.733 (XINMIN) also discloses an active noise canceling system comprising a loudspeaker arranged in triple glazing, in the air gap separating two panes. The speaker and the control microphone are in the same plane. The invention aims to remedy this state of affairs. In particular, an object of the invention is to improve noise attenuation in a multi-glazed window of the known type of the aforementioned prior art.

Another object of the invention is to obtain attenuation of noise in a wide frequency band.

The solution proposed by the invention is a multi-glazed window formed by a frame made of profiles supporting at least two panes separated by an air knife, said window having a longitudinal median plane and incorporating an active noise reduction device from a noise source, which device comprises:

at least one loudspeaker which is in the form of an elongate rectangle parallelepiped shaped hollow body, one face of which is constituted at least partially by a vibrating membrane arranged between the two adjacent panes so as to vibrate and generate a counter-noise in the air space,

an actuator associated with the membrane, which actuator is adapted to induce a vibratory movement to said membrane,

at least one control microphone carried by the frame, which microphone is installed in the air space for sensing the acoustic signals in said air space,

a control electronics adapted to control the actuator as a function of the acoustic signals picked up by the control microphone. This window is remarkable in that:

the hollow body forms one of the profiles of the frame,

the membrane is disposed in the middle of the two panes, symmetrically with respect to the longitudinal median plane of the window,

- The control microphone is offset from the longitudinal center plane of the window so that it is closer to the window that is farthest from the source of noise than the other window.

Thanks to this position of the control microphone, the applicant was able to note, surprisingly, that the attenuation of the noise was efficient and stable, in a wider frequency band than that indicated in the patent EP 0,710,946 mentioned above.

Other advantageous features of the invention are listed below. Each of these features may be considered alone or in combination with the outstanding features defined above, and may be subject, where applicable, to one or more divisional patent applications: - The control microphone can be installed on the profile formed by the hollow body of the speaker, said control microphone being adjacent to the membrane.

In an alternative embodiment, the control microphone is installed on a profile that is remote from the profile formed by the hollow body of the loudspeaker.

- The control microphone is advantageously oriented in a direction that is perpendicular to the direction of propagation, in the air space, acoustic signals from the noise source.

- The control electronics advantageously comprises a feedback filtering means having an input connected to the control microphone and an output connected to the actuator.

- Preferably, at least one reference microphone is carried by the frame, which reference microphone is installed outside the air gap, at the window which is closest to the source of noise; the control electronics in this case comprises an anticipatory filtering means, having an input connected to the reference microphone and an output connected to the actuator.

- The control microphone and the reference microphone can be carried on the same profile or each with a separate profile.

The reference microphone is advantageously oriented in a direction which is parallel to the direction of propagation of the acoustic signals coming from the noise source.

- Preferably, the control electronics comprises a summing means having a first input, a second input and an output connected to the actuator; the feedback filtering means comprises an input connected to the control microphone and an output connected to the first input of the summing means; and the anticipatory filtering means comprises an input connected to the reference microphone and an output connected to the second input of the summing means. The anticipation filtering means may be of the adaptive type and comprise: a first input connected to the control microphone; and a second input connected to the reference microphone.

The anticipatory filtering means may also be of the non-adaptive type.

- The speaker can be a linear speaker or a circular speaker. esorjtion _^ s figy _res;

Other advantages and characteristics of the invention will appear better on reading the description of a preferred embodiment which will follow, with reference to the accompanying drawings, carried out as indicative and non-limiting examples and in which:

FIG. 1 is a diagrammatic front view of a multi-glazed window with an active noise reduction device comprising a linear loudspeaker,

FIG. 2 is a diagrammatic sectional view along A-A of the window of FIG. 1, according to a first embodiment,

FIG. 3 is a diagrammatic sectional view along A-A of the window of FIG. 1, according to a second embodiment,

FIG. 4 is a diagrammatic sectional view along A-A of the window of FIG. 1, according to a third embodiment,

FIG. 5 is a diagrammatic sectional view along A-A of the window of FIG. 1, according to a fourth embodiment,

FIG. 6 is a schematic front view of a multi-glazed window with an active noise reduction device comprising a plurality of circular loudspeakers,

- Figure 7 is a graph showing the acoustic attenuation likely to be provided by a window according to the invention. Preferred embodiments of the invention

The present invention relates to a multi-glazed window, which is characterized by a particular design of the active noise reduction device that it incorporates.

