JP5982728B2 - Active noise control device and active noise control method - Google Patents

Description

  The present invention relates to an active noise control device, an active noise control method, an active noise control program, and a recording medium on which the active noise control program is recorded.

  2. Description of the Related Art In recent years, active noise control (ANC) devices that generate noise (control sound) having an opposite phase to noise (control target sound) and perform noise suppression control have attracted attention. Such an active noise control device is used, for example, to control engine sound as noise that can be heard in the passenger compartment of a vehicle with control sound output from a speaker, and to reduce engine sound at a passenger's ear position.

  In order to prevent the divergence of the control system employed for such an active noise control device, the power of the control sound signal (and thus the power of the control sound) is detected, and when the detection result exceeds a predetermined threshold value, A technique for determining that divergence has occurred has been proposed (see Patent Document 1: hereinafter referred to as “Conventional Example 1”). In the technique of the conventional example 1, when it is determined that the divergence of the control system has occurred, the control sound output is stopped.

  In the technique of Conventional Example 1, the output voltage of the drive amplifier that drives the speaker that outputs the control sound is detected. If the detected voltage value is large with respect to the power of the control sound signal intended in design, the predetermined threshold value is increased. On the other hand, if the detected voltage value is small with respect to the power of the control sound signal intended by design, the predetermined threshold value is decreased. In this way, by controlling the predetermined threshold value, erroneous determination regarding the presence / absence of divergence due to variations in the operation characteristics unique to the apparatus due to variations in the characteristics of the constituent elements is prevented.

  Although not a technique for dealing with control system divergence, the phase characteristics of the modeled transfer function from the adaptive filter output to the error signal input are convolved as a control system control algorithm in the active noise control device. A technique that employs a Filtered-X LMS (Least Mean Square) algorithm that generates a control sound signal is proposed (refer to Patent Document 2: hereinafter referred to as “conventional example 2”). In the technique of Conventional Example 2, the Filtered-X LMS algorithm is adopted, and the step size parameter value is set to “Large” in the transient state based on the average value of the measurement result of the error sensor for measuring the control result over a certain period of time. By making asymptotically variable so as to be “small” in the steady state, the convergence is fast and stable control is possible in the steady state.

JP-A-5-303387 JP 2001-255877 A

  In the technique of Conventional Example 1 described above, a value larger than the normal power value of the control sound of the normal control operation assumed in design is adopted as the predetermined threshold value. For this reason, after the volume of the abnormal control sound accompanying the divergence of the control system has become considerably large, it is determined that the control system diverges and the control sound output is stopped. As a result, there is a case in which abnormal noise or explosive noise occurs before the control sound output is stopped.

  Also, in a control system that employs the Filtered-X LMS algorithm, the phase characteristics of the modeled transfer function used in the calculation and the actual phase characteristics may be affected by factors such as changes in the sound propagation environment accompanying temperature changes. When the absolute value of the phase difference exceeds π / 2, the operation becomes unstable and diverges.

  Furthermore, a so-called blowing sound that is generated when wind is applied to a microphone that is an error sensor, and a percussive sound that is generated when the microphone is struck become disturbances that disturb the operation of the control system. That is, when such a disturbance occurs, the control system performs a control operation to mute the sound including the disturbance, so that the control sound becomes loud and may be abnormal.

  For this reason, there is a need for a technique that can effectively suppress the generation of abnormal noise or explosion during active noise control. Meeting this requirement is one of the problems to be solved by the present invention.

  The present invention has been made in view of the above circumstances, and an object thereof is to provide a new active noise control device and an active noise control method capable of effectively suppressing the generation of abnormal noise or explosion. And

  The invention according to claim 1 is a transfer function of a path from the output of the control sound signal by the adaptive notch filter unit to the input of the error signal obtained from the sound collection result at a predetermined position to the adaptive notch filter unit, A determination signal generation unit that generates a determination signal based on at least one of a cosine value and a sine value of a phase difference between the path and the modeled transfer function; based on the determination signal, And an output characteristic control unit that controls the output characteristic.

  The invention according to claim 11 is an active noise control method used in an active noise control device including a determination signal generation unit and an output characteristic control unit, wherein the determination signal generation unit is adaptive. A transfer function of a path from the output of the control sound signal from the notch filter unit to the input of the error signal obtained from the sound collection result at a predetermined position to the adaptive notch filter unit, and a modeled transfer function of the path A determination signal generation step for generating a determination signal based on at least one of a cosine value and a sine value of a phase difference between the output signal and the output characteristic control unit based on the determination signal; An active noise control method comprising: an output characteristic control step for controlling the noise.

  A twelfth aspect of the present invention is an active noise control program that causes an arithmetic means to execute the active noise control method according to the eleventh aspect.

  A thirteenth aspect of the present invention is a recording medium in which the active noise control program according to the twelfth aspect of the present invention is recorded so as to be readable by the calculation means.

1 is a block diagram schematically showing a configuration of an active noise control device according to an embodiment of the present invention. It is a block diagram which shows the structure of the adaptive notch filter part of FIG. It is a block diagram which shows the structure of the signal generation part for determination of FIG. 3 is a flowchart for explaining a mute control process by an output characteristic control unit of FIG. 1. It is a flowchart for demonstrating the disturbance monitoring process of FIG. 3 is a flowchart for explaining phase correction control processing by an output characteristic control unit in FIG. 1.

  Hereinafter, an embodiment of the present invention will be described with reference to FIGS. Hereinafter, an active noise control device that silences engine sound leaking into the vehicle interior by noise control using the Filtered-X LMS algorithm will be described as an example. In the following description and drawings, the same or equivalent elements are denoted by the same reference numerals, and redundant description is omitted.

