JP2013112140A - Active vibration noise control apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an active vibration noise control apparatus in which cancellation control can be executed, with respect to vibration noise of which the frequency characteristics change, while following up that change.SOLUTION: A second adaptive notch filter inputs a reference signal thereto and outputs a second control signal therefrom. A phase angle variation dθ is calculated between a phase angle θ, on a complex plane, of a filter coefficient W2 (Rw2, Iw2) of the second adaptive notch filter and a previous phase angle θold which is calculated in last updating, and a target frequency Fc of the reference signal is changed in accordance with the phase angle variation dθ.

Description

  The present invention relates to an active vibration noise control device that cancels out vibration noise in a passenger compartment generated when a vehicle travels, for example, by output of cancellation vibration noise.

  When the vehicle travels, for example, the vibration of the wheels is transmitted to the vehicle body via the suspension, so that road noise (including vibration and noise. Hereinafter, collectively referred to as “vibration noise”) is generated in the vehicle interior. Occur. In view of this, various active vibration noise control devices have been proposed in which canceling vibration noise having a phase opposite to that of the vibration noise is output from a speaker to cancel the vibration noise.

  For example, in Patent Document 1, a signal component having a predetermined frequency is extracted from an error signal obtained from a microphone provided in a vehicle interior using an adaptive notch filter, and the amplitude and phase of a control signal generated based on the signal component are extracted. There has been proposed an active vibration noise control apparatus for adjusting the noise. Thereby, the amount of calculation processing for canceling vibration noise can be significantly reduced, and the manufacturing cost of the device can be suppressed.

JP 2009-45554 A

  The present invention has been made in connection with the technical idea disclosed in Patent Document 1 described above, and is an active type capable of executing canceling control that follows the change of vibration noise whose frequency characteristics change. An object is to provide a vibration noise control device.

  An active vibration noise control device according to the present invention includes a vibration noise canceling means for outputting a canceling vibration noise based on a canceling signal for the vibration noise, and residual vibration noise due to interference between the vibration noise and the canceling vibration noise as an error signal. An error signal detecting means for detecting, and an active vibration noise control means for receiving the error signal and generating the canceling signal, wherein the active vibration noise control means outputs a reference signal of a predetermined frequency. A reference signal generating means for generating, an adaptive notch filter for inputting the reference signal and outputting a control signal for generation of the cancellation signal, and an amplitude or phase adjustment value corresponding to the frequency of the reference signal are stored. And an amplitude / phase adjusting means for generating the cancellation signal by adjusting the amplitude or phase of the control signal, and a correction error signal by subtracting the control signal from the error signal. A correction error signal generating means for generating the filter coefficient update means for sequentially updating the filter coefficient of the adaptive notch filter based on the reference signal and the correction error signal so that the correction error signal is minimized, A correction cancellation signal generation unit that generates a correction cancellation signal by correcting the cancellation signal based on a transfer characteristic from the vibration noise cancellation unit to the error signal detection unit; and subtracts the correction cancellation signal from the error signal. A non-control-time detection signal generating means for generating a non-control-time detection signal, and a filter coefficient defined on a complex plane, wherein the reference signal is input and a second control signal is output. A notch filter, second correction error signal generating means for generating a second correction error signal by subtracting the second control signal from the non-control detection signal, and the reference signal Second filter coefficient updating means for sequentially updating the filter coefficient of the second adaptive notch filter based on the second correction error signal so that the second correction error signal is minimized, and the second adaptive notch filter. The phase angle change amount between the phase angle of the filter coefficient on the complex plane and the phase angle calculated at the previous update is calculated, and the frequency of the reference signal is calculated according to the phase angle change amount. Frequency switching means for switching.

  In this way, the phase angle change amount between the phase angle on the complex plane of the filter coefficient of the second adaptive notch filter and the phase angle calculated at the previous update is calculated, and the phase angle change amount is calculated as the phase angle change amount. Since the frequency switching means is provided to switch the frequency of the reference signal accordingly, it is possible to monitor the amount of change of the phase angle on the complex plane of the filter coefficient one after another. And accurately grasp. Thus, even if there is a change in the frequency characteristics of the vibration noise, the vibration noise cancellation control that follows the change can be executed.

  The frequency switching means calculates a frequency change amount based on the sampling period of the error signal and the phase angle change amount, and the frequency of the reference signal when the frequency change amount falls below a lower threshold value. Is preferably maintained. Thereby, when the amount of frequency change falls below the lower limit threshold, it is possible to suppress the occurrence of different noise associated with frequency switching.

  Furthermore, it is preferable that the frequency switching means maintains the frequency of the reference signal when the amount of frequency change exceeds an upper threshold value that is greater than the lower threshold value. Thereby, when the amount of frequency change exceeds the upper limit threshold, it is possible to suppress the occurrence of different noise due to excessive control.

