Active noise control device

Abstract

This invention relates to an active noise control device. The active noise control device for controlling the output of a canceled tone to eliminate noise includes: a control signal generator configured to generate a control signal based on an input signal; and a frequency converter configured to convert the sampling frequency of the control signal to generate the canceled tone output signal. The frequency converter is configured to: generate a first upsampled signal by increasing the sampling frequency of the control signal; generate a correction signal by correcting the first upsampled signal based on a predicted value of the first upsampled signal; and generate the canceled tone output signal by performing low-pass filtering on the correction signal or a signal based on the correction signal.

Description

Technical Field

[0001] This invention relates to an active noise control device. Background Technology

[0002] Active noise control devices that control the output of cancel tone to eliminate noise are generally known. For example, an active noise control device includes a control signal generator that generates a control signal based on an input signal, and a frequency converter that converts the sampling frequency of the control signal.

[0003] JP2010-288246A discloses a technique for converting the sampling frequency of an input signal (differential code), wherein the sampling frequency of the input signal is increased and then noise in the input signal is removed by low-pass filtering.

[0004] As mentioned above, increasing the sampling frequency of the input signal reduces the smoothness (waveform smoothness) of the input signal. When low-pass filtering is performed on an input signal with reduced smoothness, the computational load required for low-pass filtering increases, and the delay (delay time) of the output signal relative to the input signal increases. If this technique with a large delay is applied to an active noise control device, the responsiveness of the canceled tone to noise may decrease, resulting in a reduction in noise reduction performance. Summary of the Invention

[0005] In view of the above background, the object of the present invention is to provide an active noise control device that can suppress noise and delay simultaneously when switching the sampling frequency of the control signal.

[0006] To achieve this objective, one aspect of the present invention provides an active noise control device for controlling the output of a canceled tone to eliminate noise. The active noise control device includes: a control signal generator configured to generate a control signal based on an input signal; and a frequency converter configured to convert the sampling frequency of the control signal to generate the canceled tone output signal. The frequency converter is configured to: generate a first upsampled signal by increasing the sampling frequency of the control signal; generate a correction signal by correcting the first upsampled signal based on a predicted value of the first upsampled signal; and generate the canceled tone output signal by performing low-pass filtering on the correction signal or a signal based on the correction signal.

[0007] Therefore, based on the above aspects, an active noise control device can be provided that can simultaneously suppress noise and delay when switching the sampling frequency of the control signal. Attached Figure Description

[0008] Figure 1This is a schematic diagram showing a vehicle to which an active noise control device according to an embodiment of the present invention is applied; Figure 2 This is a block diagram illustrating an active noise control device according to this embodiment; Figure 3 It is a waveform diagram showing the time variation of the amplitude of the correction signal; Figure 4 It is a waveform diagram showing the time variation of the amplitude of the uncorrected signal; Figure 5 It is a waveform diagram showing the time variation of the amplitude of the second upsampled signal and the output signal; Figure 6 It is a waveform diagram showing the relationship between frequency and noise reduction amount; Figure 7A This is an explanatory diagram illustrating the process flow of sampling-based processing; Figure 7B This is an explanatory diagram illustrating the process of frame-based processing; and Figure 8 This is a block diagram illustrating an active noise control device according to another embodiment. Detailed Implementation

[0009] Hereinafter, an active noise control device 1 (hereinafter simply referred to as "noise control device 1") according to an embodiment of the present invention will be described with reference to the accompanying drawings. In the description, the symbol "^" (cap) attached to the reference numerals indicates an identification value or estimated value. In the drawings, the symbol "^" is arranged above the reference numerals, while in the description, the symbol "^" is arranged after the reference numerals.

[0010] <Vehicle 3>

[0011] Figure 1 This is a schematic diagram showing a vehicle 3 incorporating a noise control device 1. The vehicle 3 includes a body 4 defining the exterior of the vehicle 3. A passenger compartment 5 is formed inside the body 4. The passenger compartment 5 is provided with a plurality of passenger seats 7.