[Constitution]
FIG. 1 is a block diagram showing a schematic configuration of an active noise control (ANC) apparatus 100 according to an embodiment. As shown in FIG. 1, the active noise control device 100 includes a frequency detection unit 110, a sine wave generation unit 115, a mute processing unit 120, and a π / 2 delay unit 125. The active noise control apparatus 100 includes a reference signal generation unit 130, an adaptive notch filter unit 140, and a sound output unit 150. Furthermore, the active noise control apparatus 100 includes a sound collection unit 160, a determination signal generation unit 170, and an output characteristic control unit 180.

  The frequency detection unit 110 receives an engine rotation signal that is sent from a rotation detection sensor (not shown) that detects rotation of the engine as a noise source and is input to the input port IPT. Then, the frequency detection unit 110 detects the frequency of the engine rotation signal, converts the detection result into the engine rotation frequency, and then sends it to the sine wave generation unit 115 as the detection angular frequency DQ.

  In the present embodiment, the engine rotation signal is an analog signal. The frequency detection unit 110 detects the frequency of the engine rotation signal by digital signal processing after the analog signal is converted into a digital signal by a built-in AD (Analogue to Digital) converter. In the present embodiment, signal processing after the frequency detection unit 110 and before generation of a control sound signal y described later is performed by digital signal processing.

The sine wave generator 115 receives the detection angular frequency DQ sent from the frequency detector 110. Then, the sine wave generator 115 generates a sine wave signal X represented by the following equation (1) based on the detected angular frequency DQ, and sends the generated sine wave signal X to the mute processing unit 120. send.
X (t) = sin (ω · t) (1)

  Note that the angular frequency ω is any angular frequency that is a natural number multiple of the detected angular frequency DQ that is the angular frequency of the component having a large level in the engine sound when the engine rotates at the detected angular frequency DQ. That is, the angular frequency obtained by multiplying the natural number selected in advance by the detected angular frequency DQ is the angular frequency ω.

The mute processing unit 120 receives the sine wave signal X sent from the sine wave generation unit 115. Then, the mute processing unit 120 performs soft mute processing on the sine wave signal X according to the mute control designation MC sent from the output characteristic control unit 180, and generates a signal x0 represented by the following equation (2). The generated signal x0 is sent to the π / 2 delay unit 125, the reference signal generation unit 130, and the adaptive notch filter unit 140.
_{x0 (t) = G M ·} sin (ωt) ... (2)
Here, the value G _M is the mute factor.

The mute processing section 120, when the mute is designated by the mute control designation MC may attack time is short soft mute processing, i.e., to perform the soft mute processing abruptly mute coefficient G _M decreases. Also, the mute processing section 120, when the unmute is designated by the mute control designation MC is relatively long soft muting recovery time, i.e., to perform the soft mute processing gently mute coefficient G _M increases .

In the period in which the mute processing in the mute processing section 120 is not performed, the value G _M is "1", the signal x0 is adapted to a sine wave signal X itself.

The π / 2 delay unit 125 receives the signal x0 sent from the mute processing unit 120. The π / 2 delay unit 125 delays the phase of the signal x0 by π / 2 to generate the signal x1 represented by the following equation (3), and the generated signal x1 is referred to as the reference signal generation unit 130 and This is sent to the adaptive notch filter unit 140.
_{x1 (t) = - G M} · cos (ωt) ... (3)

  Note that the signal x0 and the signal x1 described above are reference signals for the adaptive notch filter unit 140.

The reference signal generator 130 receives the signal x0 sent from the mute processor 120 and the signal x1 sent from the π / 2 delay unit 125. In the reference signal generation unit 130, the phase correction units 131 and 132 perform phase correction of the signals x0 and x1 under the control of the output characteristic control unit 180, and the following equations (4) and (5) The reference signals r0 and r1 that can be expressed as follows are generated, and the generated reference signals r0 and r1 are sent to the adaptive notch filter unit 140 and the determination signal generation unit 170.
_{r0 (t) = G M ·} sin (ωt + Δθ) ... (4)
_{r1 (t) = - G M} · cos (ωt + Δθ) ... (5)

  Here, the phase Δθ is a transfer function C (gain characteristic: G, phase characteristic: dθ (hereinafter, simply referred to as “phase dθ”) from the output of the adaptive notch filter unit 140 to the error signal input of the adaptive notch filter unit 140. The phase characteristics of the transfer function C ^ modeled as

The adaptive notch filter unit 140 includes a signal x0 sent from the mute processing unit 120, a signal x1 sent from the π / 2 delay unit 125, reference signals r0 and r1 sent from the reference signal generation unit 130, and The error signal e sent from the sound collection unit 160 is received. Then, the adaptive notch filter unit 140 generates a control sound signal y represented by the following equation (6) using the LMS algorithm based on the signals x0 and x1, the reference signals r0 and r1, and the error signal e. The generated control sound signal y is sent to the sound output unit 150 and the determination signal generation unit 170.
y (t) = A · sin (ωt + θ) (6)
That is, the transfer function of the adaptive notch filter unit 140 has a gain A _F (= A / G _M ) and a phase characteristic θ (also simply referred to as “phase θ”).

Here, the gain A _F is set such that a predetermined maximum gain value A _MAX is an upper limit value. The maximum gain value A _MAX, like the product of the minimum value of the mute factor _{G M (= A MAX · G} M) is sufficiently smaller than 1, it is predetermined.

  The configuration of the adaptive notch filter unit 140 will be described later.

