Signal control device, noise reduction device, signal control program, signal control method

Abstract

A noise reduction signal is generated to produce a suitable noise reduction sound without installing an error microphone at the sound elimination position. [Solution] The signal control device generates a noise reduction signal for generating a noise reduction sound from a speaker by adaptive signal processing according to the noise from the noise source, based on a noise signal corresponding to the sound acquired by the reference microphone and a virtual noise signal obtained by subtracting a virtual feedback signal corresponding to the feedback sound propagated from the speaker to the reference microphone from the noise signal.

Description

Technical Field

[0001] The present disclosure relates to a signal control device for controlling a signal for generating a noise reduction sound for reducing noise, a noise reduction device including the signal control device, and a method for controlling the signal. The present disclosure also relates to a signal control program for realizing the above-described signal control device by a computer.

Background Art

[0002] There is known a technology of active noise control (ANC) that generates a sound having a phase opposite to that of noise in order to reduce the noise at a predetermined position. In ANC, when the characteristics of the noise change due to a change in the surrounding environment of the predetermined position or the like, the characteristics of the sound for reducing the noise also change. Therefore, controlling the characteristics of the sound generated from the speaker according to the characteristics of the noise leads to reduction of the noise at the above-described predetermined position regardless of the characteristics of the noise. Patent Document 1 discloses a technology in which a monitor error microphone is installed in a target space where noise is to be reduced, sound is generated from the speaker into the target space so that the sound acquired by the monitor error microphone is reduced, and sound according to the characteristics of the noise is generated from the speaker.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the invention described in Patent Document 1, it is necessary to always install an error microphone on the side of the target space where noise is to be reduced rather than the speaker for ANC. As a result, in this invention, the space of the target space is narrowed, or the labor of always installing the error microphone in the target space is generated.

[0005] One aspect of this disclosure provides active noise control that does not require an error microphone. [Means for solving the problem]

[0006] To solve the above problems, a signal control device according to one aspect of the present disclosure includes: a first signal generation unit that generates a virtual feedback signal which virtually reproduces a signal corresponding to a feedback sound propagated from a speaker that generates a noise reduction sound for reducing sound pressure at a sound silencing position to a reference microphone located between a noise source and the speaker; a second signal generation unit that generates a virtual noise signal by subtracting the virtual feedback signal from a noise signal corresponding to a sound acquired by the reference microphone; and a noise reduction signal generation unit that generates a noise reduction signal for generating the noise reduction sound from the speaker by adaptive signal processing according to the noise from the noise source based on the noise signal and the virtual noise signal.

[0007] Furthermore, a noise reduction device according to one aspect of the present disclosure comprises the signal control device, the reference microphone, and the speaker.

[0008] Each aspect of the present disclosure may be implemented by a computer, in which case a signal control program for implementing the signal control device by a computer and a computer-readable recording medium on which the program is recorded also fall within the scope of the present disclosure.

[0009] Furthermore, a signal control method according to one aspect of the present disclosure includes the steps of: generating a virtual feedback signal that virtually reproduces a signal corresponding to a feedback sound propagated from a speaker that generates a noise reduction sound to reduce sound pressure at a sound silencing position to a reference microphone located between a noise source and the speaker; generating a virtual noise signal by subtracting the virtual feedback signal from a noise signal corresponding to a sound acquired by the reference microphone; and generating a noise reduction signal for generating the noise reduction sound from the speaker by adaptive signal processing according to the noise from the noise source based on the noise signal and the virtual noise signal. [Effects of the Invention]

[0010] According to one aspect of this disclosure, a noise reduction signal for generating a noise reduction sound having characteristics suitable for the characteristics of the noise can be generated without installing an error microphone at the sound elimination position. [Brief explanation of the drawing]

[0011] [Figure 1] This is a schematic diagram showing the noise reduction device, noise source, and sound silencing position during the control stage of the noise reduction device according to the embodiment. [Figure 2] This is a block diagram showing the noise reduction device, noise source, and sound reduction position during the control stage of the noise reduction device according to the embodiment. [Figure 3] This is a schematic diagram showing the noise reduction device, noise source, and sound elimination position during the tuning stage of the noise reduction device according to the embodiment. [Figure 4] This is a block diagram showing the noise reduction device, noise source, and sound reduction position during the tuning stage of the noise reduction device according to the embodiment. [Figure 5] This graph shows the change over time in the amount of sound pressure reduction at the sound elimination position by the noise reduction devices according to Example 1 and Example 2. [Modes for carrying out the invention]

[0012] [Embodiment] Hereinafter, one embodiment of the present disclosure will be described in detail with reference to the drawings. In this disclosure, the components shown in each drawing are for illustrative purposes only, and the actual shape and scale of the components are not limited to those shown in each drawing of this disclosure.

[0013] <Noise source and sound silencing location> In this embodiment, both the sound and the signal described later can be represented as a discrete-time signal a(n) with discrete time n as a variable. Generally, there is a correlation between the amplitude of a sound and the sound pressure of that sound. Therefore, reducing the amplitude of a sound, in other words, the mean square of the discrete-time signal representing the sound, is equivalent to reducing the sound pressure of that sound.

[0014] Figure 1 is a schematic diagram showing the noise reduction device 1 according to this embodiment. In this embodiment, an example of the control stage of the noise reduction device 1 will be described, in which the noise reduction device 1 reduces the sound pressure of noise x(n) generated from a noise source X as a primary sound source at the silencing position E. Figure 1 also shows the noise source X and the silencing position E.

[0015] In this embodiment, noise source X generates noise x(n). Noise x(n) is a discrete signal representation of the sound produced by noise source X. In this embodiment, noise x(n) may vary depending on n. Noise source X may be urban ambient noise, or noise generated by means of transportation such as automobiles.

[0016] In this embodiment, the sound-dampening position E is located in the propagation direction of at least a portion of the noise x(n) from the noise source X. For example, the sound-dampening position E may be located in a position surrounded by a wall EW having an opening EH on the side of the noise source X. For example, the sound-dampening position E may be located inside a house or inside a means of transportation such as an automobile.

[0017] For simplicity, in this embodiment, it is assumed that only noise generated from the noise source and noise reduction sounds (described later) used to reduce the sound pressure of said noise propagate to the silencing position E. In this embodiment, the influence of sounds other than the aforementioned noise and noise reduction sounds, such as voices that a person located at the silencing position E would like to hear, is ignored in the discussion.

[0018] <Overview of Vibration Reduction Device> In the present embodiment, as shown in FIG. 1, the noise reduction device 1 includes a signal control device 11, a reference microphone 12, and a speaker 13 as a secondary sound source. In particular, the noise reduction device 1 includes the reference microphone 12 on the side of the noise source X with respect to the speaker 13. In other words, the reference microphone 12 is located between the noise source X and the speaker 13.