[0012] Carriage 5 is equipped with multiple speakers 9 (in Figure 1 Only one speaker 9 is shown in the diagram. Multiple speakers 9 are examples of noise cancellation output devices. Each speaker 9 reproduces audio content (music, radio, voice guidance, etc.) and outputs a cancellation tone y that is out of phase with the noise d. The cancellation tone y output from the speaker 9 interferes with the noise d, causing the cancellation tone y and the noise d to cancel each other out, thereby reducing the noise d. For example, the speaker 9 is installed in a location other than the passenger seat 7 in the carriage 5. In another embodiment, the speaker 9 may be installed on the passenger seat 7.

[0013] Carriage 5 is equipped with multiple error microphones 11 ( Figure 1Only one error microphone 11 is shown in the diagram. Each error microphone 11 generates an error signal e based on noise d and canceled tone y. For example, the error microphone 11 is mounted on the passenger seat 7. In another embodiment, the error microphone 11 may be mounted in a location other than the passenger seat 7 in the carriage 5.

[0014] Multiple wheels 13 are arranged below the vehicle body 4. A suspension 14 is arranged between the vehicle body 4 and each wheel 13. An acceleration sensor 15 is provided on the suspension 14. The acceleration sensor 15 detects the acceleration of the suspension 14 based on noise d and generates a reference signal r based on the acceleration of the suspension 14. In another embodiment, the acceleration sensor 15 may be arranged at a location other than the suspension 14 of the vehicle 3. For example, the acceleration sensor 15 may be attached to the frame of the vehicle body 4.

[0015] <ECU 21>

[0016] See Figure 1 Vehicle 3 is equipped with an integrated electronic control unit (ECU) 21. The ECU 21, together with the speaker 9, error microphone 11 and acceleration sensor 15, constitutes an active noise cancellation system 22.

[0017] ECU 21 includes a processor 24 (e.g., CPU or MPU) and a memory 25 (e.g., ROM or RAM). The processor 24 executes a program stored in the memory 25, causing multiple functional units of ECU 21 to perform their functions. ECU 21 is equipped with multiple functional units, including an audio control unit 27, a noise control unit 1, and a signal combiner 28.

[0018] The audio control device 27 controls the speaker 9 to reproduce audio content (music, radio, voice guidance, etc.). The audio control device 27 generates an audio signal m corresponding to the audio content and outputs the generated audio signal m to the signal combiner 28.

[0019] The noise control device 1 controls the output of the cancelling tone y via the loudspeaker 9. The noise control device 1 generates an output signal So (hereinafter referred to as "output signal So") for the cancelling tone y based on the error signal e and the reference signal r, and outputs the generated output signal So to the signal combiner 28. The noise control device 1 will be described in detail below.

[0020] Signal combiner 28 combines the audio signal m and the output signal So, and outputs the combined audio signal m and the output signal So to speaker 9. In response, speaker 9 reproduces the audio content corresponding to the audio signal m and outputs a canceled tone y corresponding to the output signal So. It should be noted that in Figure 2 Signal combiner 28 is omitted.

[0021] <Noise Control Device 1>

[0022] Reference Figure 2 The noise control device 1 includes: a control signal generator 31 that generates a control signal u based on a reference signal r and an error signal e (both of which are examples of input signals); and a frequency converter 32 that converts the sampling frequency of the control signal u.

[0023] <Control Signal Generator 31>

[0024] The control signal generator 31 of the noise control device 1 includes a control filter unit 34, a reference signal correction unit 35, and a control update unit 36.

[0025] The control filter unit 34 is composed of a control filter W. For example, the control filter W is composed of a finite impulse response (FIR) filter or a single-frequency adaptive notch filter (SAN). The control filter unit 34 generates a control signal u by filtering the reference signal r using the control filter W. The control filter unit 34 outputs the generated control signal u to the frequency converter 32.

[0026] The reference signal correction unit 35 is composed of a secondary path filter C^. The secondary path filter C^ is a filter that represents an estimate of the transfer function C of the secondary path from the speaker 9 to the error microphone 11. For example, the secondary path filter C^ is composed of an FIR filter or a SAN filter. The reference signal correction unit 35 corrects the reference signal r by filtering the reference signal r using the secondary path filter C^. The reference signal correction unit 35 outputs the corrected reference signal r to the control update unit 36.

[0027] The control update unit 36 ​​adaptively updates the control filter W using an adaptive algorithm such as the least mean square (LMS) algorithm. More specifically, the control update unit 36 ​​adaptively updates the control filter W such that the error signal e output from the error microphone 11 is minimized.