  The sound output unit 150 includes a DA (Digital to Analogue) conversion unit, a power amplifier, and a speaker. The sound output unit 150 receives the control sound signal y sent from the adaptive notch filter unit 140. In the sound output unit 150 that receives the control sound signal y, the DA conversion unit converts the control sound signal y into an analog signal, and supplies the converted analog signal to the speaker via the power amplification unit. As a result, a control sound is output from the speaker.

  The sound collection unit 160 includes a microphone and an AD conversion unit that are installed at predetermined positions in the vehicle interior. The sound collection result by the microphone in the sound collection unit 160 is converted into a digital signal by the AD conversion unit. The digital signal thus converted is sent to the adaptive notch filter unit 140 and the determination signal generation unit 170 as an error signal e.

Here, since the characteristics of the transfer function C from the output of the adaptive notch filter unit 140 to the error signal input of the adaptive notch filter unit 140 are the gain G and the phase dθ as described above, the error signal e is given by It is represented by the formula (7).
e (t) = A · sin (ωt + θ + dθ)) (7)
When the target sound is present at a predetermined position where the microphone is installed, the result of the adaptive operation is reflected as changes in the gain A _F (= A / G _M ) and the phase θ of the adaptive notch filter unit 140. The

  The determination signal generation unit 170 includes reference signals r0 and r1 sent from the reference signal generation unit 130, filter coefficients W0 and W1 and control sound signal y sent from the adaptive notch filter unit 140, and a sound collection unit. The error signal e sent from 160 is received. Based on the reference signals r0 and r1, the filter coefficients W0 and W1, the control sound signal y, and the error signal e, the determination signal generation unit 170 determines whether the adaptive notch filter unit 140 divides adaptive control and introduces disturbance. And a second determination signal SV for determining the phase difference δθ (= Δθ−dθ) between the phase Δθ and the phase dθ described above. The first determination signal CV and the second determination signal SV thus generated are sent to the output characteristic control unit 180.

  The configuration of the determination signal generation unit 170 will be described later.

  The output characteristic control unit 180 receives the first determination signal CV and the second determination signal SV sent from the determination signal generation unit 170. Then, the output characteristic control unit 180 generates a mute control designation MC based on the first determination signal CV, and sends the generated mute control designation MC to the mute processing unit 120. Further, the output characteristic control unit 180 generates a phase correction specification PC based on the second determination signal SV, and sends the generated phase correction specification PC to the reference signal generation unit 130.

  The mute control process and the phase correction control process, which are output characteristic control processes by the output characteristic control unit 180, will be described later.

  Next, the configuration of the adaptive notch filter unit 140 will be described. As illustrated in FIG. 2, the adaptive notch filter unit 140 includes filter coefficient update units 141 and 142, adaptive filter units 143 and 144, and an addition unit 145.

  The filter coefficient updating unit 141 receives the reference signal r0 sent from the reference signal generation unit 130 and the error signal e sent from the sound collection unit 160. Then, the filter coefficient update unit 141 generates a filter coefficient W0 using an LMS algorithm that uses an instantaneous value of the mean square error based on the reference signal r0 and the error signal e, and the generated filter coefficient W0 is an adaptive filter unit. 143 and the determination signal generator 170.

  The filter coefficient updating unit 142 receives the reference signal r 1 sent from the reference signal generating unit 130 and the error signal e sent from the sound collecting unit 160. Then, the filter coefficient update unit 142 generates a filter coefficient W1 based on the reference signal r1 and the error signal e using an LMS algorithm that uses an instantaneous value of the mean square error, and the generated filter coefficient W1 is an adaptive filter unit. 144 and the determination signal generator 170.

Note that W0 generated by the filter coefficient updating unit 141 is expressed by the following equation (8), and W1 generated by the filter coefficient updating unit 142 is expressed by the following equation (9). .
W0 (t) = A _F · cos θ (8)
W1 (t) = − A _F · sin θ (9)

  The adaptive filter unit 143 receives the signal x0 sent from the mute processing unit 120 and the filter coefficient W0 sent from the filter coefficient update unit 141. Then, adaptive filter section 143 multiplies signal x0 and filter coefficient W0 and sends the multiplication result to addition section 145.

  The adaptive filter unit 144 receives the signal x1 sent from the π / 2 delay unit 125 and the filter coefficient W1 sent from the filter coefficient update unit 142. Then, adaptive filter section 144 multiplies signal x1 and filter coefficient W1 and sends the multiplication result to addition section 145.

The adder 145 receives the multiplication result sent from the adaptive filter unit 143 and the multiplication result sent from the adaptive filter unit 144. Then, the adder 145 adds the multiplication results of both to generate the control sound signal y. That is, the adding unit 145 generates the control sound signal y (see the above-described equation (6)) by performing the calculation of the following equation (10).
x0 (t) · W0 (t) + x1 (t) · W1 (t)
= G _M · A _F · (sin (ωt) · cos θ + cos (ωt) · sin θ)
= A · sin (ωt + θ) = y (t) (10)
The control sound signal y generated in this way is sent to the sound output unit 150 and the determination signal generation unit 170.

  Next, the configuration of the determination signal generation unit 170 will be described. As shown in FIG. 3, the determination signal generation unit 170 includes a level detection unit 210, a first generation unit 220, and a second generation unit 230.

  The level detection unit 210 receives the control sound signal y sent from the adaptive notch filter unit 140. Then, the level detection unit 210 detects the amplitude value A of the control sound signal y (see the above formula (6) or (10)). The amplitude value A that is the detection result is sent to the first generation unit 220 and the second generation unit 230.

  The first generation unit 220 includes the reference signals r0 and r1 sent from the reference signal generation unit 130, the filter coefficients W0 and W1 sent from the adaptive notch filter unit 140, and the error signal e sent from the sound collection unit 160. , And the amplitude value A sent from the level detector 210 is received. Then, the first generation unit 220 generates the first determination signal CV based on the reference signals r0 and r1, the filter coefficients W0 and W1, the error signal e, and the amplitude value A.