  Furthermore, it is preferable to further include an amplitude phase switching unit that switches the adjustment value stored in the amplitude phase adjustment unit in response to the frequency switching unit switching the frequency of the reference signal. As a result, the state in which the frequency of the reference signal is switched can be immediately reflected in the cancellation signal, and the control followability is further improved.

  According to the active vibration noise control device of the present invention, the phase angle change amount between the phase angle on the complex plane of the filter coefficient of the second adaptive notch filter and the phase angle calculated at the previous update. And a frequency switching means for switching the frequency of the reference signal according to the amount of change in the phase angle is provided, so that the amount of change in the phase angle on the complex plane of the filter coefficient can be monitored successively. The trend of change in frequency characteristics can be easily and accurately grasped from the quantity. Thereby, the cancellation control which followed the change with respect to the vibration noise from which a frequency characteristic changes can be performed.

1 is a schematic block diagram of an active vibration noise control apparatus according to the present embodiment. It is a flowchart with which operation | movement description of the ANC apparatus shown in FIG. 1 is provided. It is a 1st detailed block diagram of the active vibration noise control part shown in FIG. It is a 2nd detailed block diagram of the active vibration noise control part shown in FIG. It is a detailed flowchart for demonstrating the update method of the object frequency in step S11 of FIG. It is a schematic explanatory drawing showing the calculation method of the phase angle variation | change_quantity in the complex space of the filter coefficient of a 2nd adaptive notch filter. FIG. 7A is a spectrum diagram of an error signal before execution of ANC control. FIG. 7B is a frequency characteristic diagram of a SAN-type bandpass filter suitable for the error signal shown in FIG. 7A. FIG. 7C is a frequency characteristic diagram of the first adaptive notch filter corresponding to the frequency characteristic of the SAN-type bandpass filter shown in FIG. 7B. FIG. 8A is a spectrum diagram of an error signal in which the frequency characteristic shown in FIG. 7A is changed. FIG. 8B is a frequency characteristic diagram of a SAN bandpass filter suitable for the error signal shown in FIG. 8A. FIG. 9A is a frequency characteristic diagram of the first adaptive notch filter when the frequency switching process is not performed. FIG. 9B is a frequency characteristic diagram showing the sensitivity function of the ANC device when the frequency switching process is not performed. FIG. 9C is a spectrum diagram of the error signal after the ANC control is executed when the frequency switching process is not performed. FIG. 10A is a frequency characteristic diagram of the first adaptive notch filter when the frequency switching process is performed. FIG. 10B is a frequency characteristic diagram showing the sensitivity function of the ANC device when the frequency switching process is performed. FIG. 10C is a spectrum diagram of an error signal after execution of ANC control when frequency switching processing is performed.

  Preferred embodiments of the active vibration noise control device according to the present invention will be described below with reference to the accompanying drawings.

  As shown in FIG. 1, an active vibration noise control device {hereinafter referred to as an ANC (Adaptive Noise Control) device 10. } Is mounted on the vehicle 11. The ANC apparatus 10 includes an active vibration noise control unit 14 (active vibration noise control means), a microphone 16 (error signal detection means), and a speaker 18 (vibration noise canceling means).

  The microphone 16 inputs various sounds generated inside and outside the vehicle 11. The input sound includes vibration noise NS caused by wheel vibration received from the road surface 12 and canceling vibration noise CS for canceling the vibration noise NS. In this case, the microphone 16 detects residual vibration noise due to interference between the vibration noise NS and the canceling vibration noise CS as an input signal (hereinafter referred to as an error signal A) to the active vibration noise control unit 14. In the illustrated example, the microphone 16 is provided above the passenger compartment 13 of the vehicle 11 (specifically, in the vicinity of an unillustrated occupant's listening point).

  Note that “vibration noise” in this specification refers to all elastic waves propagating through an elastic body. That is, it is not limited to a narrow meaning like an audible sound (an elastic wave having an audible frequency and propagating in the air). For example, when detecting vibration, a vibration sensor or the like may be used instead of the microphone 16.

  The speaker 18 outputs canceling vibration noise CS based on an output signal from the active vibration noise control unit 14 (hereinafter referred to as canceling signal B). Specifically, the speaker 18 outputs a canceling vibration noise CS having a phase opposite to that of the vibration noise NS having a predetermined frequency as a main component, thereby suppressing the degree of generation of the vibration noise NS due to the wave interference effect. Let In the illustrated example, the speaker 18 is provided in the vicinity of the kick panel around the seat in the passenger compartment 13.

  The active vibration noise control unit 14 performs predetermined signal processing on the input error signal A to obtain a cancellation signal B, and then outputs the cancellation vibration noise CS via the speaker 18, thereby generating vibration noise. Control to actively cancel NS (hereinafter referred to as “ANC control”) is performed. The active vibration noise control unit 14 is configured by a microcomputer, a DSP (Digital Signal Processor), or the like. Various processes can be realized by a CPU (Central Processing Unit) executing programs stored in a memory such as a ROM based on input of various signals.