[0019] Based on the sound acquired by the reference microphone 12, the noise reduction device 1 generates a noise reduction sound y(n) from the speaker 13. For example, ideally, the noise reduction sound y(n) generated from the speaker 13 has the same amplitude as the sound propagating to the sound cancellation position E including the noise x(n) and has an opposite phase. Thereby, the noise reduction device 1 reduces the amplitude of the sound at the sound cancellation position E by the noise reduction sound y(n) from the speaker 13, and thus reduces the sound pressure at the sound cancellation position E.

[0020] Based on the sound acquired by the reference microphone 12, the signal control device 11 generates a noise signal u(n) described later. The noise signal u(n) is a discrete-time signal representing the sound acquired by the reference microphone 12 at the discrete time n.

[0021] Also, based on the noise signal u(n), the signal control device 11 generates a noise reduction signal w(n) described later, inputs the noise reduction signal w(n) to the speaker 13, and causes the speaker 13 to generate a noise reduction sound y(n). The noise reduction sound y(n) corresponds to the signal representing the noise reduction signal w(n) input to the speaker 13 at the discrete time n.

[0022] <Sound propagation path> The principle of reducing the sound pressure at the sound cancellation position E by the noise reduction device 1 will be described in more detail with reference to FIG. 2. FIG. 2 is a block diagram showing the noise reduction device 1 in the control stage together with the noise source X and the sound cancellation position E.

[0023] As described above, noise source X generates noise x(n). Noise x(n) from noise source X changes into noise dv(n) by the time it propagates to the silenced position E. Noise dv(n) can be determined from noise x(n) using the first transfer function T1, which corresponds to the propagation path from noise source X to silenced position E.

[0024] In this embodiment, both the transfer function and the filter function described later are expressed as functions corresponding to the transfer function and filter function, with z as the variable. For example, suppose a sound corresponding to a discrete-time signal a(n) propagates along a predetermined propagation path. In this case, the discrete-time signal becomes a'(n), obtained by inverse Z-transforming the product of a function A(z) obtained by Z-transforming the discrete-time signal a(n) and the transfer function corresponding to the propagation path. Also, for example, suppose a signal represented by a discrete-time signal b(n) is processed by a predetermined filter. In this case, the processed signal becomes a discrete-time signal b'(n), obtained by inverse Z-transforming the product of a function B(z) obtained by Z-transforming the discrete-time signal b(n) and the filter function corresponding to the filter.

[0025] As described above, in this embodiment, the sound propagated along the propagation path and the signal processed by the filter are obtained by performing a Z-transform on the original sound or signal, and then performing an inverse Z-transform on the discrete-time signal. By performing calculations in this manner, calculations on various signals can be easily carried out by digital processing, for example, by a computer. However, this embodiment is not limited to this, and calculations on each signal, including signals representing various sounds, may be performed by directly performing numerical processing on each value of the signal. For example, calculations on each signal may be performed in the time domain. In this embodiment, the signal output by processing the signal with a filter may be calculated, for example, by a convolution operation between the filter coefficients of the filter and the signal.

[0026] If the first transfer function T1 is the function P(z), then the function dv(z) obtained by Z-transforming the noise dv(n) can be expressed as dv(z) = x(z)P(z), where x(z) is the function obtained by Z-transforming the noise x(n).

[0027] In this embodiment, noise dv(n) is described as sound propagating from noise source X through opening EH to sound silencing position E, but this is not limited to this, and the propagation path of noise x(n) from noise source X to sound silencing position E is not particularly limited. For example, in this embodiment, noise x(n) from noise source X may propagate as noise dv(n) through wall EW to sound silencing position E located in the space separated by wall EW.

[0028] Furthermore, noise x(n) from noise source X changes into noise r(n) by the time it propagates to the reference microphone 12. Let the transfer function corresponding to the propagation path from noise source X to reference microphone 12 be the second transfer function T2, and let the second transfer function T2 be represented by the function R(z). In this case, the function r(z) obtained by Z-transforming noise r(n) can be expressed as r(z) = x(z)R(z) using the function x(z).

[0029] On the other hand, speaker 13, which receives the noise reduction signal w(n) output by signal control device 11 using the method described later, generates a noise reduction sound y(n) as described above. The noise reduction sound y(n) from speaker 13 changes into a noise reduction sound yv(n) by the time it propagates to the silenced position E. Let the transfer function corresponding to the propagation path from speaker 13 to silenced position E be the third transfer function T3, and let the third transfer function T3 be represented by the function S(z). In this case, the function yv(z) obtained by Z-transforming the noise reduction sound yv(n) is expressed as yv(z)=y(z)S(z) using the function y(z).

[0030] Let the sound at the silencing position E be the error sound e(n), which is a combination of the noise dv(n) and the noise reduction sound yv(n). As described above, if we ignore sounds other than the sound from noise source X and the sound from speaker 13, the error sound e(n) is a discrete signal representation of the sound that is a combination of the sound from noise source X and the sound from speaker 13.

[0031] In this case, the function e(z) obtained by Z-transforming the error sound e(n) is expressed as e(z)=dv(z)+yv(z)=x(z)P(z)+y(z)S(z). The noise reduction device 1 reduces the sound pressure at the silenced position E by generating a noise reduction sound y(n) from the speaker 13 such that the error sound e(n) approaches 0 as closely as possible, regardless of n. Ideally, the speaker 13 generates the noise reduction sound y(n) such that the noise reduction sound yv(n) has the same amplitude and opposite phase as the noise dv(n).

[0032] As shown in Figure 1, the noise reduction sound y(n) from speaker 13 propagates through the aperture EH to the reference microphone 12. The sound propagating from speaker 13 to reference microphone 12 is denoted as the feedback sound yf(n). The transfer function corresponding to the propagation path from speaker 13 to reference microphone 12 is denoted as the fourth transfer function T4, and the fourth transfer function T4 is represented by the function F(z). In this case, the function yf(z) obtained by Z-transforming the feedback sound yf(n) is expressed as yf(z)=y(z)F(z) using the function y(z).

[0033] In this embodiment, the feedback sound yf(n) is described as sound propagated from speaker 13 through aperture EH to reference microphone 12, but is not limited to this. In particular, in this embodiment, the propagation path of the feedback sound yf(n) is not particularly limited, as long as the noise reduction sound y(n) generated from speaker 13 propagates to reference microphone 12 and is acquired as the feedback sound yf(n). For example, in this embodiment, the noise reduction sound y(n) from speaker 13 located in a space separated by wall EW may propagate through the wall EW to reference microphone 12 and be acquired as the feedback sound yf(n).

[0034] <Signal control device: Overview> Referring to Figure 2, the signal control device 11 comprises a first signal generation unit 31, a second signal generation unit 32, a first adaptive filter 41, and a first update unit 61. In particular, the signal control device 11 includes the first adaptive filter 41 and the first update unit 61 as a noise reduction signal generation unit that generates a noise reduction signal w(n). The signal control device 11 may also include an acquisition unit 21 and an output unit 51.