The window itself is of the known type. In Figure 1, it consists of a frame 19, or frame formed of sections 19a, 19b, 19c, 19d enclosing a glazed panel 4. The frame 19 is preferably of rectangular or square shape, but can be polygonal, have one or more curved edges, etc. In Figure 2, the panel 4 is formed by two adjacent panes V1 and V2 separated by an air knife L. The noise reduction device is used for active noise control. It generates in the air space a sound level equivalent to the ambient sound level to be controlled, including noise from a source of noise S.

An active noise reduction device may be in the form of a piezoelectric actuator or loudspeaker. A linear loudspeaker of the type described in the aforementioned US Pat. No. 6,285,773 (Carme) is preferably used and to which the person skilled in the art can refer if necessary. This type of linear speaker can indeed be easily housed in a reduced volume and in particular in a narrow space, while having a performance comparable to that of a conventional speaker conical membranes. The geometric shape and the particular arrangement of the constituent elements of the linear loudspeaker offer a very satisfactory performance. In particular, given the long length of the membrane, it moves a large mass of air during its vibration, which provides a good performance in the low frequencies. The linear speaker also generates a sound wave whose phase is uniform over the entire width of the glazing. Figure 6 illustrates an alternative embodiment not covered by the invention wherein the linear speaker is replaced by several circular speakers installed side by side in the section 19a. One can for example use ASC A loudspeakers marketed by the plaintiff. Using a linear speaker, however, reduces the number of loudspeakers to achieve equivalent noise reduction.

In the remainder of the description, the generic term "loudspeaker" is used, whether the latter is a loudspeaker as such or a piezoelectric actuator.

The noise reduction device may comprise a single HP linear speaker disposed on only one of the sides 19a of the frame 19, or several speakers disposed respectively on the different sides 19a, 19b, 19c, 19d of said frame. The choice of the number of HP speakers and their arrangement in the frame 19 depends on the sound field to attenuate, by superposition, noises propagating in the air gap L, to increase the sound insulation of the double glazing.

FIG. 2 gives a schematic representation of a linear loudspeaker, which is externally like a hollow body 1 in the form of an elongate rectangular parallelepiped, having for example a length of 50 cm to 2 m, a width of 2 cm to 4 cm and a depth of 2 cm to 4 cm. The body 1 can be made of aluminum, steel, plastic, or any other material that is suitable for those skilled in the art and advantageously forms one of the profiles of the frame 19. In FIG. 2, the body 1 forms the profile horizontal 19a which is located at the bottom of the frame 19. At least one side of the loudspeaker HP is formed at least partially by a vibrating membrane 7 arranged between the two panes adjacent V1, V2 so as to vibrate and generate a counter-noise in the air gap L. This membrane 7 is flat and for the case of a linear speaker, it is elongated. It extends preferably over the entire length of the body 1. The membrane 7 is disposed in the middle of the two panes V1, V2 symmetrically with respect to the longitudinal median plane P of the window.

An actuator January 1 is associated with the membrane 7. This actuator January 1 is adapted to induce a vibratory movement to the membrane 7. It may be a piezoelectric actuator or more conventionally an actuator using a magnet arrangement and coil electrically excited to cause the vibration of the membrane 7 which generates the noise.

At least one control microphone 21, or error microphone, is carried by the frame 19. In FIG. 2, this microphone 21 is installed in the air knife L to pick up the acoustic signals propagating in the latter. By way of example, it is possible to use a control microphone 21 of the PUI Audio brand bearing the reference POM-2246L-C33-Ret manufactured by the company PUI Audio.

The microphone 21 sends a signal representative of the noise in the air gap L to a control electronics 23. Therefore, the control electronics 23 emits a control signal to the actuator January 1 according to the acoustic signals received by the microphone 21. This active noise reduction device increases the sound insulation of the double glazing. According to the invention, and as illustrated in FIG. 2, the control microphone 21 is installed in the air gap L, offset from the longitudinal median plane P, so that it is closer to the window V2 which is the farthest from the source of noise S than the other window V1. For example if the source of noise S is the ambient noise prevailing outside a room, a room or a cabin, (for example the noise of vehicles traveling in a street or on a road, the noise of an airplane engine, ...), the V1 pane is that which is located outside the room, the room or of the cabin, and the glass V2 is the one that is installed inside the room, the room or the cabin. If the source of noise S is the ambient noise prevailing inside a room, a room or a cabin (for example the music of a discotheque), the window V1 is the one which is located at inside the room, the room or the cabin, and the window V2 is the one that is installed outside the room, the room or the cabin.