  The active vibration noise control unit 14 is set by a frequency setting unit 20 that sets a predetermined frequency (hereinafter referred to as a target frequency Fc) that is an object of ANC control from an arbitrary frequency band, and the frequency setting unit 20. A reference signal generator 22 (reference signal generator) that generates a reference signal X whose main component is the target frequency Fc, and a SAN (Single Adaptive Notch) filter for the reference signal X generated by the reference signal generator 22 And a first adaptive notch filter 24 (adaptive notch filter) that obtains a first control signal O1 (control signal).

  The active vibration noise control unit 14 subtracts the first control signal O1 output via the first adaptive notch filter 24 from the error signal A input via the microphone 16 to obtain a first correction error signal E1 ( Based on the subtractor 26 (correction error signal generation means) for obtaining the correction error signal), the reference signal X generated by the reference signal generator 22, and the first correction error signal E1 output from the subtractor 26. And a first filter coefficient updating unit (filter coefficient updating means) that sequentially updates the filter coefficient W1 of the first adaptive notch filter 24 so that the one correction error signal E1 is minimized. Note that a SAN type band pass filter 30 is configured by combining the first adaptive notch filter 24 and the subtractor 26. That is, the first correction error signal E1 corresponds to a signal obtained by removing a frequency component having a predetermined width centered on the target frequency Fc from the frequency components over a relatively wide band included in the error signal A.

  The active vibration noise control unit 14 includes an amplitude phase adjustment unit 36 (amplitude phase adjustment unit) that generates the cancellation signal B by adjusting the amplitude or phase of the first control signal O1 using the amplitude or phase adjustment value. The correction phase signal is generated by correcting the canceling signal B based on the transfer characteristic from the speaker 18 to the microphone 16 and the amplitude phase switching unit 38 (amplitude phase switching means) that switches the adjustment value according to the target frequency Fc. A correction cancellation signal generation unit 40 (correction cancellation signal generation means) to perform, and a subtractor 42 (non-control detection signal generation means) that generates a non-control detection signal Awoc by subtracting the correction cancellation signal Ab from the error signal A And further comprising.

  The active vibration noise control unit 14 applies a SAN filter to the reference signal X generated by the reference signal generation unit 22 to obtain a second control signal O2, and a subtractor 42. A subtractor 46 (second correction error signal generation) that subtracts the second control signal O2 input through the second adaptive notch filter 44 from the input non-control detection signal Awoc to obtain a second correction error signal E2. And the second adaptive notch filter 44 so that the second correction error signal E2 is minimized based on the reference signal X from the reference signal generator 22 and the second correction error signal E2 from the subtractor 46. A second filter coefficient update unit 48 (second filter coefficient update means) that sequentially updates the coefficient W2. The SAN type band pass filter 50 is configured by combining the second adaptive notch filter 44 and the subtractor 46. That is, the second correction error signal E2 corresponds to a signal obtained by removing a frequency component having a predetermined width centered on the target frequency Fc from each frequency component over a relatively wide band included in the non-control detection signal Awoc. .

  The active vibration noise control unit 14 is based on the filter coefficient holding unit 52 that holds the filter coefficient W2 successively updated by the second filter coefficient updating unit 48, and the filter coefficient W2 supplied from the filter coefficient holding unit 52. And a frequency switching unit 54 (frequency switching means) that determines whether or not the target frequency Fc needs to be updated or an update amount thereof.

  The ANC device 10 according to the present embodiment is basically configured as described above. The operation of this apparatus will be described below with reference to the flowchart of FIG. By the way, the reference signal X, the filter coefficient W1, and the filter coefficient W2 shown in FIG. 1 are defined on the complex plane and have components of a real part and an imaginary part, respectively. Steps S1 to S8 in FIG. 2 will be described in detail while paying attention to the signal processing flow of the real part component and the imaginary part component with reference to the first detailed block diagram in FIG. For convenience of illustration, FIG. 3 omits some of the components shown in FIG.

  In step S <b> 1, the microphone 16 detects residual vibration noise in the passenger compartment 13 and inputs it as an error signal A. The error signal A includes not only the vibration noise NS described above but also the cancellation vibration noise CS output from the speaker 18 for canceling the vibration noise NS.

  In step S2, the reference signal generator 22 generates a reference signal X whose main component is the target frequency Fc. Prior to generation of the reference signal X, the frequency setting unit 20 sets a frequency to be subjected to ANC control (that is, a target frequency Fc). For example, the frequency setting unit 20 may be configured such that the control target range is 50 Hz to 300 Hz and can be set at 1 Hz intervals. Thereafter, the frequency setting unit 20 drives and controls the reference signal generation unit 22 in accordance with the set target frequency Fc.

  The reference signal generator 22 corresponds to the real part reference signal generator 60 that generates the real part reference signal Rx {= cos (2πFc · t)} corresponding to the real part of the reference signal X, and the imaginary part of the reference signal X. And an imaginary part reference signal generation unit 62 that generates an imaginary part reference signal Ix {= sin (2πFc · t)}. In this case, the reference signal X is expressed by a trigonometric function with respect to time (t), that is, X (t) = exp (i2πFc · t). Note that i is an imaginary unit.