[0035] The acquisition unit 21 acquires a noise signal u(n) corresponding to the sound acquired by the reference microphone 12. The sound acquired by the reference microphone 12 is a composite sound of noise r(n) propagated from the noise source X and feedback sound yf(n) propagated from the speaker 13. The noise signal u(n) corresponds to a signal that represents the sound acquired by the reference microphone 12 at discrete time n as a discrete signal. Therefore, the function u(z) obtained by Z-transforming the noise signal u(n) can be expressed as u(z) = r(z) + yf(z) using the functions r(z) and yf(z).

[0036] The first signal generation unit 31 generates a virtual feedback signal y^f(n) which virtually reproduces the signal corresponding to the feedback sound yf(n) propagating from the speaker 13 to the reference microphone 12. For example, the first signal generation unit 31 includes a virtual filter which virtually reproduces the fourth transfer function T4, which is the transfer function from the speaker 13 to the reference microphone 12.

[0037] In this disclosure, letters with the symbol "^" (hat) attached should ideally be represented by placing the symbol "^" directly above the letter. However, for the sake of presentation in this specification, letters with the symbol "^" attached will be represented by placing the symbol "^" after the letter.

[0038] The filter function corresponding to the virtual filter is represented by the function F^(z). Function F^(z) is a function that virtually reproduces the function F(z). The first signal generation unit 31 generates a virtual feedback signal y^f(n) by processing the noise reduction signal w(n) with the virtual filter. The function y^f(z) obtained by Z-transforming the virtual feedback signal y^f(n) is expressed as y^f(z)=w(z)F^(z) using the functions w(z) and F^(z).

[0039] The second signal generation unit 32 includes an adder that combines the noise signal u(n) and a signal with the opposite phase to the virtual feedback signal y^f(n) to generate a virtual noise signal r^(n). In other words, the second signal generation unit 32 generates the virtual noise signal r^(n) by subtracting the virtual feedback signal y^f(n) from the noise signal u(n). Therefore, the function r^(z) obtained by Z-transforming the virtual noise signal r^(n) can be expressed as r^(z) = u(z) - y^f(z) using the functions u(z) and y^f(z).

[0040] The signal obtained by subtracting the signal corresponding to the feedback sound yf(n) from the noise signal u(n) corresponds to the signal corresponding to the noise component r(n) of the sound acquired by the reference microphone 12. Therefore, the virtual noise signal r^(n) corresponds to a signal that virtually reproduces the signal corresponding to the noise component r(n) of the noise signal u(n).

[0041] The first adaptive filter 41 processes the virtual noise signal r^(n) to generate a noise reduction signal w(n). For example, the first adaptive filter may be an FIR (Finite Impulse Response) filter, and the filter coefficients of the first adaptive filter are determined to process the virtual noise signal r^(n) to generate a noise reduction signal w(n). In particular, the determination of the filter coefficients of the first adaptive filter 41 is determined by the first update unit 61, which will be described later.

[0042] The filter function corresponding to the first adaptive filter 41 is represented by the function W(z). Therefore, the function w(z) can be expressed using the functions y^f(z) and W(z) as w(z) = y^f(z)W(z).

[0043] The output unit 51 outputs a noise reduction signal w(n) to the speaker 13, which generates a noise reduction sound y(n) from the speaker 13. In particular, the output unit 51 generates a noise reduction sound y(n) from the speaker 13 that corresponds to the noise reduction signal w(n) at discrete time n.

[0044] <Signal control device: First update unit> The first update unit 61 updates the parameters of the first adaptive filter 41 based on the noise signal u(n) and the virtual noise signal r^(n). In particular, the first update unit 61 sequentially updates the parameters of the first adaptive filter so that the error sound e(n) of the sound at the silenced position E approaches 0 as close as possible to 0, regardless of n, by a method described later. For example, in this embodiment, the first update unit 61 updates the filter coefficients of the first adaptive filter 41, which is an FIR filter, by updating the filter coefficients of the first adaptive filter 41 based on the noise signal u(n) and the virtual noise signal r^(n).

[0045] The first update unit 61 includes a comparison filter 71, an internal reference signal generation unit 72, and a coefficient update unit 73. The internal reference signal generation unit 72 includes a second adaptive filter 42 and a signal synthesis unit 74.

[0046] The comparison filter 71 has filter coefficients corresponding to the transfer function of one of the sound propagation paths described above. In particular, during the control phase, the comparison filter 71 functions as a first comparison filter having filter coefficients that virtually reproduce the fourth transfer function T4, which is the transfer function from speaker 13 to reference microphone 12. During the control phase, the comparison filter 71 may have the same filter coefficients as the virtual filter of the first signal generation unit 31. In other words, during the control phase, the filter function corresponding to the comparison filter 71 may be the function F^(z).

[0047] The comparison filter 71 receives a virtual noise signal r^(n) from the second signal generation unit 32. The comparison filter 71 processes the virtual noise signal r^(n) to generate the first comparison signal r^1(n). Therefore, the function r^1(z) obtained by Z-transforming the first comparison signal r^1(n) can be expressed as r^1(z) = r^(z)F^(z) using the functions r^(z) and F^(z).

[0048] The second adaptive filter 42 of the internal reference signal generation unit 72 processes the virtual noise signal r^(n) to generate the intermediate signal yh(n). For example, the second adaptive filter 42 may be an FIR (Finite Impulse Response) filter, and the filter coefficients of the second adaptive filter 42 are determined to process the virtual noise signal r^(n) to generate the intermediate signal yh(n). The specific method for determining the filter coefficients of the second adaptive filter 42 will be described in detail later.

[0049] For example, suppose the filter function corresponding to the second adaptive filter 42 is function H(z). In this case, the function yh(z) obtained by Z-transforming the intermediate signal yh(n) can be expressed as yh(z)=r^(z)H(z) using the functions r^(z) and H(z).

[0050] The signal synthesis unit 74 of the internal reference signal generation unit 72 includes an adder that, upon receiving the noise signal u(n) and the intermediate signal yh(n), synthesizes the noise signal u(n) and the intermediate signal yh(n) to generate the internal reference signal eh(n). Therefore, the function eh(z) obtained by Z-transforming the internal reference signal eh(n) can be expressed as eh(z) = u(z) + yh(z) using the functions u(z) and yh(z). Thus, the internal reference signal generation unit 72 generates the internal reference signal eh(n) based on the noise signal u(n) and the virtual noise signal r^(n).

[0051] The coefficient update unit 73 receives the first comparison signal r^1(n) and the internal reference signal eh(n) as inputs. In other words, the coefficient update unit 73 receives the first comparison signal r^1(n), which is obtained by adapting the filter function F^(z) corresponding to the fourth transfer function T4 to the virtual noise signal r^(n) input to the first adaptive filter 41. The coefficient update unit 73 also receives the internal reference signal eh(n).