[0028] <Frequency Converter 32>

[0029] The frequency converter 32 of the noise control device 1 includes a first upsampling unit 41, a correction signal generation unit 42, a second upsampling unit 43, and a filtering processing unit 44.

[0030] The first upsampling unit 41 generates a first upsampled signal Su1 by increasing the sampling frequency of the control signal u output from the control signal generator 31. For example, the first upsampling unit 41 increases the sampling frequency of the control signal u by performing an integer-multiple sampling-frequency conversion on the control signal u. The first upsampling unit 41 outputs the generated first upsampled signal Su1 to the correction signal generation unit 42.

[0031] The correction signal generation unit 42 generates a correction signal Si by correcting the first upsampled signal Su1 based on the predicted value of the first upsampled signal Su1. The process of generating the correction signal Si by the correction signal generation unit 42 will be described in detail later. The correction signal generation unit 42 outputs the generated correction signal Si to the second upsampled unit 43.

[0032] The second upsampling unit 43 generates a second upsampling signal Su2 by increasing the sampling frequency of the correction signal Si. For example, the second upsampling unit 43 increases the sampling frequency of the correction signal Si by performing an integer-multiple sampling-frequency conversion on the correction signal Si. The increase rate of the sampling frequency of the second upsampling signal Su2 relative to the sampling frequency of the correction signal Si is higher than the increase rate of the sampling frequency of the first upsampling signal Su1 relative to the sampling frequency of the control signal u. The second upsampling unit 43 outputs the generated second upsampling signal Su2 to the filtering processing unit 44.

[0033] The filtering unit 44 is composed of a low-pass filter L. The filtering unit 44 generates the output signal So by performing low-pass filtering on the second upsampled signal Su2 (an example of a signal based on the correction signal Si). Therefore, performing low-pass filtering on the second upsampled signal Su2 reduces aliasing noise included in the second upsampled signal Su2.

[0034] <Generation and processing of correction signal Si>

[0035] Figure 3 This is a waveform diagram showing the time variation of the amplitude of the correction signal Si. Figure 3 In the time interval an (n=1, 2, 3...), the time interval an (hereinafter referred to as "basic time interval an") represents the moment of each basic time interval Ts. In an orthogonal coordinate system (hereinafter referred to as "time-amplitude coordinate system") using the time and amplitude of the signal as two coordinate axes, the coordinates of the point representing the amplitude Xan of the signal at a certain basic time interval an are represented as (an, Xan). Figure 3In the time-amplitude coordinate system, time bn (n=1, 2, 3...) represents the moment when half a time interval Ts / 2 has elapsed since the basic time interval an (hereinafter referred to as "half-time interval bn"). In the time-amplitude coordinate system, the coordinates of the point representing the amplitude Xbn of the signal at a certain half-time interval bn are represented as (bn, Xbn).

[0036] The first upsampling unit 41 generates a first upsampling signal Su1 for each basic time period Ts and outputs the generated first upsampling signal Su1 to the correction signal generation unit 42. When the first upsampling signal Su1 with an amplitude of Xa1 is input to the correction signal generation unit 42 at the basic time period a1, the correction signal generation unit 42 sets the amplitude of the correction signal Si to Xa1. Between the basic time period a1 and the basic time period a2, the correction signal generation unit 42 maintains the amplitude of the correction signal Si at Xa1 through sample-and-hold processing.

[0037] When a first upsampled signal Su1 with an amplitude Xa2 is input at the basic time period a2, the correction signal generation unit 42 switches the amplitude of the correction signal Si from Xa1 to Xa2. Furthermore, the correction signal generation unit 42 obtains a predicted value of the amplitude of the first upsampled signal Su1 at the half-time period b2 based on the time change of the amplitude Xa2 of the first upsampled signal Su1 at the basic time period a2 relative to the amplitude Xa1 of the first upsampled signal Su1 at the basic time period a1.