  In step S <b> 3, the first adaptive notch filter 24 generates the control signal O to be supplied to the subtractor 26 side and the amplitude phase adjustment unit 36 side based on the reference signal X from the reference signal generation unit 22. Hereinafter, a specific configuration and operation of the first adaptive notch filter 24 will be described.

  The first adaptive notch filter 24 includes a first filter 64 whose real part filter coefficient Rw1 is variably set, a second filter 66 whose imaginary part filter coefficient Iw1 is variably set, and an output signal on the first filter 64 side. And a subtractor 68 for subtracting the output signal on the second filter 66 side. The first filter 64 scales the amplitude component of the real part reference signal Rx (cosine wave signal) input from the real part reference signal generation unit 60 side by Rw times and outputs it to the subtracter 68 side. The second filter 66 scales the amplitude component of the imaginary part reference signal Ix (sine wave signal) input from the imaginary part reference signal generation unit 62 by Iw times, and outputs it to the subtractor 68 side. Thereafter, the subtracter 68 subtracts the output signal (= Iw1 · Ix) on the second filter 66 side from the output signal (= Rw1 · Rx) on the first filter 64 side. As a result, the first adaptive notch filter 24 outputs the first control signal O1 (= Rw1 · Rx−Iw1 · Ix).

  In step S4, the subtractor 26 uses the first control signal O1 (see step S3) inputted through the first adaptive notch filter 24 from the error signal A (see step S1) inputted through the microphone 16. The first correction error signal E1 is generated by subtraction. In this case, the first correction error signal E1 from which a frequency component having a predetermined width centered on the target frequency Fc is removed is obtained by the action of the SAN type bandpass filter 30.

  In step S5, the first filter coefficient update unit 28 updates the filter coefficient W1 of the first adaptive notch filter 24. Hereinafter, a specific configuration and operation of the first adaptive notch filter 24 will be described.

  The first filter coefficient updating unit 28 corresponds to the real part multiplier 70 and the gain adjuster 72 that are used for updating the real part filter coefficient Rw1 corresponding to the real part of the filter coefficient W1, and the imaginary part of the filter coefficient W1. An imaginary part multiplier 74 and a gain adjuster 76 are provided for updating the imaginary part filter coefficient Iw1. That is, in the present configuration example, the first filter coefficient updating unit 28 updates the filter coefficient W1, that is, the real part filter coefficient Rw1 and the imaginary part filter coefficient Iw1, according to an LMS (Least Mean Square) algorithm. The update algorithm is not limited to this method, and various optimization methods may be adopted.

The real part multiplier 70 multiplies the real part reference signal Rx input from the real part reference signal generation part 60 side by the first correction error signal E1 input from the subtractor 26 side, and supplies it to the gain adjuster 72 side. Output. The gain adjuster 72 multiplies the amplitude component of the multiplication signal by μ1 and outputs it to the first filter 64 side. Here, the constant μ1 corresponds to a step size parameter. Then, the first filter 64 adds the update amount (= + μ1 · Rx · E1) acquired from the first filter coefficient update unit 28 to the real filter coefficient Rw1 at the current time, so that a new real part filter is obtained. A coefficient Rw1 is obtained. That is, the real part filter coefficient Rw1 is updated according to the following equation (1).
Rw1 ← Rw1 + μ1, Rx, E1 (1)

On the other hand, the imaginary part multiplier 74 multiplies the imaginary part reference signal Ix input from the imaginary part reference signal generation part 62 side by the first correction error signal E1 input from the subtractor 26 side, and gain adjuster 76. Output to the side. The gain adjuster 76 multiplies the amplitude component of the multiplication signal by μ1, inverts the phase (adjusts by π), and outputs the result to the second filter 66 side. After that, the second filter 66 adds the update amount (= −μ1 · Ix · E1) acquired from the first filter coefficient update unit 28 to the imaginary part filter coefficient Iw1 at the current time, thereby creating a new imaginary part. A filter coefficient Iw1 is obtained. That is, the imaginary part filter coefficient Iw1 is updated according to the following equation (2).
Iw1 ← Iw1-μ1, Ix, E1 (2)

  In step S <b> 6, the amplitude / phase adjustment unit 36 generates the cancellation signal B by adjusting the amplitude and / or phase of the first control signal O <b> 1 input from the first adaptive notch filter 24 side.

  The amplitude phase adjustment unit 36 uses an amplitude adjuster 78 that adjusts the amplitude of the first control signal O1 using a first adjustment value Gfb that is a parameter for adjusting the amplitude, and a second adjustment value θfb that is a parameter for adjusting the phase. For adjusting the phase of the first control signal O 1, a first storage unit 82 for storing the first adjustment value Gfb to be supplied to the amplitude adjuster 78, and a second to be supplied to the phase adjuster 80. And a second storage unit 84 that stores the adjustment value θfb. That is, the first control signal O1 is adjusted in amplitude by the amplitude adjuster 78, adjusted in phase by the phase adjuster 80, and then supplied to the speaker 18 as the canceling signal B.