[0052] As described above, the filter coefficients of the second adaptive filter 42 of the internal reference signal generation unit 72 are configured to generate an intermediate signal yh(n) through processing of the virtual noise signal r^(n), which, when combined with the noise signal u(n), generates the internal reference signal eh(n). As will be described later, the filter coefficients of the second adaptive filter 42 are optimized so that the mean square of the internal reference signal eh(n) is minimized when the filter coefficients of the first adaptive filter 41 are updated so that the sound pressure at the silencing position E is minimized. In other words, the internal reference signal eh(n) is the signal whose mean square is minimized when the sound pressure at the silencing position E is minimized, that is, when the mean square of the signal obtained by combining the noise dv(n) and the noise reduction sound yv(n) is minimized.

[0053] Therefore, the coefficient update unit 73 can optimize the filter coefficients of the first adaptive filter 41 using an adaptive algorithm, taking the first comparison signal r^1(n) as input and the internal reference signal eh(n) as the error. In other words, the coefficient update unit 73 can optimize the filter coefficients of the first adaptive filter 41 by updating the filter coefficients of the first adaptive filter 41 based on the first comparison signal r^1(n) and the internal reference signal eh(n).

[0054] In particular, the coefficient update unit 73 updates the filter coefficients of the first adaptive filter 41 using an adaptive algorithm, for example, so that the mean square of the internal reference signal eh(n) approaches zero. In other words, the coefficient update unit 73 updates the filter coefficients of the first adaptive filter 41 using an adaptive algorithm so that the amplitude of the internal reference signal eh(n) approaches zero. The updating of the filter coefficients of the first adaptive filter 41 by the coefficient update unit 73 may be performed using, for example, the NLMS (Normalized Least Mean Squares) algorithm.

[0055] However, the method for updating the filter coefficients of the first adaptive filter 41 by the coefficient update unit 73 is not limited to this, as long as it is a method that reduces the amplitude of the internal reference signal eh(n), such as a method that reduces the squared value of the internal reference signal eh(n). For example, the algorithm for updating the filter coefficients of the first adaptive filter 41 by the coefficient update unit 73 may be an LMS (least mean squares) algorithm, an RLS (recursive least squares) algorithm, or a steepest descent method, etc.

[0056] Therefore, by determining the filter coefficients, which are parameters of the first adaptive filter 41, so that the mean square of the internal reference signal eh(n) approaches 0, the noise reduction device 1 can reduce the error sound e(n) at the silencing position E.

[0057] As mentioned above, the filter coefficients, which are parameters of the first adaptive filter 41, are sequentially updated by the first update unit 61. Therefore, the noise reduction device 1 can sequentially reduce the error sound e(n) at the silencing position E even when there is a change in the noise x(n) from the noise source X. Thus, the noise reduction device 1 can reduce the sound pressure at the silencing position E regardless of the time-dependent change in the noise x(n) from the noise source X.

[0058] As a result, the signal control device 11 can update and optimize the filter coefficients of the first adaptive filter 41 using the adaptive algorithm described above, via the first update unit 61. Therefore, the signal control device 11 can realize the noise reduction signal w(n) for the noise reduction device 1 to generate a noise reduction sound y(n) having characteristics suitable for the noise x(n) characteristics, with a simpler configuration.

[0059] In this embodiment, the signal control device 11, based on the noise signal u(n) and the virtual noise signal r^(n), sequentially optimizes the parameters of the first adaptive filter 41 by adaptive signal processing in accordance with the time-dependent change in noise x(n) from the noise source X, using the first update unit 61. In addition, the signal control device 11 generates a noise reduction signal w(n) by processing the virtual noise signal r^(n) with the first adaptive filter 41, whose parameters are sequentially optimized by adaptive signal processing.

[0060] Therefore, the noise reduction signal generation unit of the signal control device 11 generates a noise reduction signal w(n) by adaptive signal processing corresponding to the noise x(n) from the noise source X, based on the noise signal u(n) and the virtual noise signal r^(n). In other words, the signal control device 11 implements a signal control method that generates a noise reduction signal w(n) by adaptive signal processing corresponding to the noise x(n) from the noise source X, based on the noise signal u(n) and the virtual noise signal r^(n).

[0061] <Virtual microphone> As described above, the parameters of the second adaptive filter 42, including the filter coefficients, are predetermined such that the intermediate signal yh(n) generated by the second adaptive filter 42 is combined with the noise signal u(n) to form the internal reference signal eh(n). The tuning stage of the second adaptive filter 42, which determines the parameters including the filter coefficients of the second adaptive filter 42, will be explained in detail with reference to Figures 3 and 4.

[0062] Figure 3 is a schematic diagram showing the noise reduction device 1 according to this embodiment during the tuning stage, along with the noise source X and the sound silencing position E. Figure 4 is a block diagram showing the noise reduction device 1 according to this embodiment during the tuning stage, along with the noise source X and the sound silencing position E.

[0063] During the tuning phase, a virtual microphone VM is temporarily installed inside the sound-canceling position E, for example, inside the wall section EW. The virtual microphone VM acquires sound at the sound-canceling position E. The virtual microphone VM is removed from the sound-canceling position E during the control phase.

[0064] Since the virtual microphone VM is installed at the silenced position E, the transfer function corresponding to the propagation path from the noise source X to the virtual microphone VM can be considered identical to the first transfer function T1 described above. Also, the transfer function corresponding to the propagation path from speaker 13 to the virtual microphone VM can be considered identical to the third transfer function T3 described above. Therefore, the virtual microphone VM acquires noise dv(n) and noise-reduced sound yv(n) as sound at the silenced position E.

[0065] Ideally, the out-of-phase sounds of noise dv(n) and noise reduction sound yv(n) are identical, so at the silenced position E, noise dv(n) and noise reduction sound yv(n) cancel each other out. Therefore, ideally, the amplitude of the sound acquired by the virtual microphone VM at the silenced position E is 0, in other words, the sound pressure at the silenced position E is 0. On the other hand, if there is a difference between the out-of-phase sounds of noise dv(n) and noise reduction sound yv(n), the virtual microphone VM acquires an error sound e(n) at the silenced position E, which is a combination of noise dv(n) and noise reduction sound yv(n).

[0066] <Adjusting the second adaptive filter> During the tuning phase, the signal control device 11 includes an error signal acquisition unit 81 and a second update unit 62. During the control phase, in other words, when the noise reduction device 1 is used to reduce the sound pressure at the silencing position E, the signal control device 11 does not necessarily need to include the error signal acquisition unit 81 and the second update unit 62.

[0067] The error signal acquisition unit 81 acquires an error signal ev(n) corresponding to the error sound e(n) acquired by the virtual microphone VM. The error signal ev(n) corresponds to the error sound e(n) acquired by the virtual microphone at discrete time n. Therefore, the relationship e(z) = ev(z) holds between the function e(z) obtained by Z-transforming the error sound e(n) and the function ev(z) obtained by Z-transforming the error signal ev(n).