[0038] More specifically, the correction signal generation unit 42 generates a straight line (hereinafter referred to as "projection line La2") in the time-amplitude coordinate system. This straight line passes through the point (a1, Xa1) representing the amplitude Xa1 of the first upsampled signal Su1 at the basic time interval a1 and the point (a2, Xa2) representing the amplitude Xa2 of the first upsampled signal Su1 at the basic time interval a2. That is, the correction signal generation unit 42 generates the projection line La2 by linearly approximating the time variation of the amplitude of the first upsampled signal Su1 using a first-order Taylor series expansion. The correction signal generation unit 42 obtains the amplitude Xb2 at the intersection of the projection line La2 and the straight line Lb2 representing the half-time interval b2 as the predicted value of the amplitude of the first upsampled signal Su1 at the half-time interval b2. As described above, the correction signal generation unit 42 calculates the predicted value of the first upsampled signal Su1 by extrapolation processing (extrapolation operation) based on the amplitudes Xa1 and Xa2 of the first upsampled signal Su1 at the basic time periods a1 and a2 (i.e., the past values ​​of the first upsampled signal Su1).

[0039] Between the basic time period a2 and the half-time period b2, the correction signal generation unit 42 maintains the amplitude of the correction signal Si at Xa2 through sample-and-hold processing. At the half-time period b2, the correction signal generation unit 42 switches the amplitude of the correction signal Si from Xa2 to Xb2.

[0040] Starting from the basic time period a3, the correction signal generation unit 42 repeats the same process as described above. That is, the correction signal generation unit 42 obtains the predicted value of the amplitude of the first upsampled signal Su1 at the next half time period bn (the midpoint between the current basic time period an and the next basic time period a(n+1)) based on the time change of the amplitude Xan of the first upsampled signal Su1 at the current basic time period an relative to the amplitude Xa(n-1) of the first upsampled signal Su1 at the previous basic time period a(n-1).

[0041] More specifically, the correction signal generation unit 42 generates a projection line Lan in the time-amplitude coordinate system. This projection line Lan passes through the point (a(n-1), Xa(n-1)) representing the amplitude Xa(n-1) of the first upsampled signal Su1 at the previous basic time interval a(n-1) and the point (an, Xan) representing the amplitude Xan of the first upsampled signal Su1 at the current basic time interval an. That is, the correction signal generation unit 42 generates the projection line Lan by linearly approximating the time variation of the amplitude of the first upsampled signal Su1 using a first-order Taylor series expansion. The correction signal generation unit 42 obtains the amplitude Xbn at the intersection of the projection line Lan and the straight line Lbn representing the half-time interval bn as the predicted value of the amplitude of the first upsampled signal Su1 at the next half-time interval bn.

[0042] Between the basic time period an and the half-time period bn, the correction signal generation unit 42 maintains the amplitude of the correction signal Si at Xan through sample-and-hold processing. At the half-time period bn, the correction signal generation unit 42 switches the amplitude of the correction signal Si from Xan to Xbn. As described above, the correction signal generation unit 42 generates the correction signal Si by correcting the first upsampled signal Su1 based on the predicted value of the amplitude of the first upsampled signal Su1 at the second half-time period bn.

[0043] <Effect>

[0044] Figure 4 This is a waveform showing the time variation of the amplitude of the signal without the aforementioned correction (hereinafter referred to as the "uncorrected signal"). The amplitude of the uncorrected signal is held at Xan by a sample-and-hold process from the fundamental time interval an to the next fundamental time interval a(n+1). Figure 3 and Figure 4The comparison clearly shows that the waveform of the corrected signal Si is smoother than that of the uncorrected signal. That is, the corrected signal Si is smoother than the uncorrected signal.

[0045] Figure 5 It is a waveform diagram showing the time variation of the amplitude of the second upsampled signal Su2 (the signal just before the low-pass filtering process) and the output signal So (the signal just after the low-pass filtering process). Figure 5 The term "uncorrected" in the text indicates that the first upsampled signal Su1 was not corrected based on the amplitude-based prediction. Figure 5 The term "correction" in this context indicates that the amplitude-based predicted value of the first upsampled signal Su1 is corrected. When the first upsampled signal Su1 is corrected based on the amplitude-based predicted value, the smoothness of the second upsampled signal Su2 is improved, and the computational cost of low-pass filtering is reduced compared to the case where the first upsampled signal Su1 is not corrected based on the amplitude-based predicted value. Therefore, when the first upsampled signal Su1 is corrected based on the amplitude-based predicted value, the delay of the output signal So is reduced by Z (ms) compared to the case where the first upsampled signal Su1 is not corrected based on the amplitude-based predicted value.