Thanks to this position of the control microphone 21, the Applicant has surprisingly found that the noise attenuation was effective and stable, in a wider frequency band than that indicated in patent EP 0,710,946 supra. This phenomenon is explained below with reference to FIG. 7.

Good results are obtained when the control microphone 21 is oriented in a direction which is perpendicular to the direction of propagation, in the air space L, acoustic signals from the source of noise S. The control microphone 21 is thus oriented in a direction which is parallel to the direction of movement of the membrane 7, that is to say parallel to the longitudinal median plane P of the window. In this arrangement, it appears that the control microphone 21 satisfactorily collects the residual acoustic signal which is used as an error signal in the feedback filtering described hereinafter in the description. This residual acoustic signal is a combination of the residual noise reaching the glass V2 and a noise generated by the speaker HP which is ideally the inverted copy of the noise to be removed from the source S.

In FIG. 2, the control microphone 21 is installed on the section 19a formed by the hollow body 1 of the HP speaker. More particularly, the control microphone 21 is adjacent to the membrane 7. This configuration simplifies the design of the active noise reduction device insofar as all its components are grouped together in one and the same profile 19a. The control microphone 21 may, however, be installed on a profile 19b which is remote from the profile 19a formed by the hollow body 1 of the loudspeaker HP, as shown schematically in FIG. 3. In this figure, the control microphone 21 is arranged on a horizontal section 9b which is opposite to the horizontal section 19a formed by the hollow body 1 of the HP speaker. Of course, the control microphone 21 can be installed on one of the vertical profiles 19c or 19d, while the hollow body 1 of the HP speaker forms one of the horizontal sections 19a or 19b, and vice versa.

In FIGS. 2 and 3, the control electronics 23 comprises a non-adaptive type FB feedback filtering means having an input FBe connected to the control microphone 21 and an output FBs connected to the control microphone 21. actuator 11.

The active feedback attenuation technique is based on a feedback loop arranged to generate an active attenuation of the sound waves propagating in the air gap L. The signal measured by the control microphone 21 is injected at the actuator 1 1 through the feedback filtering means FB which corrects said signal to attempt to cancel its energy. This retroactive technique makes it possible to obtain an acoustic attenuation with a certain gain, without generating instability in a frequency band of treatment. Most often, this treatment frequency band corresponds to low frequencies, for example sound waves in the frequency band ranging from 0 to 400 Hz and more particularly from 70 Hz to 400 Hz.

The control electronics 23 advantageously comprises: pre-amplification means comprising an input connected to the control microphone 21 and an output connected to the input FBe of the feedback filtering means FB; and amplification means comprising an input connected to the output FBs of the feedback filtering means FB, and an output connected to the actuator 11. This control electronics 23 constitutes here a feedback loop arranged to generate an active acoustic attenuation without generating instability in a selected frequency band. For example, the frequency band in which the feedback filtering means is effective without generating instability in the sense of Nyquist, is of the order of 0 to 600 Hz for sound waves and more particularly 70 Hz to 600 Hz .

In practice, the feedback filtering means FB comprises a plurality of active analog filters of order greater than or equal to 1, arranged to generate a transfer function making it possible to avoid instabilities in the frequency band 0-600 Hz and more. particularly in the band 70-600 Hz in the Nyquist sense, and the transfer function of the filtering means FB is determined such that the phase of said transfer function does not pass through the value 0 in this band.

However, a pumping effect appears above 600 Hz which results in an increase in the noise level compared to the action of the passive attenuation means alone, that is to say the panel 4 alone. This phenomenon is well known to those skilled in the art, and constitutes a non-linearity (performance degradation) compared to the results expected from observation of the open-loop system.