  The first generation unit 220 having such a function includes multiplication units 221 and 222 and an addition unit 223. The first generation unit 220 includes a division unit 224 and a smoothing unit 225.

  The multiplication unit 221 receives the reference signal r0 sent from the reference signal generation unit 130, the filter coefficient W0 sent from the adaptive notch filter unit 140, and the error signal e sent from the sound collection unit 160. Then, the multiplication unit 221 multiplies the reference signal r0, the filter coefficient W0, and the error signal e, and sends the multiplication result to the addition unit 223.

  The multiplication unit 222 receives the reference signal r1 sent from the reference signal generation unit 130, the filter coefficient W1 sent from the adaptive notch filter unit 140, and the error signal e sent from the sound collection unit 160. Then, the multiplication unit 222 multiplies the reference signal r1, the filter coefficient W1, and the error signal e, and sends the multiplication result to the addition unit 223.

The adder 223 receives the multiplication result sent from the multiplier 221 and the multiplication result sent from the multiplier 222. Then, the adding unit 223 adds these multiplication results to generate a signal u1. The signal u1 generated in this way is expressed by the following equation (11).
u1 (t) = (r0 (t) .W0 (t) + r1 (t) .W1 (t)). e (t)
= A · (sin (ωt + Δθ + θ)) · e (t) (11)
The signal u1 generated in this way is sent to the division unit 224.

The division unit 224 receives the signal u1 sent from the addition unit 223 at the input terminal I1, and receives the amplitude value A sent from the level detection unit 210 at the input terminal I2. Then, the division unit 224 generates a signal u2 expressed by the following equation (12), and sends the generated signal u2 to the smoothing unit 225.
u2 (t) = u1 (t) / A
= (Sin (ωt + Δθ + θ)) · e (t) (12)

  The smoothing unit 225 receives the signal u2 sent from the division unit 224. Then, the smoothing unit 225 performs the time averaging process on the signal u2 over time (π / ω), thereby smoothing the signal u2 and generating the first determination signal CV. The generated first determination signal CV is sent to the output characteristic control unit 180.

When the smoothing unit 225 smoothes the signal u2 as described above, since the components other than the DC component in the signal u2 are substantially zero, the first determination signal CV is substantially equal to the DC component in the signal u2. That is, the first determination signal CV can be expressed by the following equations (13) and (14).
CV (t) ≈K · cos (Δθ−dθ) = K · cos δ (13)
K = A · G / 2> 0 (14)

  By the way, in this embodiment, when the disturbance is not mixed in the error signal e, the change in the control target sound is gentle compared to the control performance, and therefore the control target sound is constant locally. I can say that. For this reason, the mute control in the present embodiment performs adaptive control that increases the amplitude value A when the gain G decreases due to a sound field dip or the like, and makes the control sound at the microphone position constant, that is, The value (A · G) is constant. For this reason, it can be said that the first determination signal CV has the dependency on the amplitude value A of the control sound signal y removed when disturbance is not mixed in the error signal e.

  The second generation unit 230 includes reference signals r0 and r1 sent from the reference signal generation unit 130, filter coefficients W0 and W1 sent from the adaptive notch filter unit 140, and an error signal e sent from the sound collection unit 160. , And the amplitude value A sent from the level detector 210 is received. Then, the second generation unit 230 generates the second determination signal SV based on the reference signals r0 and r1, the filter coefficients W0 and W1, the error signal e, and the amplitude value A.

  The second generation unit 230 having such a function includes multiplication units 231 and 232 and a subtraction unit 233. In addition, the second generation unit 230 includes a division unit 234 and a smoothing unit 235.

  The multiplication unit 231 receives the reference signal r0 sent from the reference signal generation unit 130, the filter coefficient W1 sent from the adaptive notch filter unit 140, and the error signal e sent from the sound collection unit 160. Then, the multiplication unit 231 multiplies the reference signal r0, the filter coefficient W1, and the error signal e, and sends the multiplication result to the subtraction unit 233.

  The multiplication unit 232 receives the reference signal r 1 sent from the reference signal generation unit 130, the filter coefficient W 0 sent from the adaptive notch filter unit 140, and the error signal e sent from the sound collection unit 160. Then, the multiplication unit 232 multiplies the reference signal r1, the filter coefficient W0, and the error signal e, and sends the multiplication result to the subtraction unit 233.

The subtraction unit 233 receives the multiplication result sent from the multiplication unit 231 at the + input terminal and receives the multiplication result sent from the multiplication unit 232 at the −input terminal. Then, the subtraction unit 233 subtracts the multiplication result sent from the multiplication unit 232 from the multiplication result sent from the multiplication unit 231 to generate the signal v1. The signal v1 generated in this way is expressed by the following equation (15).
v1 (t) = (r0 (t) .W1 (t) -r1 (t) .W0 (t)). e (t)
= −A · (cos (ωt + Δθ + θ)) · e (t) (15)
The signal v1 thus generated is sent to the division unit 234.

The division unit 234 receives the signal v1 sent from the subtraction unit 233 at the input terminal I1, and receives the amplitude value A sent from the level detection unit 210 at the input terminal I2. Then, the division unit 234 generates a signal v2 represented by the following equation (16), and sends the generated signal v2 to the smoothing unit 235.
v2 (t) = v1 (t) / A
= − (Cos (ωt + Δθ + θ)) · e (t) (16)

  The smoothing unit 235 receives the signal v <b> 2 sent from the division unit 234. Then, the smoothing unit 235 smoothes the signal v2 by performing time averaging processing on the signal v2 over time (π / ω), and generates the second determination signal SV. The generated second determination signal SV is sent to the output characteristic control unit 180.