[0046] Figure 6 The waveforms show the relationship between frequency and noise reduction in the example and comparative examples. Figure 6 The phrase "no low-pass filter" in the text indicates that the frequency converter 32 does not include a low-pass filter L. Figure 6 The phrase "with low-pass filter" indicates that the frequency converter 32 includes a low-pass filter L. Figure 6 The content of "uncorrected" and "corrected" in the text is the same as... Figure 5 The content is the same as in [previous example], so its description is omitted. In this example (with a low-pass filter, corrected), the amount of noise reduction in the high-frequency band Fh (an example of the band as the target for noise reduction) is increased compared to Comparative Example 1 (without a low-pass filter) and Comparative Example 2 (with a low-pass filter, uncorrected). This shows that high noise reduction performance can be achieved by performing low-pass filtering on the signal obtained by correcting the first upsampled signal Su1 using amplitude-based prediction values.

[0047] There are two types of signal processing methods: sample-based processing and frame-based processing. Figure 7A This is an explanatory diagram illustrating the process flow of sampling-based processing, and Figure 7B This is an illustrative diagram showing the flow of frame-based processing. (Reference) Figure 7A In sample-based processing, the input signal is processed on a sample-by-sample basis to generate the output signal. Therefore, the output signal is delayed by one sample relative to the input signal. (See reference...) Figure 7BIn frame-based processing, frames containing multiple samples of the input signal are generated, and these multiple samples are batch-processed on a frame-by-frame basis to generate the output signal. Therefore, the output signal is delayed by two frames relative to the input signal. Thus, compared to sample-based processing, frame-based processing increases the delay of the output signal relative to the input signal.

[0048] Traditionally, noise control device 1 and audio control device 27 are mounted on separate ECU 21, with sample-based processing used for noise control device 1 and frame-based processing used for audio control device 27. However, according to this embodiment, since noise control device 1 and audio control device 27 are mounted on the same ECU 21, the signal processing methods for noise control device 1 and audio control device 27 need to be identical. Therefore, frame-based processing applies not only to audio control device 27 but also to noise control device 1. Thus, frequency converter 32 simultaneously converts the sampling frequency of the control signal u for multiple samples through frame-based processing (batch processing) of the control signal u for multiple samples.

[0049] For the reasons mentioned above, if frame-based processing is used for noise control device 1, there is a risk of increased delay in the output signal So compared to sampling-based processing. If the delay in the output signal So increases, the responsiveness of the canceled tone y to the noise d may decrease, resulting in reduced noise reduction performance.

[0050] Therefore, the frequency converter 32 generates a correction signal Si by correcting the first upsampled signal Su1 based on the predicted value of the first upsampled signal Su1. This improves the smoothness of the correction signal Si and reduces the computational load required for low-pass filtering. Therefore, even when the noise control device 1 employs frame-based processing, the delay of the output signal So can be reduced, and high noise reduction performance can be achieved.

[0051] <Modified Implementation Method>

[0052] In the above embodiment, the filtering unit 44 generates the output signal So by performing low-pass filtering on the second upsampled signal Su2 (an example of a signal based on the correction signal Si). (Refer to...) Figure 8 According to another embodiment, the filtering unit 44 can generate the output signal So by performing low-pass filtering on the correction signal Si itself. That is, according to another embodiment, the second upsampling unit 43 can be omitted.

[0053] In the above embodiments, the rate of increase of the sampling frequency of the second upsampled signal Su2 relative to the sampling frequency of the correction signal Si is higher than the rate of increase of the sampling frequency of the first upsampled signal Su1 relative to the sampling frequency of the control signal u. In another embodiment, the rate of increase of the sampling frequency of the second upsampled signal Su2 relative to the sampling frequency of the correction signal Si can be equal to or less than the rate of increase of the sampling frequency of the first upsampled signal Su1 relative to the sampling frequency of the control signal u. When the sampling frequency of the control signal u is Fu, the sampling frequency of the first upsampled signal Su1 is FSu1, and the sampling frequency of the second upsampled signal Su2 is FSu2, the following equations (1) to (3) are maintained.

[0054] Fu <FSu1<FSu2(1)

[0055] FSu1=n×Fu (n=1, 2,...) (2)

[0056] FSu2=m×FSu1 (m=1, 2,...) (3)

[0057] In the above embodiment, the noise control device 1 is applied to the passenger compartment 5 of the vehicle 3. Alternatively, in another embodiment, the noise control device 1 can be applied to the interior space of a moving object other than the vehicle 3 (e.g., a ship or an aircraft), or the noise control device 1 can be applied to the interior space of a fixed object (e.g., a house).