  In consideration of the additivity of the trigonometric function, the calculation result obtained by adjusting the amplitude or phase of the first control signal O1 is calculated by adjusting the amplitude or phase of the real part reference signal Rx and the imaginary part reference signal Ix, respectively. Match the result. Therefore, the amplitude phase adjustment unit 36 acquires the real part reference signal Rx and the imaginary part reference signal Ix from the reference signal generation unit 22, adjusts the amplitude or phase of these signals, and then combines them to cancel the signal B May be generated.

  In addition, the amplitude phase switching unit 38 responds to the fact that the frequency switching unit 54 has switched the target frequency Fc of the reference signal X (details will be described later), so that the amplitude phase adjusting unit 36 (the first storage unit 82, the second storage unit 82). The adjustment values (first adjustment value Gfb, second adjustment value θfb) stored in the storage unit 84) may be switched. As a result, the state in which the target frequency Fc is switched to Fc ′ can be immediately reflected via the cancellation signal B, and the control followability is further improved.

  In step S7, the correction cancellation signal generation unit 40 generates the correction cancellation signal Ab by correcting the control signal B input from the amplitude phase adjustment unit 36 side. The correction cancellation signal generation unit 40 is configured by, for example, an FIR (Finite Impulse Response) filter, and has a filter characteristic that simulates a transmission characteristic of a sound representing a transmission path from the speaker 18 to the microphone 16. Thereby, the correction cancellation signal Ab simulates the cancellation signal at the position of the microphone 16.

  In step S8, the subtractor 42 subtracts the correction cancellation signal Ab (see step S7) input through the correction cancellation signal generation unit 40 from the error signal A (see step S1) input through the microphone 16. Thus, the non-control detection signal Awoc is generated. This non-control detection signal Awoc is the error signal A (that is, the vibration noise NS) that would be detected by the microphone 16 under the condition that the cancellation vibration noise CS was not output (ANC control was not executed). (Simulated signal).

  Next, steps S9 to S11 in FIG. 2 will be described in detail with reference to the second detailed block diagram in FIG. 4 while paying attention to the signal processing flow of the real part component and the imaginary part component. For convenience of illustration, FIG. 4 omits some of the components shown in FIG.

  In step S <b> 9, the second adaptive notch filter 44 generates the second control signal O <b> 2 based on the reference signal X from the reference signal generator 22. Hereinafter, a specific configuration and operation of the second adaptive notch filter 44 will be described.

  The second adaptive notch filter 44 includes a first filter 90 in which the real part filter coefficient Rw2 is variably set, a second filter 92 in which the imaginary part filter coefficient Iw2 is variably set, and an output signal on the first filter 90 side. And a subtractor 94 for subtracting the output signal on the second filter 92 side. The first filter 90 scales the amplitude component of the real part reference signal Rx (cosine wave signal) input from the real part reference signal generation unit 60 side by Rw times, and outputs it to the subtractor 94 side. The second filter 92 scales the amplitude component of the imaginary part reference signal Ix (sine wave signal) input from the imaginary part reference signal generation unit 62 by Iw times, and outputs it to the subtractor 94 side. Thereafter, the subtracter 94 subtracts the output signal (= Iw2 · Ix) on the second filter 92 side from the output signal (= Rw2 · Rx) on the first filter 90 side. As a result, the second adaptive notch filter 44 outputs the second control signal O2 (= Rw2 · Rx−Iw2 · Ix).

  In step S10, the subtractor 46 uses the second control signal O2 (step S10) input via the second adaptive notch filter 44 from the non-control detection signal Awoc (see step S8) input via the subtractor 42. The second correction error signal E2 is generated by subtracting (see S9). In this case, a second correction error signal E2 from which a frequency component having a predetermined width centered on the target frequency Fc is removed is obtained by the action of the SAN bandpass filter 50.

  In step S <b> 10, the second filter coefficient update unit 48 updates the filter coefficient W <b> 2 of the second adaptive notch filter 44. Hereinafter, a specific configuration and operation of the second adaptive notch filter 44 will be described.

  The second filter coefficient updating unit 48 corresponds to the real part multiplier 96 and the gain adjuster 98 that are used for updating the real part filter coefficient Rw2 corresponding to the real part of the filter coefficient W2, and the imaginary part of the filter coefficient W2. The imaginary part multiplier 100 and the gain adjuster 102 are provided for updating the imaginary part filter coefficient Iw2. That is, in the present configuration example, the second filter coefficient updating unit 48 updates the filter coefficient W2, that is, the real part filter coefficient Rw2 and the imaginary part filter coefficient Iw2, respectively, according to an LMS (Least Mean Square) algorithm. The update algorithm is not limited to this method, and various optimization methods may be adopted.