Here, when the signal v2 is smoothed, the components other than the DC component in the signal v2 are substantially zero, so the second determination signal CV is substantially equal to the DC component in the signal v2. That is, the second determination signal SV can be expressed by the following equations (17) and (18).
SV (t) ≈−K · sin (Δθ−dθ) = − K · sin δ (17)
K = A · G / 2> 0 (18)

  By the way, the silencing control in the present embodiment is a control for keeping the value (A · G) constant when the disturbance is not mixed in the error signal e as described above. For this reason, it can be said that the second determination signal SV has the dependency on the amplitude value A of the control sound signal y removed when disturbance is not mixed in the error signal e.

[Operation]
Next, the operation of the active noise control apparatus 100 configured as described above will be described mainly focusing on the control sound output control processing by the output characteristic control unit 180.

  When receiving the engine rotation signal input to the input port IPT, the frequency detection unit 110 detects the frequency of the engine rotation signal. Then, the frequency detection unit 110 converts the detection result into the engine rotation frequency and then sends the detection result to the sine wave generation unit 115 as the detection angular frequency DQ (see FIG. 1).

  The sine wave generator 115 that has received the detection angular frequency DQ generates the sine wave signal X represented by the above (1) based on the detection angular frequency DQ. Then, the sine wave generation unit 115 sends the generated sine wave signal X to the mute processing unit 120 (see FIG. 1).

  Upon receiving the sine wave signal X, the mute processing unit 120 performs soft mute processing on the sine wave signal X in accordance with the mute control designation MC sent from the output characteristic control unit 180, and is expressed by the above equation (2). A signal x0 is generated. The mute processing unit 120 then sends the generated signal x0 to the π / 2 delay unit 125, the reference signal generation unit 130, and the adaptive notch filter unit 140 (see FIG. 1).

  Upon receiving the signal x0, the π / 2 delay unit 125 delays the phase of the signal x0 by π / 2, and generates the signal x1 represented by the above equation (3). Then, the π / 2 delay unit 125 sends the generated signal x1 to the reference signal generation unit 130 and the adaptive notch filter unit 140 (see FIG. 1).

  Receiving the signals x0 and x1, the reference signal generation unit 130 performs phase correction of the signals x0 and x1 under the control of the output characteristic control unit 180, and is expressed by the above equations (4) and (5). The reference signals r0 and r1 are generated. Then, the reference signal generation unit 130 sends the generated reference signals r0 and r1 to the adaptive notch filter unit 140 and the determination signal generation unit 170 (see FIG. 1).

  The adaptive notch filter unit 140 that has received the signals x0, x1 and the reference signals r0, r1 refers to the error signal e received from the sound collection unit 160 at that time, and uses the LMS algorithm to A control sound signal y expressed by the following formula is generated. Then, the adaptive notch filter unit 140 sends the generated control sound signal y to the sound output unit 150 and the determination signal generation unit 170 (see FIGS. 1 and 2).

  The sound output unit 150 that has received the control sound signal y outputs a control sound from the speaker according to the control sound signal y. Note that the sound collection unit 160 transmits the result of sound collection by the microphone in a state where the control sound is output from the speaker to the adaptive notch filter unit 140 and the determination signal generation unit 170 as an error signal e.

  As described above, in parallel with the mute operation by noise control performed by adopting the Filtered-X LMS algorithm, the determination signal generation unit 170 performs the first determination represented by the expression (13) as described above. A signal CV is generated, and a second determination signal SV expressed by equation (17) is generated. Then, the determination signal generation unit 170 sends the generated first determination signal CV and second determination signal SV to the output characteristic control unit 180.

  Upon receiving the first determination signal CV and the second determination signal SV, the output characteristic control unit 180 performs control sound output characteristic control based on the first determination signal CV and the second determination signal SV. Such control sound output characteristic control includes mute control based on the first determination signal CV and phase correction control based on the second determination signal SV.

<Mute control processing>
First, mute control processing by the output characteristic control unit 180 will be described. Initially, the mute processing unit 120 does not perform the mute processing, and the mute processing unit 120 outputs the signal X as the signal x0.

  In the mute control, as shown in FIG. 4, first, in step S <b> 11, the output characteristic control unit 180 determines the value of the first determination signal CV sent from the determination signal generation unit 170 (hereinafter “value”). CV)). Subsequently, in step S12, the output characteristic control unit 180 determines whether or not the value CV is smaller than the first threshold value VT1, so that the absolute value of the phase difference δ in the above-described equation (13) becomes π / 2. It is determined whether there is a possibility that the control by the Filtered-X LMS algorithm employed in the active noise control apparatus 100 may become unstable.

  Here, the sign of the value CV is the cosine value (= cos δ) of the phase difference δ that becomes negative when the absolute value of the phase difference δ exceeds π / 2, as shown by the above-described equations (13) and (14). ). However, since the expression (13) is an approximate expression, the sign of the value CV does not completely match the sign of the cosine value (= cos δ) of the phase difference δ.

  Therefore, in determining whether the absolute value of the phase difference δ at which control by the Filtered-X LMS algorithm becomes unstable may exceed π / 2, in the present embodiment, the first threshold value VT1 is set to a negative value. It is a predetermined value. The first threshold value VT1 is determined in advance based on experiments, simulations, experiences, and the like from the viewpoint that it is possible to determine whether the absolute value of the phase difference δ may exceed π / 2.

  If the result of the determination in step S12 is negative (step S12: N), the process proceeds to step S19 described later. On the other hand, when the result of the determination in step S12 is affirmative (step S12: Y), the process proceeds to step S13.