[0068] The second update unit 62 updates the filter coefficients, which are parameters of the second adaptive filter 42. Specifically, the second update unit 62 receives a virtual noise signal r^(n) and an internal reference signal eh(n) as inputs. The second update unit 62 takes the virtual noise signal r^(n) as input and the internal reference signal eh(n) as an error, and optimizes the filter coefficients of the second adaptive filter 42 using an adaptive algorithm. In other words, the second update unit 62 optimizes the filter coefficients of the second adaptive filter 42 by updating them based on the virtual noise signal r^(n) and the internal reference signal eh(n).

[0069] In particular, the second update unit 62 updates the filter coefficients of the second adaptive filter 42 based on the virtual noise signal r^(n) and the internal reference signal eh(n) so as to minimize the mean square of the internal reference signal eh(n). The algorithm for updating the filter coefficients of the second adaptive filter 42 by the second update unit 62 may be the same as the algorithm for updating the filter coefficients of the first adaptive filter 41 by the coefficient update unit 73 described above.

[0070] Furthermore, during the tuning phase, the comparison filter 71 functions as a second comparison filter having filter coefficients that virtually reproduce the third transfer function T3, which is the transfer function from speaker 13 to virtual microphone VM. During the tuning phase, the filter function corresponding to the comparison filter 71 may be the function S^(z). The filter function corresponding to the comparison filter 71 may be changed as appropriate by appropriately changing the filter coefficients of the comparison filter 71 between the tuning phase and the control phase.

[0071] Even during the tuning stage, the comparison filter 71 receives a virtual noise signal r^(n) from the second signal generation unit 32. The comparison filter 71 processes the virtual noise signal r^(n) to generate a second comparison signal r^2(n). Therefore, the function r^2(z) obtained by Z-transforming the second comparison signal r^2(n) can be expressed as r^2(z)=r^(z)S^(z) using the functions r^(z) and S^(z).

[0072] During the tuning phase, the coefficient update unit 73 receives the second comparison signal r^2(n) instead of the first comparison signal r^1(n). Also during the tuning phase, the coefficient update unit 73 receives the error signal ev(n) acquired by the error signal acquisition unit 81 instead of the internal reference signal eh(n).

[0073] In other words, during the tuning phase, the coefficient update unit 73 receives a second comparison signal r^2(n), which is obtained by adapting the filter function S^(z) corresponding to the third transfer function T3 to the virtual noise signal r^(n) input to the first adaptive filter 41. Also during the tuning phase, the coefficient update unit 73 receives an error signal ev(n), which corresponds to the error between the noise dv(n) and the noise reduction sound yv(n) at the silencing position E.

[0074] Therefore, the coefficient update unit 73 can optimize the filter coefficients of the first adaptive filter 41 using an adaptive algorithm, taking the second comparison signal r^2(n) as input and the error signal ev(n) as the error. In other words, the coefficient update unit 73 can optimize the filter coefficients of the first adaptive filter 41 by updating the filter coefficients of the first adaptive filter 41 based on the second comparison signal r^2(n) and the error signal ev(n).

[0075] In this embodiment, during the tuning stage, the filter function corresponding to the first adaptive filter is updated to the function Wt(z) by the first update unit 61. Therefore, during the tuning stage, the function w(z) is replaced with w(z)=y^f(z)Wt(z) using the functions y^f(z) and Wt(z).

[0076] Except as stated above, the updating of the filter coefficient of the first adaptive filter by the first update unit 61 during the tuning phase may be performed in the same manner as the updating of the filter coefficient of the first adaptive filter 41 by the first update unit 61 during the control phase. Also, except as stated above, during the tuning phase, each part of the noise reduction device 1 performs the same operation as it did during the control phase.

[0077] The signal control device 11 generates an internal reference signal eh(n) using an internal reference signal generation unit 72 that includes a second adaptive filter 42. This allows the signal control device 11 to generate an internal reference signal eh(n) with a more accurate value. In particular, the signal control device 11 updates the filter coefficients of the first adaptive filter 41 using a first update unit 61 based on the actually acquired error signal ev(n), while simultaneously determining the filter coefficients of the second adaptive filter 42 using a second update unit 62. This allows the signal control device 11 to generate an internal reference signal eh(n) with an even more accurate value.

[0078] The signal control device 11 can determine the filter coefficients of the second adaptive filter 42 using the adaptive algorithm described above, via the second update unit 62. Therefore, the signal control device 11 can determine the appropriate filter coefficients of the second adaptive filter 42 with a simpler configuration.

[0079] Furthermore, during the tuning phase, the signal control device 11 can update and optimize the filter coefficients of the first adaptive filter 41 using the first update unit 61, which includes a comparison filter 71 that functions as a second comparison filter. As a result, during the tuning phase, the signal control device 11 can determine the filter coefficients of the second adaptive filter 42 with a simpler configuration and with greater accuracy.

[0080] <Considerations of the first adaptive filter during the tuning phase> Through an examination of the error signal ev(n) during the tuning phase, we will consider the filter function Wt(z) of the first adaptive filter 41. The error signal ev(n) corresponds to the error sound e(n) acquired by the virtual microphone VM, and the error sound e(n) is a sound synthesized from noise dv(n) and noise reduction sound yv(n). Therefore, the function ev(z) obtained by Z-transforming the error signal ev(n) is expressed as shown in equation (1) below.

[0081]

number

[0082] Here, since F^(z) is a filter function that virtually reproduces the fourth transfer function T4, F(z), ideally F^(z) = F(z) holds. In this case, equation (1) above can be replaced with equation (2) below.

[0083]

number

[0084] During the tuning phase, the filter coefficients of the filter function Wt(z) of the first adaptive filter 41 are updated so that the mean square of the error signal ev(n) is minimized, ideally to zero. In equation (2) above, for the mean square of the error signal ev(n) to be zero regardless of the noise x(n), the expression inside the curly braces on the right-hand side of equation (2) must be zero. Therefore, the filter function Wt(z) of the first adaptive filter 41 optimized during the tuning phase satisfies equation (3) below.

[0085]

number

[0086] In other words, the first update unit 61 updates the filter coefficients of the first adaptive filter 41 during the tuning phase so that the filter function Wt(z) satisfies equation (3) above.

[0087] <Considerations of the second adaptive filter during the tuning phase> Next, we will consider the filter function H(z) of the second adaptive filter 42 through our examination of the internal reference signal eh(n) during the tuning stage. During the tuning stage, the internal reference signal eh(n) is a signal obtained by combining the noise signal u(n) and the intermediate signal yh(n). Therefore, assuming that F^(z)=F(z) holds as described above, the function eh(z) obtained by Z-transforming the internal reference signal eh(n) is expressed as shown in equation (4) below.