[0058] The specific implementation methods have been described above, but the present invention is not limited to the above-described implementation methods or modified implementation methods, and can be widely modified and implemented.

[0059] <Summary of Implementation Methods>

[0060] An active noise control device 1 for controlling the output of canceled tone y to cancel noise d, the active noise control device comprising: a control signal generator 31 configured to generate a control signal u based on input signals r and e; and a frequency converter 32 configured to convert the sampling frequency of the control signal u to generate an output signal So for canceled tone y, the frequency converter 32 being configured to: generate a first upsampled signal Su1 by increasing the sampling frequency of the control signal u; generate a correction signal Si by correcting the first upsampled signal Su1 based on a predicted value of the first upsampled signal Su1; and generate the output signal So for canceled tone y by performing low-pass filtering on the correction signal Si or a signal Su2 based on the correction signal Si.

[0061] According to this aspect, the frequency converter 32 can suppress noise by performing low-pass filtering on the extrapolated signal Si or the signal Su2 based on the extrapolated signal Si. Furthermore, the frequency converter 32 generates a correction signal Si by correcting the first upsampled signal Su1 based on the predicted value of the first upsampled signal Su1. This improves the smoothness (waveform smoothness) of the correction signal Si and reduces the computational load required for low-pass filtering. Therefore, the delay of the output signal So relative to the input signals r and e can be reduced (delay time), and the responsiveness of the canceled tone y to noise d can be improved, thereby achieving high noise reduction performance. As described above, when the sampling frequency of the control signal u is converted, noise and delay can be suppressed simultaneously.

[0062] The frequency converter 32 is configured to generate a second upsampled signal Su2 by increasing the sampling frequency of the correction signal Si; and to generate an output signal So for canceling tone y by performing low-pass filtering on the second upsampled signal Su2.

[0063] Based on the above, the sampling frequency of the control signal u can be increased to the target value by performing two upsampling processes (processes that increase the sampling frequency). Therefore, compared to increasing the sampling frequency of the control signal u to the target value by performing only one upsampling process, the sampling frequency of the first upsampled signal Su1 (the signal after the first upsampling process) can be reduced. Therefore, the computational load required for generating the correction signal Si based on the first upsampled signal Su1 can be reduced, thereby further reducing the delay of the output signal So relative to the input signals r and e.

[0064] The frequency converter 32 is configured to: increase the sampling frequency of the control signal u by performing an integer-multiple sampling-frequency conversion on the control signal u; and increase the sampling frequency of the correction signal Si by performing an integer-multiple sampling-frequency conversion on the correction signal Si.

[0065] According to this aspect, by performing integer-multiple sampling-frequency conversion in both upsampling processes, the high-frequency components included in the second upsampled signal Su2 (the signal after the second upsampling process) can be reduced, and the smoothness of the second upsampled signal Su2 can be improved. Therefore, the computational cost required for low-pass filtering based on the second upsampled signal Su2 can be further reduced, thereby further reducing the delay of the output signal So relative to the input signals r and e.

[0066] The frequency converter 32 is configured to set the increase rate of the sampling frequency of the second upsampled signal Su2 relative to the sampling frequency of the correction signal Si to be higher than the increase rate of the sampling frequency of the first upsampled signal Su1 relative to the sampling frequency of the control signal u.

[0067] Based on this, the sampling frequency of the first upsampled signal Su1 can be further reduced. Therefore, the computational load required for generating the correction signal Si based on the first upsampled signal Su1 can be further reduced, thereby further reducing the delay of the output signal So relative to the input signals r and e.

[0068] The frequency converter 32 is configured to generate an output signal So for canceling tone y by performing low-pass filtering on the correction signal Si.

[0069] According to this aspect, the configuration of the active noise control device 1 can be simplified compared to the case where a low-pass filter is applied to the signal obtained by performing a second upsampling process on the correction signal Si.

[0070] The frequency converter 32 is configured to: obtain a predicted value of the amplitude of the first upsampled signal Su1 based on the time variation of the amplitude of the first upsampled signal Su1; and generate a correction signal Si by correcting the first upsampled signal Su1 based on the predicted value of the amplitude of the first upsampled signal Su1.