To remedy this, it is advantageous to combine active attenuation by feedback with active attenuation by anticipation. In FIG. 4, the control electronics 23 comprises for this purpose an FF (FEEDFORWARD) anticipation filtering means, having an FFe input connected to a reference microphone 22 and an FFs output connected to the actuator. 11. By way of example, a reference microphone 22 of the PUI Audio brand bearing the reference POM-2246L-C33-Ret manufactured by the company PUI Audio can be used. In this proactive attenuation technique, a reference acoustic field, upstream of the propagation of the acoustic field in the air gap L, is detected by the reference microphone 22, then processed by the filtering means FF so to determine the command to be applied to the actuator

1 1.

To optimize the signal processing, provision is made for: preamplification means comprising an input connected to the reference microphone 22 and an output connected to the input FFe of the forward filtering means FF; and amplification means comprising an input connected to the output FFs of the forward filtering means FF, and an output connected to the actuator 11.

In FIG. 4, the control electronics 23 comprises summing means 24 having: a first input 24e 1 connected to the output FBs of the feedback filtering means FB; a second input 24e2 connected to the output FFs of the forward filtering means FF; and an output 24s connected to the actuator 11. The output signal of the summing means 24 which is applied to the actuator 11 is thus a linear combination of the signals coming from the filtering channels by feedback and anticipation. Advantageously, amplification means are provided comprising an input connected to the output 24s of the summing means 24, and an output connected to the actuator 11.

The anticipation technique is articulated around the FF anticipatory filtering means of non-adaptive or adaptive type. Compared to non-adaptive filtering, adaptive filtering is more efficient in terms of noise attenuation, but requires more computing power and higher realization cost. In the case where the forward filtering means FF is of the non-adaptive type, its transfer function is a fixed function which is preset and which does not vary.

With an adaptive FF anticipatory filtering means, the transfer function is dynamically modified, continuously, by an algorithm for real-time analysis of the acoustic signal from the source S. The coefficients of the FF anticipatory filtering means are adapted in real time according to an algorithm chosen so as to minimize the energy of the vibrations picked up by the control microphone 21 as a function of the energy of the reference vibrations picked up by the reference microphone 22.

This adaptive filtering is shown diagrammatically in FIG. 5, where the forward filtering means FF comprises: a first input FFe1 connected to the control microphone 21; and a second FFe2 input connected to the reference microphone 22. In practice, the FF anticipatory filtering means comprises adaptive-type finite impulse response filters. The coefficients of these filters are updated in real time by a minimization algorithm which takes into account the signals picked up by the control microphone 21. For example, the minimization algorithm is of the least mean squares type, also called LMS for " LEAST MEAN SQUARES "or more preferably of the least mean squares type with filtered reference, also called FXLMS for" Filtered-X Least Mean Squares ".

In a previous initialization step, the transfer function of the so-called secondary path between the loudspeaker HP and the control microphone 21 is measured, sampled and saved in the memory of a processor of the control electronics 23. previously measured transfer function will then be used in the calibration phase for the adaptation of anticipatory filter elements. This step is performed in a manner known to those skilled in the art.

The "hybrid" type active attenuation obtained according to the invention results from a combination of the anticipation and feedback filtering means in which the anticipatory filtering is grafted onto the feedback filtering or vice versa. This makes it possible to linearize the retroactive attenuation in a wider frequency band than the frequency band (0-600 Hz and more particularly 70-600 Hz) processed directly by the FB feedback filtering means, to accelerate the convergence. of the minimization algorithm, and to improve the robustness of the FF advance filtering means. This improves the active attenuation gain in an enlarged band that can go up to 4000 Hz, by suppressing the pumping effect mentioned above. In FIGS. 4 and 5, the reference microphone 22 is carried by the frame 19. Unlike the control microphone 21, it is installed outside the air gap L, at the window V1 which is closest to the source of noise S. The reference microphone 22 can thus optimally capture the copy of the noise to be removed from the source S and transmit this signal to the control electronics 23.

Good results are obtained when the reference microphone 22 is oriented in a direction that is parallel to the direction of propagation of the acoustic signals from the noise source S. The reference microphone 22 is thus oriented in a direction which is perpendicular to the direction of displacement of the membrane 7, that is to say perpendicular to the longitudinal median plane P of the window. In this arrangement, it appears that the reference microphone 22 satisfactorily collects the acoustic signal from the noise source S, without being disturbed by the noise generated by the speaker HP. To simplify the design of the noise reduction device, the reference microphone 22 and the control microphone 21 are carried by the same profile 19a. However, it can be provided that the control microphone 21 and the reference microphone 22 are each carried by a separate section. The reference microphone 22 may for example be arranged on a horizontal section 19b which is opposite to the horizontal section 19a formed by the hollow body 1 of the high betting HP and the control microphone 21. It can also be installed on the one of the vertical profiles 19c or 19d, while the speaker HP and the control microphone 21 are installed on one of the horizontal sections 19a or 19b, and vice versa.