[0088]

number

[0089] During the tuning phase, the filter coefficients of the filter function H(z) of the second adaptive filter 42 are updated so that the mean square of the internal reference signal eh(n) is minimized, ideally to zero. In equation (4) above, for the mean square of the internal reference signal eh(n) to be zero regardless of the noise x(n), the expression in the curly braces on the right side of equation (4) must be zero. Therefore, the optimized filter function H(z) of the second adaptive filter 42 satisfies equation (5) below.

[0090]

number

[0091] In other words, the second update unit 62 updates the filter coefficients of the second adaptive filter 42 during the tuning stage so that the filter function H(z) satisfies equation (5) above, thereby determining the filter coefficients.

[0092] By determining the filter function H(z) in this way, the second adaptive filter 42 is obtained by processing the input virtual noise signal r^(n) to obtain an intermediate signal yh(n) from which an internal reference signal eh(n) is obtained by combining it with the noise signal u(n). In other words, the filter coefficients of the second adaptive filter 42 are determined such that the mean square of the internal reference signal eh(n) is minimized when the filter coefficients of the first adaptive filter 41 are updated during the tuning stage so that the sound pressure at the silenced position E is minimized.

[0093] <Consideration of the first adaptive filter in the control phase> Returning to Figures 1 and 2, let us consider the filter function W(z) of the first adaptive filter 41 during the control phase. As mentioned above, if F^(z)=F(z) holds, then during the control phase, the function eh(z) obtained by Z-transforming the internal reference signal eh(n) is expressed as shown in equation (6) below.

[0094]

number

[0095] During the control phase, the first update unit 61 updates the filter coefficients of the filter function W(z) of the first adaptive filter 41 so that the mean square of the internal reference signal eh(n) is minimized, ideally to zero. In equation (6) above, for the mean square of the internal reference signal eh(n) to be zero regardless of the noise x(n), the expression inside the curly braces on the right side of equation (6) must be zero. Therefore, the filter function W(z) of the first adaptive filter 41 optimized during the control phase satisfies (7) below.

[0096]

number

[0097] As described above, the filter function H(z) of the second adaptive filter 42 is determined to satisfy equation (5) above. Therefore, by substituting equation (5) above into equation (7) above, equation (8) below is obtained.

[0098]

number

[0099] In other words, the filter function W(z) of the first adaptive filter 41 optimized during the control phase is the same as the filter function Wt(z) of the first adaptive filter 41 optimized during the tuning phase. As described above, during the tuning phase, the filter coefficients of the first adaptive filter 41 are updated so that the mean square of the error signal ev(n) is minimized, in other words, so that the sound pressure at mute position E is minimized. Also, during the control phase, the filter coefficients of the first adaptive filter 41 are updated so that the mean square of the internal reference signal eh(n) is minimized. Therefore, from equation (8) above, it is clear that the filter coefficients of the second adaptive filter 42 are determined so that the sound pressure at mute position E is minimized when the mean square of the internal reference signal eh(n) is minimized during the control phase.

[0100] Therefore, by updating the filter coefficients of the first adaptive filter 41 as described above during the control phase, the first adaptive filter 41 can process the virtual noise signal r^(n) and generate a noise reduction signal w(n) that minimizes the mean square of the error signal ev(n). Thus, the noise reduction device 1 equipped with the signal control device 11 that has undergone the tuning phase described above can minimize the mean square of the error signal ev(n) corresponding to the sound at the silencing position E during the control phase.

[0101] As described above, the signal control device 11 can generate a noise reduction signal w(n) to reduce the sound pressure at the silenced position E without requiring an error microphone that is always installed on the side of the silenced position E that is closer to the speaker 13. In other words, the signal control device 11 can realize active noise control using the first adaptive filter 41 without requiring an error microphone.

[0102] Therefore, the signal control device 11 can generate a noise reduction signal w(n) for generating a noise reduction sound y(n) according to the characteristics of the noise x(n), while reducing the effort required to install a sound collection device at the noise reduction position E, or reducing the narrowing of the space near the noise reduction position E. In other words, the above-described signal control method realizes the generation of a noise reduction signal w(n) for generating a noise reduction sound y(n) according to the characteristics of the noise x(n), while reducing the effort required to install a sound collection device at the noise reduction position E, or reducing the narrowing of the space near the noise reduction position E. The noise reduction device 1 equipped with the signal control device 11 can reduce the sound pressure at the noise reduction position E regardless of the characteristics of the noise x(n), without installing a sound collection device at the noise reduction position E.

[0103] In particular, the first signal generation unit 31 includes a virtual filter represented by a filter function F^(z) that virtually reproduces the transfer function F(z) from the speaker 13 to the reference microphone 12. As described above, the validity of equation (8) is premised on the fact that F(z)=F^(z), and bringing F^(z) closer to F(z) contributes to further reduction of sound pressure at the silenced position E. For this reason, by providing the first signal generation unit 31 including the virtual filter described above, the signal control device 11 can improve the accuracy of reproducing the virtual feedback signal y^f(n) corresponding to the feedback sound yf(n), or simplify the configuration necessary for such reproduction.

[0104] As described above, the virtual noise signal r^(n) is the signal obtained by subtracting the virtual feedback signal y^f(n) from the noise signal u(n). The signal control device 11 generates a noise reduction signal w(n) by processing the virtual noise signal r^(n) with the first adaptive filter 41. Therefore, the signal control device 11 improves the accuracy of generating a suitable noise reduction signal w(n) or simplifies the configuration required for generating the virtual feedback signal y^f(n) by improving the accuracy of reproducing the virtual feedback signal y^f(n) or by simplifying the configuration required for such reproduction.

[0105] In this embodiment, the signal control device 11 generates a noise reduction signal w(n) while updating the filter coefficients of the first adaptive filter 41 using the feedback sound transmitted from the speaker 13 to the reference microphone 12. Therefore, the larger the feedback sound, or in other words, the larger the transfer function F(z), the more accurately the signal control device 11 can generate the noise reduction signal w(n) for reducing the sound pressure at the silenced position E.

[0106] On the other hand, the signal control device 11 according to this embodiment may include an amplifier for amplifying the noise signal u(n) and an amplifier for amplifying the virtual noise signal r^(n). In this case, the signal control device 11 can generate a noise reduction signal w(n) with reduced accuracy even when the feedback sound is small.

[0107] <Examples of implementation using software> The function of the signal control device 11 (hereinafter referred to as "the device") is a program that causes the device to function as a computer, and can be realized by a signal control program that causes the computer to function as each control block of the device. In particular, the signal control program causes the computer to function as the first signal generation unit 31, the second signal generation unit 32, the first adaptive filter 41, and the first update unit 61.

[0108] In this case, the device includes a computer having, as hardware for executing the program, at least one control device such as a processor and at least one storage device such as memory. By executing the program using this control device and storage device, each of the functions described in each of the embodiments is realized.

[0109] The above program may be recorded on one or more computer-readable recording media, not temporary ones. These recording media may or may not be provided by the above device. In the latter case, the program may be supplied to the above device via any wired or wireless transmission medium.