  In step S13, the output characteristic control unit 180 starts monitoring the duration of the state of “CV <VT1”. Subsequently, in step S14, the output characteristic control unit 180 reads the value CV.

  Next, in step S15, the output characteristic control unit 180 determines whether or not the value CV is smaller than the first threshold value VT1. If the result of this determination is affirmative (step S15: Y), the process proceeds to step S16.

  In step S16, the output characteristic control unit 180 determines whether or not the duration of the state of “CV <VT1” is equal to or longer than the predetermined time TTH. Here, the predetermined time TTH is determined in advance based on experiments, simulations, experiences, and the like from the viewpoint that the possibility that the absolute value of the phase difference δ exceeds π / 2 is almost certain.

  If the result of the determination in step S16 is negative (step S16: N), the process returns to step S14. Thereafter, if the state of “CV <VT1” continues, the processes of steps S14 to S16 are repeated. Then, when the duration of the state of “CV <VT1” is equal to or longer than the predetermined time TTH, the process proceeds to step S17.

In step S <b> 17, the output characteristic control unit 180 sends the mute control designation MC designating the mute setting to the mute processing unit 120. As a result, the mute processing section 120, the attack time is short soft mute processing, i.e., the soft mute processing rapidly mute coefficient G _M decreases is executed.

Here, when the mute factor G _M is reduced by the soft mute processing, the above-described (4), as indicated by (5), the reference signal r0, r1 also becomes attenuated. For this reason, since the update of the filter coefficients W0 and W1 is slow, the control sound signal y is rapidly attenuated. As a result, the volume of the control sound decreases rapidly. As described above, the maximum gain value A _MAX of the adaptive notch filter 140, because the product of the minimum value of the mute factor _{G M (= A MAX · G} M) is sufficiently smaller than 1, muting factor G _{Even if} the gain A _F of the adaptive notch filter unit 140 reaches the maximum gain value A _MAX after _M becomes the minimum value, the volume of the control sound is sufficiently small.

  When step S17 is executed as described above, the output characteristic control unit 180 ends the mute control process. As a result, the volume of the control sound is maintained in a sufficiently small state.

  If the determination result in step S15 described above is negative (step S15: N), the process proceeds to step S18. In step S18, the output characteristic control unit 180 ends the monitoring of the duration of the state of “CV <VT1”. Subsequently, in step S19, the output characteristic control unit 180 executes disturbance monitoring processing. Then, the process returns to step S11. The contents of the disturbance monitoring process in step S19 will be described later.

  By executing the processes of steps S11 to S19 as described above, the output characteristic control unit 180 executes a mute control process.

  Next, the disturbance monitoring process in step S19 will be described. In the disturbance monitoring process, as shown in FIG. 5, first, in step S21, the output characteristic control unit 180 determines whether or not the value CV is larger than a predetermined threshold value VT2 (> 0). It is determined whether or not a disturbance due to a large disturbance sound is mixed in the error signal e. The reason for determining whether the disturbance is mixed by this determination is that the value CV swings positively and negatively with a fluctuation width corresponding to the magnitude of the disturbance sound in synchronization with the disturbance sound.

  Note that the predetermined threshold value VT2 is determined in advance based on experiments, simulations, experiences, and the like from the viewpoint of determining the possibility that the volume of the control sound output for muting the disturbance noise becomes too large.

  If the result of the determination in step S21 is affirmative (step S21: Y), the process proceeds to step S22. In step S22, the output characteristic control unit 180 determines whether or not the mute processing unit 120 has been set to mute. If the result of determination in step S22 is affirmative (step S22: Y), the output characteristic control unit 180 has already taken measures to prevent the volume of the control sound from becoming abnormally high. And the process of step S19 is terminated.

  If the result of the determination in step S22 is negative (step S22: N), the process proceeds to step S23. In step S23, the output characteristic control unit 180 sends a mute control designation MC designating mute setting to the mute processing unit 120 to the mute processing unit 120 in order to prevent the volume of the control sound from increasing abnormally. . Then, the output characteristic control unit 180 ends the process of step S19.

  If the result of the determination in step S21 described above is negative (step S21: N), the process proceeds to step S24. In step S24, the output characteristic control unit 180 determines whether or not the mute processing unit 120 is set to mute. If the result of determination in step S24 is negative (step S24: N), the output characteristic control unit 180 ends the process of step S19.

  If the result of the determination in step S24 is affirmative (step S24: Y), the process proceeds to step S25. In this step S25, the output characteristic control unit 180 determines that the mute processing for preventing the volume of the control sound from becoming abnormally high is not necessary, and sends a mute control signal designating mute release to the mute processing unit 120. The MC is sent to the mute processing unit 120. Then, the output characteristic control unit 180 ends the process of step S19.

When unmute is specified, the mute processing section 120 performs a soft mute processing gently mute coefficient G _M increases. When muting factor G _M increases gradually by soft mute processing, the level of the reference signal r0, r1 gradual recovery, update rate of the filter coefficients W0, W1 also gradually recovered. Further, the levels of the signals x0 and x1 are also gradually recovered. As a result, the control returns to the normal control sound generation process in which the mute process is not performed in a manner that does not give the vehicle passenger a sense of incongruity.

  By performing the processes in steps S21 to S25, the output characteristic control unit 180 performs disturbance monitoring, alignment, and mute setting and mute release processing based on the disturbance monitoring result. Then, when the process of step S19 ends, the process returns to step S11 in FIG. 4 described above.

<Phase correction control processing>
Next, phase correction control processing by the output characteristic control unit 180 will be described.