The real part multiplier 96 multiplies the real part reference signal Rx input from the real part reference signal generation part 60 side by the second correction error signal E2 input from the subtractor 46 side, and supplies it to the gain adjuster 98 side. Output. The gain adjuster 98 multiplies the amplitude component of the multiplication signal by μ2 and outputs it to the first filter 90 side. Here, the constant μ2 corresponds to a step size parameter. Then, the first filter 90 adds the update amount (= + μ 2 · Rx · E 2) acquired from the second filter coefficient update unit 48 to the current real part filter coefficient Rw 2, thereby creating a new real part filter. A coefficient Rw2 is obtained. That is, the real part filter coefficient Rw2 is updated according to the following equation (3).
Rw2 ← Rw2 + μ2, Rx, E2 (3)

On the other hand, the imaginary part multiplier 100 multiplies the imaginary part reference signal Ix input from the imaginary part reference signal generation unit 62 side by the second correction error signal E2 input from the subtractor 46 side, and the gain adjuster 102. Output to the side. The gain adjuster 102 multiplies the amplitude component of this multiplication signal by [mu] 2 and inverts the phase (adjusted by [pi]), and then outputs it to the second filter 92 side. Thereafter, the second filter 92 adds the update amount (= −μ2 · Ix · E2) acquired from the second filter coefficient update unit 48 to the imaginary part filter coefficient Iw2 at the current time, thereby obtaining a new imaginary part. A filter coefficient Iw2 is obtained. That is, the imaginary part filter coefficient Iw2 is updated according to the following equation (4).
Iw2 ← Iw2-μ2, Ix, E2 (4)

  Thereafter, the filter coefficient holding unit 52 (first holding unit 104) holds the real part filter coefficient Rw2 updated in step S10. Further, the filter coefficient holding unit 52 (second holding unit 106) holds the imaginary part filter coefficient Iw2 updated in step S10.

  In step S11, the frequency switching unit 54 determines the next target frequency Fc ′ with reference to the target frequency Fc set in step S2. In this step, the target frequency Fc may be updated (Fc ′ ≠ Fc) and may not be updated (Fc ′ = Fc). The calculation process in this step may be referred to as “frequency switching process”.

  Hereinafter, a method for updating the target frequency Fc, in other words, a specific operation of the frequency switching unit 54 will be described with reference to a flowchart of FIG. 5 and a schematic explanatory diagram of FIG.

In step S <b> 21, the frequency switching unit 54 calculates the phase angle θ (0 ≦ θ <2π) on the complex space in the filter coefficient W <b> 2 (t) at the current time t of the second adaptive notch filter 44. Specifically, the phase angle θ is calculated by θ = tan ⁻¹ (Iw2 / Rw2) using the filter coefficients (Rw2, Iw2) at the target frequency Fc.

In step S22, the frequency switching unit 54 calculates a phase angle change amount dθ from the phase angle θ calculated in step S21 and the previously calculated phase angle (hereinafter referred to as the previous phase angle θold). Specifically, it is calculated by the following equation (5).
dθ = (θ−θold) mod 2π (5)

  Here, the previous phase angle θold corresponds to the phase angle θ in the latest filter coefficient W2 (t−Ts). The phase angle change amount dθ is not limited to the difference in phase angle θ, and may be any type as long as it is a parameter representing the degree of change from the previous phase angle θold. Further, the phase angle change amount dθ may be calculated by using not only the previously calculated phase angle but also the phase angles calculated in the latest plural times.

  In step S23, the frequency switching unit 54 substitutes the value of the phase angle θ calculated in step S21 for the previous phase angle θold. This previous phase angle θold is used for the next calculation in step S22.

  In step S24, the frequency switching unit 54 calculates the frequency change amount dF from the phase angle change amount dθ calculated in step S22. Specifically, it is calculated by dF = dθ / (2πTs). Ts corresponds to a sampling period (unit: s) for inputting the error signal A.

  In step S25, the frequency switching unit 54 determines whether or not an update condition for the target frequency Fc is satisfied. The frequency switching unit 54 determines whether or not the frequency change amount dF calculated in step S24 is within a predetermined range. For example, any value between 0.05 and 0.2 Hz is selected as the lower limit threshold Th1, and any value between 1 and 3 Hz is selected as the upper limit threshold Th2.

  When the frequency change amount dF satisfies the relationship Th1 ≦ | dF | ≦ Th2, a new target frequency Fc ′ is determined according to the update formula Fc ′ = Fc + γdF (step S26). Note that γ is a positive value (for example, 0 <γ <1), and corresponds to a parameter for adjusting the follow-up speed of this control.

  On the other hand, if 0 ≦ | dF | <Th1 is satisfied, Fc ′ = Fc and the target frequency Fc is maintained without being updated (step S27). When 0 ≦ | dF | <Th1, the frequency characteristic of the vibration noise NS is assumed to be stable. By not switching the target frequency Fc, it is possible to suppress the occurrence of different noise (for example, overshoot) due to excessive control.