[0110] Furthermore, some or all of the functions of each of the above control blocks can also be implemented by logic circuits. For example, an integrated circuit in which logic circuits functioning as each of the above control blocks are formed is also included in the scope of this disclosure. In addition, it is also possible to implement the functions of each of the above control blocks by, for example, a quantum computer.

[0111] Furthermore, each process described in the above embodiments may be performed by AI (Artificial Intelligence). In this case, the AI ​​may operate on the control device described above, or it may operate on other devices (for example, an edge computer or a cloud server).

[0112] <Actual measurement of noise reduction amount> Using noise reduction devices according to Examples 1 and 2, which have the same configuration as noise reduction device 1 according to this embodiment, we actually verified that the sound pressure of noise from noise source X is reduced at the sound reduction position E.

[0113] In the verification using the noise reduction device according to each embodiment, a noise source X was placed inside a soundproof room. The noise reduction device according to each embodiment is installed inside the soundproof room and includes a reference microphone 12 located 100 cm away from the noise source X, along the propagation direction of the noise X(n) from the noise source X. The noise reduction device according to each embodiment also includes a speaker 13 located inside the soundproof room, 11 cm away from the reference microphone 12, along the propagation direction of the noise X(n) from the noise source X.

[0114] The determination of the filter coefficient of the second adaptive filter 42 of the signal control device 11 in each embodiment of the noise reduction device was performed using the same method as the determination of the filter coefficient of the second adaptive filter 42 in the noise reduction device 1 according to this embodiment. In the tuning stage according to Embodiment 1, the virtual microphone VM was placed in the soundproof room at a position 10 cm away from the speaker 13, along the propagation direction of noise X(n) from the noise source X. In the tuning stage according to Embodiment 2, the virtual microphone VM was placed in the same soundproof room at a position 20 cm away from the speaker 13, along the propagation direction of noise X(n) from the noise source X. The sound silencing position E in each embodiment was the same as the installation position of the virtual microphone VM.

[0115] The noise reduction devices according to each embodiment were operated in the same manner as described above for the operation method of the noise reduction device 1 when reducing the sound pressure at the sound reduction position E using the noise reduction device 1 according to this embodiment, and the sound pressure at the sound reduction position E was measured.

[0116] Figure 5 is a graph showing the change over time in the amount of sound pressure reduction at the sound silencing position E by the noise reduction devices according to Example 1 and Example 2. Graph G1 shows the amount of sound pressure reduction at the sound silencing position E in Example 1, and graph G2 shows the amount of sound pressure reduction at the sound silencing position E in Example 2. For each graph in Figure 5, the horizontal axis represents the elapsed time (in s) from the start of operation of the noise reduction device 1, and the vertical axis represents the amount of sound pressure reduction at the sound silencing position E relative to the sound pressure at the noise source X (in dB).

[0117] As shown in the graphs in Figure 5, it is clear that the sound pressure at the silencing position E is reduced to that at the noise source X by the noise reduction device according to each embodiment. Furthermore, the noise reduction device according to each embodiment tends to reduce the sound pressure at the silencing position E more strongly as the elapsed time from the start of operation increases. This indicates that the noise reduction device according to each embodiment generates a noise reduction sound with suitable characteristics by appropriately updating, for example, the filter coefficient of the first adaptive filter 41 of the signal control device 11 in order to reduce the sound pressure at the silencing position E.

[0118] <Summary> A signal control device according to Embodiment 1 of the present disclosure comprises: a first signal generation unit that generates a virtual feedback signal which virtually reproduces a signal corresponding to a feedback sound propagated from a speaker that generates a noise reduction sound for reducing sound pressure at a sound silencing position to a reference microphone located between a noise source and the speaker; a second signal generation unit that generates a virtual noise signal by subtracting the virtual feedback signal from a noise signal corresponding to a sound acquired by the reference microphone; and a noise reduction signal generation unit that generates a noise reduction signal for generating the noise reduction sound from the speaker by adaptive signal processing according to the noise from the noise source based on the noise signal and the virtual noise signal.

[0119] The signal control device according to aspect 2 of the present disclosure may be configured such that, in aspect 1 above, the noise reduction signal generation unit includes a first adaptive filter that processes the virtual noise signal to generate the noise reduction signal, and a first update unit that updates the parameters of the first adaptive filter based on the noise signal and the virtual noise signal.

[0120] In the third aspect of the present disclosure, the signal control device may be configured such that, in the second aspect described above, the first signal generation unit includes a virtual filter that virtually reproduces a transfer function corresponding to the propagation path from the speaker to the reference microphone.

[0121] The signal control device according to aspect 4 of the present disclosure may be configured such that, in aspect 2 or 3 above, the first adaptive filter is an FIR filter, and the first update unit updates the filter coefficients of the first adaptive filter based on the noise signal and the virtual noise signal.

[0122] The signal control device according to aspect 5 of the present disclosure may be configured in aspect 4 above, to include: a first update unit that virtually reproduces a transfer function corresponding to the propagation path from the speaker to the reference microphone and processes the virtual noise signal to generate a first comparison signal; an internal reference signal generation unit that generates an internal reference signal whose mean square is minimized when the sound pressure at the silenced position is minimized based on the noise signal and the virtual noise signal; and a coefficient update unit that updates the filter coefficients of the first adaptive filter based on the first comparison signal and the internal reference signal.

[0123] The signal control device according to aspect 6 of the present disclosure may be configured such that, in aspect 5 above, the internal reference signal generation unit includes a second adaptive filter that processes the virtual noise signal to generate an intermediate signal, and a signal combining unit that combines the noise signal and the intermediate signal to generate the internal reference signal.

[0124] A signal control device according to Embodiment 7 of the present disclosure, in Embodiment 6 above, comprises a second update unit that determines the parameters of the second adaptive filter by updating the parameters of the second adaptive filter during the tuning stage of the second adaptive filter, and an error signal acquisition unit that acquires an error signal corresponding to the sound acquired by a virtual microphone temporarily installed at the sound-dampening position during the tuning stage, wherein during the tuning stage, the first update unit updates the filter coefficients of the first adaptive filter based on the error signal and the virtual noise signal, and the second update unit determines the parameters of the second adaptive filter by updating the parameters of the second adaptive filter based on the internal reference signal and the virtual noise signal.

[0125] The signal control device according to embodiment 8 of the present disclosure may be configured such that, in embodiment 7 above, the second adaptive filter is an FIR filter, and the second update unit determines the filter coefficients of the second adaptive filter by updating the filter coefficients of the second adaptive filter during the tuning stage.

[0126] The signal control device according to aspect 9 of the present disclosure, in aspect 8 described above, may be configured such that the first update unit virtually reproduces a transfer function corresponding to the propagation path from the speaker to the virtual microphone and processes the virtual noise signal to generate a second comparison signal, and in the tuning stage, the coefficient update unit updates the filter coefficients of the first adaptive filter based on the second comparison signal and the error signal, and the second update unit updates the filter coefficients of the second adaptive filter based on the virtual noise signal and the internal reference signal.