  In the phase correction control, as shown in FIG. 6, first, in step S <b> 31, the output characteristic control unit 180 sets the value of the second determination signal SV sent from the determination signal generation unit 170 (hereinafter, “ Value SV). Subsequently, in step S32, the output characteristic control unit 180 determines a phase correction amount in the reference signal generation unit 130, that is, a change amount of the phase Δθ (hereinafter referred to as “phase correction change amount”) based on the value SV. To do.

  Here, when there is no disturbance or when the disturbance is small, the value SV is expressed by a constant (−K (<K) in the sine value (= sin δ) of the phase difference δ, as expressed by the equations (17) and (18) described above. 0)). Therefore, when normal control sound output control is performed without performing mute setting by the mute control process described above, the value SV is uniquely determined according to the value of the phase difference δ. It is decided.

  That is, when control for normal control sound output is performed, it can be said that the sign of the value SV is opposite to the sign of the phase difference δ if there is no disturbance or is small. Further, it can be said that the absolute value of the value SV increases as the absolute value of the phase difference δ increases.

  Therefore, in this embodiment, the output characteristic control unit 180 assumes that control for normal control sound output is performed, and controls the control sound output while setting the phase difference δ to 0 as the final target. In order to maintain the stability of the operation, the change amount of the phase Δθ is determined so that the phase difference δ gradually approaches 0. That is, when the value SV is a positive value, a negative phase correction change amount is determined, and when the value SV is a negative value, a positive phase correction change amount is determined.

  Next, in step S <b> 33, the output characteristic control unit 180 specifies the determined phase correction change amount to the reference signal generation unit 130. Receiving the designation of the phase correction change amount, the reference signal generation unit 130 generates reference signals r0 and r1 obtained by changing the phase Δθ by the designated phase change amount.

  When step S33 ends, the process returns to step S31. Thereafter, by repeating the processing of steps S31 to S33, when control for normal control sound output is performed, control for control sound output becomes unstable and causes divergence. Reference signals r0 and r1 for preventing the absolute value of the phase difference δ from exceeding π / 2 can be generated. For this reason, the divergence of the control system accompanying changes in environmental conditions such as temperature is prevented.

  As described above, in the present embodiment, the reference signal generator 130 is based on the modeled transfer function from the output of the adaptive notch filter 140 to the error signal input of the adaptive notch filter 140, based on the reference signal x0. , X1 are corrected to generate reference signals r0, r1. Then, the adaptive notch filter unit 140 generates reference signals r0 and r1 and a control sound signal y based on the reference signals x0 and x1 and the error signal e which is a sound collection result by the sound collection unit 160. In parallel with the generation of the control sound signal y by the adaptive notch filter unit 140, the determination signal generation unit 170 performs the error signal e, the reference signals r0 and r1, and the transfer function (that is, adaptive) of the adaptive notch filter unit 140. Based on the gain characteristic and the phase characteristic θ of the notch filter unit 140), the phase characteristic dθ of the transfer function from the output of the adaptive notch filter unit 140 to the error signal input of the adaptive notch filter unit 140, and the modeled transfer function The first determination signal CV and the second determination signal SV reflecting the phase difference δθ with respect to the phase characteristic Δθ is generated. Then, based on the first determination signal CV and the second determination signal SV, the output characteristic control unit 180 controls the output characteristic of the control sound so as to suppress the occurrence of abnormal noise or explosion.

  Therefore, it is possible to effectively suppress the generation of abnormal noise or explosion during active noise control.

  In the present embodiment, the first determination signal CV is a signal having a signal value reflecting the cosine value of the phase difference δθ, and the magnitude relationship between the value of the first determination signal CV and the first threshold value VT1 is satisfied. Based on this, when it is determined that the cosine value of the phase difference δθ is negative and the absolute value of the phase difference δθ exceeds π / 2, the mute processing unit 120 performs mute processing on the control sound. Let it be done. For this reason, it is possible to detect at an early stage that the control for the control sound output by the Filtered-X LMS algorithm is unstable and diverge, and to mute the control sound. Generation of sound and explosion can be suppressed.

  In the present embodiment, when determining that the absolute value of the phase difference δθ exceeds π / 2, the phase difference δθ is determined from the magnitude relationship between the value of the first determination signal CV and the first threshold value VT1. When the state where the cosine value is estimated to be negative continues for a predetermined time TTH or more, it is determined that the cosine value of the phase difference δθ is negative. For this reason, it can be determined with increased certainty that the cosine value of the phase difference δθ is negative.

  Further, in the present embodiment, when it is determined that the disturbance noise sound is collected by the sound collection unit 160 based on the magnitude relationship between the value of the first determination signal CV and the second threshold value VT2, the mute is performed. The processing unit 120 performs a mute process on the control sound. For this reason, mixing of a large disturbance noise sound can be easily detected, and the volume of the control sound can be prevented from becoming abnormally high due to the silence of the large disturbance noise sound.

  In the present embodiment, the second determination signal SV is a signal having a signal value reflecting the sine value of the phase difference δθ, and the absolute value of the phase difference δθ is based on the value of the second determination signal SV. The phase correction amount Δθ by the reference signal generation unit 130 is controlled so as to decrease. For this reason, it is possible to prevent the absolute value of the phase difference δ that causes divergence due to unstable control for control sound output by the Filtered-X LMS algorithm from exceeding π / 2.

[Modification of Embodiment]
The present invention is not limited to the above-described embodiment, and various modifications are possible.

  For example, in the above embodiment, when the signals u2 and v2 performed when the first and second determination signals CV and SV are generated, time average processing over time (π / ω) is performed. . On the other hand, if the time required for smoothing is within an allowable time for control for control sound output, a time average process may be performed that is an arbitrary natural number times the time (π / ω). .