Figure 7 is a graph showing the acoustic attenuation likely to be provided by a window according to the invention. Measurements were made on a 4-12-4 double glazed window (glass pane, air gap, glass thickness = 4 mm, air gap thickness = 12 mm). The curves correspond to the acoustic attenuation values in dB (ordinates) as a function of the frequency in Hz (abscissa). Table 1 below gives the different cases.

Table 1

By analyzing the curve # 1, we notice that the acoustic insulation that provides double glazing is relatively bad. The acoustic attenuation is low in the low and medium frequencies (150 Hz to 400 Hz, corresponding for example to slow road traffic noise) with a maximum fallback at the resonance frequency Fr (about 250 Hz). This resonance frequency depends on the mass of the panes V1, V2, their thickness and the nature of the elements (materials and air / gas gap) constituting the panel 4. Beyond this resonance frequency Fr, the acoustic insulation increases linearly up to the critical frequency Fc of the single panes V1 and V2 that make up the panel 4 (about 3000 Hz for a glass of 4 mm thick).

This is because the double glazing behaves like an acoustic system of the type Mass / Spring / Mass. The air gap L playing the role spring, its thickness is generally too low to create a sufficiently flexible spring and the system causes the windows V1 and V2 to resonate.

Curve n ° 2 corresponds to the case where the double glazing integrates the device of reduction of the noise. Only FEEDBACK feedback filtering is provided. The control microphone 21 is installed in the middle of the air gap L, as recommended by the patent EP 0,710,946 cited above. There is an improvement in sound insulation of about 8 dB in the range of low frequencies close to the resonance frequency Fr, in a band of about 200 Hz-350 Hz. There is also a decrease in the acoustic insulation compared to the acoustic insulation provided by double glazing alone (pumping effect beyond 650 Hz).

The curve No. 3 corresponds to the case where the double glazing integrates the noise reduction device, the control microphone 21 is now installed closer to the window V2 which is furthest from the source of noise S. Only is provided FEEDBACK feedback filtering. As on curve No. 2, there is an improvement in sound insulation of about 8 dB in the low frequency range close to the resonance frequency Fr. However, the acoustic insulation is improved in a wider band of about 150 Hz-375 Hz.

Curve n ° 4 corresponds to the case where the double glazing integrates the device of reduction of the noise. FEEDBACK feedback filtering and non-adaptive FEEDFORWARD anticipatory filtering are planned. The control microphone 21 is installed closer to the window V2. There is an improvement of the acoustic insulation of about 8 dB in the range of low frequencies close to the resonance frequency Fr, in a band of about 150 Hz-375 Hz (as on the curve No. 3). There is also an improvement in sound insulation of about 5 dB in the high frequency range close to the critical frequency Fc, which improvement is due to the anticipatory filtering. Curve n ° 5 corresponds to the case where the double glazing integrates the device of reduction of the noise. FEEDBACK feedback filtering and adaptive FEEDFORWARD predictive filtering are planned. The control microphone 21 is installed closer to the window V2. There is an improvement in sound insulation of about 10 dB in the range of low frequencies close to the resonance frequency Fr, over a wider band of about 125 Hz-400 Hz. There is also an improvement in the acoustic insulation of about 8 dB in the high frequency range close to the critical frequency Fc. The attenuation is here overall more effective compared to the curve 4. The combination of adaptive anticipation and feedback filtering improves the respective behavior of said filtering.

The arrangement of the various elements and / or means of the invention, in the embodiments described above, should not be understood as requiring such an arrangement in all implementations. In any case, it will be understood that various modifications can be made to these elements and / or means without departing from the spirit and scope of the invention. In particular :

- The window may have more than two windows, including three panes.

- Several control microphones 21 or reference 22 may be connected to the control electronics 23, these microphones being preferably installed on each of the profiles 19a, 19b, 19c, 19b of the frame 19; in this case, the control algorithm manages each channel with the objective of minimizing the pressure level on each of the error microphones, based on the information collected on the multiple reference microphones.