[0127] The noise reduction device according to embodiment 10 of the present disclosure may be configured to include the signal control device, the reference microphone, and the speaker in any of embodiments 1 to 9 described above.

[0128] The signal control program according to aspect 11 of this disclosure may be a signal control program for causing a computer to function as the signal control device in any of aspects 1 to 9 described above.

[0129] The signal control program according to aspect 12 of the present disclosure is a signal control program for causing a computer to function as the signal control device in any of aspects 2 to 9 described above, and may be configured to cause the computer to function as the first signal generation unit, the second signal generation unit, the first adaptive filter, and the first update unit.

[0130] A signal control method according to embodiment 13 of the present disclosure includes the steps of: generating a virtual feedback signal that virtually reproduces a signal corresponding to a feedback sound propagated from a speaker that generates a noise reduction sound to reduce sound pressure at a sound silencing position to a reference microphone located between the noise source and the speaker; generating a virtual noise signal by subtracting the virtual feedback signal from a noise signal corresponding to a sound acquired by the reference microphone; and generating a noise reduction signal for generating the noise reduction sound from the speaker by adaptive signal processing according to the noise from the noise source based on the noise signal and the virtual noise signal.

[0131] The signal control method according to aspect 14 of the present disclosure may be configured such that, in aspect 13 above, the step of generating the noise reduction signal includes the steps of: generating the noise reduction signal by processing the virtual noise signal with a first adaptive filter; and updating the parameters of the first adaptive filter based on the noise signal and the virtual noise signal.

[0132] This disclosure is not limited to the embodiments described above, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the different technical means disclosed in each embodiment are also included in the technical scope of this disclosure. [Explanation of symbols]

[0133] 1. Noise reduction device 11 Signal control device 12 Reference Microphone 13 speakers 31 1st signal generation section 32 Second signal generation section 41. First adaptive filter (noise reduction signal generation unit) 42. Second Adaptive Filter 61. First Update Unit (Noise Reduction Signal Generation Unit) 62 2nd update part 71 Comparison Filter 72 Internal reference signal generator 74 Signal Synthesis Unit 81 Error signal acquisition section E Mute position VM Virtual Microphone X Noise source

Claims

1. A first signal generation unit generates a virtual feedback signal that virtually reproduces a signal corresponding to the feedback sound propagating from a speaker that generates noise reduction sound to reduce sound pressure at the sound silencing position to a reference microphone located between the noise source and the speaker. A second signal generation unit generates a virtual noise signal by subtracting the virtual feedback signal from the noise signal corresponding to the sound acquired by the reference microphone, A signal control device comprising: a noise reduction signal generation unit that generates a noise reduction signal for generating the noise reduction sound from the speaker by performing adaptive signal processing according to the noise from the noise source based on the noise signal and the virtual noise signal.

2. The noise reduction signal generation unit, A first adaptive filter that processes the virtual noise signal to generate the noise reduction signal, The signal control device according to claim 1, further comprising: a first update unit that updates the parameters of the first adaptive filter based on the noise signal and the virtual noise signal.

3. The signal control device according to claim 2, wherein the first signal generation unit includes a virtual filter that virtually reproduces a transfer function corresponding to the propagation path from the speaker to the reference microphone.

4. The first adaptive filter is an FIR filter, The signal control device according to claim 2, wherein the first update unit updates the filter coefficients of the first adaptive filter based on the noise signal and the virtual noise signal.

5. The first update section is, A first comparison filter virtually reproduces the transfer function corresponding to the propagation path from the speaker to the reference microphone, and processes the virtual noise signal to generate a first comparison signal. An internal reference signal generation unit generates an internal reference signal whose mean square is minimized when the sound pressure at the silenced position is minimized, based on the aforementioned noise signal and the aforementioned virtual noise signal. The signal control device according to claim 4, comprising: a coefficient update unit that updates the filter coefficients of the first adaptive filter based on the first comparison signal and the internal reference signal.

6. The internal reference signal generation unit is, A second adaptive filter processes the aforementioned virtual noise signal to generate an intermediate signal, The signal control device according to claim 5, comprising a signal combining unit that combines the noise signal and the intermediate signal to generate the internal reference signal.

7. The system includes a second update unit that determines the parameters of the second adaptive filter by updating the parameters of the second adaptive filter during the tuning stage of the second adaptive filter, and an error signal acquisition unit that acquires an error signal corresponding to the sound acquired by a virtual microphone temporarily installed at the sound-silencing position during the tuning stage. In the aforementioned tuning stage, The first update unit updates the filter coefficients of the first adaptive filter based on the error signal and the virtual noise signal, and The signal control device according to claim 6, wherein the second update unit determines the parameters of the second adaptive filter by updating the parameters of the second adaptive filter based on the internal reference signal and the virtual noise signal.

8. The second adaptive filter is an FIR filter, The signal control device according to claim 7, wherein the second update unit determines the filter coefficients of the second adaptive filter by updating the filter coefficients of the second adaptive filter during the tuning stage.

9. The first update unit includes a second comparison filter that virtually reproduces the transfer function corresponding to the propagation path from the speaker to the virtual microphone, and processes the virtual noise signal to generate a second comparison signal. In the aforementioned tuning stage, The coefficient update unit updates the filter coefficients of the first adaptive filter based on the second comparison signal and the error signal. The signal control device according to claim 8, wherein the second update unit updates the filter coefficients of the second adaptive filter based on the virtual noise signal and the internal reference signal.

10. A noise reduction device comprising a signal control device according to any one of claims 1 to 9, the reference microphone, and the speaker.

11. A signal control program for causing a computer to function as a signal control device according to any one of claims 1 to 9.

12. A signal control program for causing a computer to function as a signal control device according to any one of claims 2 to 9, wherein the signal control program causes the computer to function as the first signal generation unit, the second signal generation unit, the first adaptive filter, and the first update unit.

13. A step of generating a virtual feedback signal that virtually reproduces a signal corresponding to the feedback sound propagating from a speaker that generates noise reduction sound to reduce sound pressure at the sound silencing position to a reference microphone located between the noise source and the speaker, A step of generating a virtual noise signal by subtracting the virtual feedback signal from the noise signal corresponding to the sound acquired by the reference microphone, A signal control method comprising the step of generating a noise reduction signal for generating the noise reduction sound from the speaker by adaptive signal processing corresponding to the noise from the noise source based on the noise signal and the virtual noise signal.

14. The step of generating the noise reduction signal is: The steps include: processing the virtual noise signal with a first adaptive filter to generate the noise reduction signal; The signal control method according to claim 13, comprising the step of updating the parameters of the first adaptive filter based on the noise signal and the virtual noise signal.

Images