  Alternatively, when | dF |> Th2 is satisfied, Fc ′ = Fc and the target frequency Fc is maintained without being updated (step S27). When | dF |> Th2 is satisfied, it is assumed that it is difficult to predict the behavior of the noise signal NS, or that the elapsed time after activation of the ANC device 10 is not sufficient. By not switching the target frequency Fc, it is possible to suppress the occurrence of different noise (for example, overshoot) due to excessive control.

  In this way, the frequency switching unit 54 sequentially determines the target frequency Fc at the predetermined sampling period Ts (step S11).

  In step S <b> 12, the speaker 18 outputs canceling vibration noise CS based on the canceling signal B from the amplitude / phase adjusting unit 36. Hereinafter, the vibration noise NS canceling control can be performed by sequentially repeating steps S1 to S12 at a predetermined sampling period Ts.

  Next, operational effects obtained by performing the above-described frequency switching process will be described with reference to FIGS. 7A to 10C. Each of FIGS. 7A to 10C represents a graph in which the horizontal axis represents frequency [Hz] and the vertical axis represents gain [dB] (logarithm of amplitude).

  FIG. 7A is a spectrum diagram of the error signal A before execution of ANC control, in other words, the vibration noise SN alone. The first spectrum SPC1 has one peak near a frequency of 45 Hz and one peak near a frequency of 70 Hz. Here, it is assumed that the peak near the frequency of 70 Hz where the spectrum intensity is maximum is suppressed using ANC control.

  FIG. 7B is a frequency characteristic diagram of the SAN type bandpass filter 30 suitable for the error signal A shown in FIG. 7A. The frequency setting unit 20 (see FIGS. 1, 3 and 4) sets the target frequency Fc to Fc = 70 Hz, so that the gain at the frequency of 70 Hz is maximum (signal loss is at the minimum level) as in this example. A filter characteristic is obtained. Thereby, it is possible to selectively extract a frequency component to be canceled out of the vibration noise NS input from the microphone 16.

  FIG. 7C is a frequency characteristic diagram of the first adaptive notch filter 24 corresponding to the frequency characteristic of the SAN-type bandpass filter 30 shown in FIG. 7B. This characteristic substantially matches the result obtained by adding the gain (unit: dB) of the SAN type bandpass filter 30 shown in FIG. 7B for each frequency to the first spectrum SPC1 shown in FIG. 7A.

  By the way, the tendency of the resonance noise may be different depending on the interaction between components constituting the suspension of the vehicle 11 and the like. For example, depending on the running state of the vehicle 11, the resonance frequency may change dynamically.

  As illustrated in FIG. 8A, it is assumed that the frequency characteristic of the error signal A (first spectrum SPC1 illustrated by a broken line) is changed while the vehicle 11 is traveling, and the resonance frequency is shifted from 70 Hz to 67 Hz. Hereinafter, the frequency characteristic after the change is referred to as a second spectrum SPC2 (illustrated by a solid line).

  FIG. 8B is a frequency characteristic diagram of the SAN-type bandpass filter 30 suitable for the error signal A shown in FIG. 8A. As in the case of FIGS. 7A and 7B, the frequency component to be canceled out of the vibration noise NS input from the microphone 16 by using a filter that maximizes the gain of the second spectrum SPC2 at the peak frequency of 67 Hz. Can be selectively extracted.

  However, when the above-described frequency switching process is not performed, the SAN type bandpass filter 30 remains the frequency characteristics shown in FIG. 7B. In this case, as shown in FIG. 9A, the frequency characteristic of the first adaptive notch filter 24 has a relatively small gain in the vicinity of 67 Hz as compared with the characteristic of FIG. 7C. As a result, a sensitivity function of the ANC device 10 as shown in FIG. 9B is obtained, and a spectrum diagram of the error signal A as shown in FIG. 9C is obtained.

  In FIG. 9C, the solid line graph shows the frequency characteristics after the ANC control is performed on the error signal A having the second spectrum SPC2. The broken line graph shows frequency characteristics after the ANC control is performed on the error signal A having the first spectrum SPC1. Thus, when the resonance frequency of the vibration noise NS shifts and slightly deviates from the target frequency Fc, the vibration noise NS in the vicinity of the resonance frequency cannot be sufficiently canceled.

  On the other hand, in the ANC device 10 according to the present embodiment, the active vibration noise control unit 14 dynamically changes the pass band of the SAN type bandpass filter 30 following the shift of the resonance frequency. Specifically, the frequency switching unit 54 calculates the frequency change amount dF (= −3 Hz), and then switches the target frequency Fc from 70 Hz to 67 Hz. As a result, the frequency characteristic of the SAN type bandpass filter 30 is changed from the characteristic indicated by the broken line in FIG. 8B to the characteristic indicated by the solid line in FIG.

  That is, when the frequency switching process described above is performed, as shown in FIG. 10A, the frequency characteristic of the first adaptive notch filter 24 has a relatively large gain in the vicinity of 67 Hz as compared with the characteristic of FIG. 9A. As a result, a sensitivity function of the ANC device 10 as shown in FIG. 10B is obtained, and a spectrum diagram of the error signal A as shown in FIG. 10C is obtained.