  Further, in the above embodiment, the signal level of the control sound signal y is detected by the level detection unit 210 that receives the control sound signal y. On the other hand, the signal level of the control sound signal y may be generated based on the filter coefficients W0 and W1.

  In the upper embodiment, the mute control process based on the first determination signal CV and the phase correction control process based on the second determination signal SV are performed in parallel. On the other hand, only one of the mute control process based on the first determination signal CV and the phase correction control process based on the second determination signal SV may be performed.

  In the upper embodiment, as a mute control process based on the first determination signal CV, the value CV and the first threshold value VT1 are compared to suppress the generation of abnormal noise or explosion caused by the divergence of the control system. And the second threshold value VT2 are compared to perform two processes in parallel, that is, a process for preventing an abnormally large control sound from occurring due to the mixing of the disturbance noise sound. I made it. On the other hand, only one of the two processes may be performed.

  In the above embodiment, the present invention is applied to an active noise control device that silences engine sound leaking into the vehicle interior, but is applied to an active noise control device that silences other types of sounds. Of course you can.

  In addition, the active noise control apparatus 100 in the above embodiment is configured as a computer system including a calculation unit including a central processing unit (CPU), a DSP (Digital Signal Processor), and the like, and is prepared in advance. A part or all of the processing in the above embodiment may be executed by executing the program on the computer system. This program is recorded on a computer-readable recording medium such as a hard disk, CD-ROM, or DVD, and is read from the recording medium and executed by the computer. The program may be acquired in a form recorded on a portable recording medium such as a CD-ROM or DVD, or may be acquired in a form distributed via a network such as the Internet. Also good.

DESCRIPTION OF SYMBOLS 100 ... Active noise control apparatus 110 ... Frequency detection part (a part of reference | standard signal production | generation part)
115 ... Sine wave generator (part of the reference signal generator)
120 ... Mute processing unit (also part of the reference signal generation unit)
DESCRIPTION OF SYMBOLS 130 ... Reference signal generation part 140 ... Adaptive notch filter part 150 ... Sound output part 160 ... Sound collection part 170 ... Determination signal generation part 180 ... Output characteristic control part 210 ... Level detection part

Claims (13)

A transfer function of the path from the output of the control sound signal by the adaptive notch filter unit to the input of the error signal obtained from the sound collection result at a predetermined position to the adaptive notch filter unit, and a modeled transfer function of the path A determination signal generation unit that generates a determination signal based on at least one of a cosine value and a sine value of a phase difference between the two;
An output characteristic control unit for controlling the output characteristic of the control sound based on the determination signal;
An active noise control device comprising:   The active noise control apparatus according to claim 1, further comprising a reference signal generation unit that generates a reference signal correlated with the control target sound. A reference signal generation unit configured to generate a reference signal by correcting the phase of the reference signal based on the modeled transfer function;
The determination signal generation unit generates the determination signal based on the error signal, the reference signal, and a transfer function of the adaptive notch filter unit,
The adaptive notch filter unit generates the control sound signal based on the reference signal, the reference signal, and the error signal.
The active noise control apparatus according to claim 2. A mute processing unit for performing a mute process on the control sound;
The determination signal includes a first determination signal having a signal value corresponding to the cosine value of the phase difference,
The output characteristic control unit controls a mute process for the control sound by the mute processing unit based on a value of the first determination signal;
The active noise control device according to claim 3.   When the output characteristic control unit determines that the cosine value of the phase difference is negative based on the magnitude relationship between the value of the first determination signal and a predetermined first threshold value, The active noise control apparatus according to claim 4, wherein the mute processing unit performs a mute process on the control sound.   In the output characteristic control unit, a state in which the cosine value of the phase difference is estimated to be negative from the magnitude relationship between the value of the first determination signal and the first threshold value has continued for a predetermined time or more. The active noise control device according to claim 5, wherein the cosine value of the phase difference is determined to be negative.   When the output characteristic control unit determines that a disturbance noise sound is collected based on the magnitude relationship between the value of the first determination signal and a predetermined second threshold value, the mute processing is performed. The active noise control apparatus according to claim 4, wherein a mute process is performed on the control sound. The determination signal generation unit includes a level detection unit that detects a signal level of the control sound signal,
The determination signal generation unit generates the first determination signal as a signal having a value that does not depend on the signal level of the control sound signal, using a detection result of the level detection unit.
The active noise control device according to any one of claims 4 to 7, wherein The determination signal includes a second determination signal having a signal value corresponding to the sine value of the phase difference,
The output characteristic control unit controls a phase correction amount by the reference signal generation unit based on a value of the second determination signal;
The active noise control device according to any one of claims 3 to 7, wherein The determination signal generation unit includes a level detection unit that detects a signal level of the control sound signal,
The determination signal generation unit generates the second determination signal as a signal having a value that does not depend on the signal level of the control sound signal, using a detection result by the level detection unit.
The active noise control apparatus according to claim 9. An active noise control method used in an active noise control device including a determination signal generation unit and an output characteristic control unit,
The determination signal generation unit includes a transfer function of a path from the output of the control sound signal from the adaptive notch filter unit to the input of the error signal obtained from the sound collection result at a predetermined position to the adaptive notch filter unit, A determination signal generation step for generating a determination signal based on at least one of a cosine value and a sine value of a phase difference between the path and the modeled transfer function;
An output characteristic control step in which the output characteristic control unit controls the output characteristic of the control sound based on the determination signal;
An active noise control method comprising:   An active noise control program for causing an arithmetic means to execute the active noise control method according to claim 11.   13. A recording medium in which the active noise control program according to claim 12 is recorded so as to be readable by an arithmetic means.

Images