The FEEDBACK feedback filter can be adaptive, for example using a type I C-FXLMS algorithm for "Internal Model Control Filtered-X Least Mean Squares".

- Regarding control algorithms in FEEDBACK mode andor

FEEDFORWARD, the processing can be either analog or digital. - Out of scope of the present invention, the advance filter FEEDFORWARD can be used alone, without FEEDBACK feedback filter.

Claims

claims

1. Multi-glazed window formed by a frame (19) formed by profiles (19a, 9b, 19c, 19d) supporting at least two panes (V1, V2) separated by an air gap (L), said window having a longitudinal median plane (P) and incorporating an active noise reduction device from a noise source (S), which device comprises:

at least one loudspeaker (HP) which is in the form of an elongate rectangular parallelepiped shaped hollow body (1), one face of which is at least partially constituted by a vibrating membrane (7) arranged between the two adjacent panes (V1 , V2) so as to vibrate and generate a counter-noise in the air gap (L),

an actuator (1 1) associated with the membrane (7), which actuator (1 1) is adapted to induce a vibratory movement to said membrane,

- at least one control microphone (21) carried by the frame (19), which microphone is installed in the air space (L) for sensing the acoustic signals in said air space (L),

a control electronics (23) adapted to control the actuator (1) as a function of the acoustic signals picked up by the control microphone (21), characterized by the fact that:

the hollow body (1) forms one of the profiles (19a) of the frame (19),

the membrane (7) is arranged in the middle of the two panes (V1, V2), symmetrically with respect to the longitudinal median plane (P) of the window,

the control microphone (21) is offset from the longitudinal median plane (P) of the window so that it is closer to the window (V2) which is furthest from the noise source (S) than from the other window (V1).

2. Window according to claim 1, wherein the control microphone (21) is installed on the profile (19a) formed by the hollow body (1) of the speaker (HP), said control microphone (21) being adjacent to the membrane (7).

3. Window according to claim 1, wherein the control microphone (21) is installed on a profile (19b) which is remote from the profile (19a) formed by the hollow body (1) of the speaker (HP).

4. Window according to one of claims 1 to 3, wherein the control microphone (21) is oriented in a direction which is perpendicular to the direction of propagation, in the air space (L), acoustic signals from the noise source (S).

5. Window according to one of claims 1 to 4, wherein the control electronics (23) comprises a feedback filtering means (FB) having an input (FBe) connected to the control microphone (21) and an output (FBs) connected to the actuator (1 1).

Window according to one of Claims 1 to 5, in which:

- at least one reference microphone (22) is carried by the frame (19), which reference microphone (22) is installed outside the air gap (L), at the window (V1 ) that is closest to the noise source (S),

- The control electronics (23) comprises an anticipatory filtering means (FF), having an input (FFe) connected to the reference microphone (22) and an output (FFs) connected to the actuator (1 1).

7. Window according to claim 6, wherein the control microphone (21) and the reference microphone (22) are carried by the same profile.

(19a).

8. Window according to claim 6, wherein the control microphone (21) and the reference microphone (22) are each carried by a separate section (19a, 19b).

9. Window according to one of claims 6 to 8, wherein the reference microphone (22) is oriented in a direction which is parallel to the direction of propagation of the acoustic signals from the noise source.

(S).

10. Window according to one of claims 6 to 9 taken in combination with claim 5, wherein:

the control electronics (23) comprise summing means (24) having a first input (24e1), a second input (24e2) and an output (24s) connected to the actuator (11),

the feedback filtering means (FB) comprises an input (FBe) connected to the control microphone (21) and an output (FBs) connected to the first input (24e1) of the summing means (24),

the anticipation filtering means (FF) comprises an input (FFe) connected to the reference microphone (22) and an output (FFs) connected to the second input (24e2) of the summing means (24).

The window according to one of claims 6 to 10, wherein the forward filtering means (FF) is of the adaptive type and comprises:

a first input (FFe1) connected to the control microphone (21),

- a second input (FFe2) connected to the reference microphone (22).

12. Window according to one of claims 6 to 10, wherein the anticipatory filtering means (FF) is of the non-adaptive type.

13. Window according to one of claims 1 to 12, wherein the speaker (HP) is a linear speaker.