  In FIG. 10C, the solid line graph shows the frequency characteristics after the ANC control is performed on the error signal A having the second spectrum SPC2. The broken line graph shows frequency characteristics after the ANC control is performed on the error signal A having the first spectrum SPC1. Thus, even when the resonance frequency of the vibration noise NS is shifted, a substantially equal canceling effect is obtained before and after the shift.

  As described above, the phase angle change between the phase angle θ on the complex plane of the filter coefficient W2 (Rw2, Iw2) of the second adaptive notch filter 44 and the previous phase angle θold calculated at the previous update. Since the frequency switching unit 54 that calculates the amount dθ and switches the target frequency Fc of the reference signal X according to the phase angle change amount dθ is provided, the change amount of the phase angle θ on the complex plane of the filter coefficient W2 is successively calculated. It is possible to monitor, and it is possible to easily and accurately grasp the trend of the change in frequency characteristics from the phase angle change amount dθ. Thereby, even if there is a change in the frequency characteristics of the vibration noise NS, it is possible to execute the cancellation control of the vibration noise NS following the change.

  The active vibration noise control unit 14 includes two adaptive notch filters, that is, a first adaptive notch filter 24 used for ANC control, and a second adaptive notch filter 44 used for update control of the target frequency Fc. It has. Since the second adaptive notch filter 44 for calculating the phase angle θ is separately provided, the calculation process is not easily affected by the state of the ANC control. Thereby, the calculation accuracy of the phase angle θ is further increased, and the effect and reliability of the frequency switching control are increased.

  In addition, this invention is not limited to embodiment mentioned above, Of course, it can change freely in the range which does not deviate from the main point of this invention.

DESCRIPTION OF SYMBOLS 10 ... ANC apparatus 11 ... Vehicle 14 ... Active type vibration noise control part 16 ... Microphone 18 ... Speaker 20 ... Frequency setting part 22 ... Reference signal generation part 24 ... 1st adaptive notch filter 26, 42, 46 ... Subtractor 28 ... 1st 1 filter coefficient update unit 30, 50 ... SAN type band pass filter 36 ... amplitude phase adjustment unit 38 ... amplitude phase switching unit 40 ... correction cancellation signal generation unit 44 ... second adaptive notch filter 48 ... second filter coefficient update unit 54 ... Frequency switching part

Claims (4)

Vibration noise canceling means for outputting canceling vibration noise based on a canceling signal for vibration noise;
Error signal detection means for detecting residual vibration noise due to interference between the vibration noise and the canceling vibration noise as an error signal;
An active vibration noise control device having active error noise control means for receiving the error signal and generating the cancellation signal,
The active vibration noise control means includes:
Reference signal generation means for generating a reference signal of a predetermined frequency;
An adaptive notch filter that receives the reference signal and outputs a control signal used to generate the cancellation signal;
Amplitude / phase adjustment means for storing an adjustment value of the amplitude or phase according to the frequency of the reference signal and generating the cancellation signal by adjusting the amplitude or phase of the control signal;
Correction error signal generating means for generating a correction error signal by subtracting the control signal from the error signal;
Filter coefficient updating means for sequentially updating filter coefficients of the adaptive notch filter based on the reference signal and the correction error signal so that the correction error signal is minimized;
A correction cancellation signal generation unit that corrects the cancellation signal based on a transfer characteristic from the vibration noise cancellation unit to the error signal detection unit to generate a correction cancellation signal;
A non-control time detection signal generating means for generating a non-control time detection signal by subtracting the correction cancellation signal from the error signal;
A second adaptive notch filter having a filter coefficient defined on a complex plane and receiving the reference signal and outputting a second control signal;
Second correction error signal generation means for generating a second correction error signal by subtracting the second control signal from the non-control detection signal;
Second filter coefficient updating means for sequentially updating filter coefficients of the second adaptive notch filter based on the reference signal and the second correction error signal so that the second correction error signal is minimized;
A phase angle change amount between the phase angle on the complex plane of the filter coefficient of the second adaptive notch filter and the phase angle calculated at the time of the previous update is calculated, and according to the phase angle change amount An active vibration noise control device comprising: frequency switching means for switching the frequency of the reference signal. The active vibration noise control apparatus according to claim 1,
The frequency switching means calculates a frequency change amount based on the sampling period of the error signal and the phase angle change amount, and maintains the frequency of the reference signal when the frequency change amount falls below a lower threshold value. An active vibration noise control device characterized by: The active vibration noise control device according to claim 2,
The frequency switching means maintains the frequency of the reference signal when the amount of change in frequency exceeds an upper threshold value that is greater than the lower threshold value. The active vibration noise control apparatus according to any one of claims 1 to 3,
An active vibration noise control apparatus, further comprising: an amplitude phase switching unit that switches the adjustment value stored in the amplitude phase adjustment unit in response to the frequency switching unit switching the frequency of the reference signal.

Images