[0071] Based on this aspect, by correcting the first upsampled signal Su1 based on the predicted value of the amplitude of the first upsampled signal Su1, the smoothness of the corrected signal Si can be improved. Therefore, the computational load required for low-pass filtering can be further reduced, thereby further reducing the delay of the output signal So relative to the input signals r and e.

[0072] The frequency converter 32 is configured to: generate a first upsampled signal Su1 for each basic time period an; and obtain a predicted value of the amplitude of the first upsampled signal Su1 at the intermediate time bn between the current basic time period an and the next basic time period a(n+1) based on the time change of the amplitude of the first upsampled signal Su1 at the current basic time period an relative to the amplitude of the first upsampled signal Su1 at the previous basic time period a(n-1).

[0073] Based on this, the smoothness of the correction signal Si can be improved without making the generation process of the correction signal Si overly complex. Therefore, the computational load required for the generation process of the correction signal Si can be reduced, thereby further reducing the delay of the output signal So relative to the input signals r and e.

[0074] The frequency converter 32 is configured to simultaneously convert the sampling frequency of the control signal u for multiple samples by batch processing the control signal u for multiple samples.

[0075] According to this, by batch processing the control signal u for multiple samples, the processing efficiency of the control signal u can be improved compared to processing the control signal u sample by sample.

[0076] The frequency converter 32 is configured to calculate the predicted value of the first upsampled signal Su1 by extrapolation processing.

[0077] Based on this, the predicted value of the first upsampled signal Su1 can be calculated with high precision.

Claims

1. An active noise control device for controlling the output of cancelling tone to eliminate noise, the active noise control device comprising: A control signal generator, the control signal generator being configured to generate a control signal based on an input signal; as well as A frequency converter configured to convert the sampling frequency of the control signal to generate the output signal for canceling the tone, wherein, The frequency converter is configured to: A first upsampled signal is generated by increasing the sampling frequency of the control signal; A correction signal is generated by correcting the first upsampled signal based on the predicted value of the first upsampled signal; and The output signal for canceling the tone is generated by performing low-pass filtering on the correction signal or a signal based on the correction signal.

2. The active noise control device according to claim 1, wherein, The frequency converter is configured to: A second upsampled signal is generated by increasing the sampling frequency of the correction signal; and The output signal for canceling the tone is generated by performing the low-pass filtering process on the second upsampled signal.

3. The active noise control device according to claim 2, wherein, The frequency converter is configured to: The sampling frequency of the control signal is increased by performing an integer-multiple sample-frequency conversion on the control signal; and The sampling frequency of the correction signal is increased by performing an integer-multiple sampling-frequency conversion on the correction signal.

4. The active noise control device according to claim 2, wherein, The frequency converter is configured to set the increase rate of the sampling frequency of the second upsampled signal relative to the sampling frequency of the correction signal to be higher than the increase rate of the sampling frequency of the first upsampled signal relative to the sampling frequency of the control signal.

5. The active noise control device according to claim 1, wherein, The frequency converter is configured to generate the output signal for canceling the tone by performing the low-pass filtering process on the correction signal.

6. The active noise control device according to any one of claims 1 to 5, wherein, The frequency converter is configured to: The predicted value of the amplitude of the first upsampled signal is obtained based on the time variation of the amplitude of the first upsampled signal; and The corrected signal is generated by correcting the first upsampled signal based on the predicted value of the amplitude of the first upsampled signal.

7. The active noise control device according to claim 6, wherein, The frequency converter is configured to: The first upsampled signal is generated for each basic time period; and The predicted value of the amplitude of the first upsampled signal at the midpoint between the current basic time period and the next basic time period is obtained based on the temporal change of the amplitude of the first upsampled signal at the current basic time period relative to the amplitude of the first upsampled signal at the previous basic time period.

8. The active noise control device according to any one of claims 1 to 5, wherein, The frequency converter is configured to simultaneously convert the sampling frequency of the control signals for the multiple samples by batch processing the control signals for the multiple samples.

9. The active noise control device according to any one of claims 1 to 5, wherein, The frequency converter is configured to calculate the predicted value of the first upsampled signal based on the first upsampled signal through extrapolation processing.

Images