JPH07230287A - Active noise control system - Google Patents

Abstract

PURPOSE:To prevent noises by the discontinuous points of the cancellation signals formed at the time of high-speed revolution of an engine by forming the cancellation signals only for a specific engine pulse among the engine pulses inputted at short intervals. CONSTITUTION:This active noise control system has adaptive filters ADSG 1 to 4 which form the cancellation signals by the engine pulses synchronized with engine noises. The system offsets and decreases the engine noises in the closed space in the room of an automobile, etc., based on the formed cancellation signals. Then, the cancellation signals of only the specific engine pulses among the engine pulses of the short intervals are formed and the other engine pulses are jumped out of reading without forming the cancellation signals if the engine pulses are input at the short intervals shorter than the specific intervals. As a result, the number of revolutions of the engine is sharply increased and even if the engine pulses are inputted at the short intervals, the noises by the discontinuous points generated in the cancellation signals are prevented.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an active noise control system (hereinafter simply referred to as ANC) for reducing engine noise of an automobile, a ship, etc. in space by using an adaptive filter to reduce the noise. ) Concerning.

[0002]

2. Description of the Related Art Conventionally, an adaptive filter for generating a cancel signal for reducing engine noise in such an ANC is based on a reference input signal which is an engine pulse synchronized with the engine noise and an error signal extracted from a microphone. To generate a cancel signal.

[0003]

However, according to the above-mentioned conventional ANC, the adaptive filter generates a cancel signal based on the engine pulse and the error signal, but the engine speed is rapid. Then, since the tap interval of the adaptive filter becomes shorter, the engine pulse becomes shorter than a predetermined interval, and there is a problem that a discontinuity occurs in the cancel signal and noise occurs.

The present invention has been made in view of the above problems, and an object thereof is an active noise control for preventing noise due to a discontinuity point of a cancel signal generated when an engine rapidly rotates. To provide a system.

[0005]

In order to achieve the above object, the present invention has an adaptive filter for generating a cancel signal with an engine pulse synchronized with engine noise, and an automobile based on the generated cancel signal. Is a system for canceling and reducing the engine noise in a closed space in a vehicle interior, etc., and when the engine pulses are input at intervals shorter than a predetermined interval, a plurality of engine pulses input at the short intervals are provided. Among these, the cancel signal is generated only for a specific engine pulse, and the other engine pulses are skipped without generating the cancel signal.

[0006]

With this configuration, when the engine speed rapidly increases and engine pulses are input at intervals shorter than the predetermined interval, the process for generating the cancel signal becomes uncertain, and the discontinuity point generated in the cancel signal occurs. This can overcome the situation where noise is generated.

[0007]

Embodiments of the active noise control system according to the present invention will be described below.

In-vehicle active noise control system (Active Noise Control)
1 System; hereinafter referred to as ANC) as an example. This is a system that reduces noise in the vicinity of the occupant's head by applying an adaptive filter to cancel engine noise in space. The features of signal processing in the ANC mode of this system are listed below.

1) A batch update type SFX-LMS algorithm was devised to reduce the calculation amount.

2) A coefficient-stabilized FIR filter (STF) is used as a measure against discontinuous sounds, divergence of an adaptive filter, and growth of unnecessary low-frequency sounds.

3) A rotation speed, rotation state sensitive type, STF, and step size changing system are adopted.

4) Adopting a sampling timing adjustment function by engine pulse interruption.

FIG. 1 is a perspective view showing the structure of this embodiment in three dimensions. In FIG. 1, an engine 100, which is a noise source of engine noise, and an engine control unit 200 that detects a rotation state and a rotation speed of the engine 100.
And a speaker 300 installed at an arbitrary position in the vehicle.
And a microphone 400 provided above each seat to pick up the noise in the vehicle, and an ANC controller 500 for controlling the entire ANC.

FIG. 2 shows a block diagram of the ANC mode.
SP1 to SP4 are cancellation sound output speakers, and MI
C1 to C4 are error microphones.

C11 to 44 are transfer functions between the speakers SP1 to SP4 and the microphones MIC1 to CH4.
1 to 44 are speaker SPs estimated by system identification
1 to 4 and transfer functions between the microphones MIC1 to MIC4.

“ADSG1 to 4” are adaptive filters (cancellation signal generators) [Adaptive Digital].
1 Signal Generator), which has been generally used conventionally.

"STF" is an ADSG coefficient stabilizing FIR filter, "UPF" is a batch update type SFX-LMS coefficient updating FIR filter, "SS1" is a step size of 1,
Fixed step size for each pass, "SS2"
Is the step size 2, which is a step size that changes depending on the rotation speed and the rotation state.

W1 (n) to W4 (n) are ADS before update
G coefficient. Reference numeral 11 is a phase inversion unit that inverts the phase of the output values output from the ADSGs 1 to 4. Reference numeral 10 is an engine speed / rotation state determination unit which is a monitoring unit and a control unit. ENGINEPULSE is ADSG1 ~
4 and a pulse synchronized with the engine speed input to the engine speed / rotation state determination unit 10.

In this system, as shown in FIG. 1, one engine 100 is used as a noise source and four error microphones 400 and four speakers 300 are used as CASE (1,
4, 4).

Next, the operation of the block diagram shown in FIG. 2 will be briefly described.

The sum of the engine noise and the cancel signal emitted from the speakers SP1 to SP4 is calculated by the microphone MIC1.
It is taken as an error signal for capturing at ~ 4. Then, the estimated transfer function (Chat) is rearranged (Data Revers).
e) Perform the convolution operation of the numerical value and the error signal. This is a feature of the batch update type SFX-LMS. This FIR
The filter is called UPF. (Details of this SFX-LMS will be described later). Further, the convolved value is multiplied by the step size 1.

The result of multiplication by the step size 1 is added for each input of the microphones MIC1 to 4 and multiplied by the step size 2 (SS2) according to the engine speed and the rotation state. This is the engine speed / rotation state sensitive step size changing method. In addition, this calculation result is A
It is an updated value of the DSG coefficient.

Current ADSG coefficients [W1 (n)] to [W
4 (n)], and the updated values are added.

In order to stabilize the addition result, STF
And the results are set as new ADSG coefficients [W1 (n + 1)] to [W4 (n + 1)]. ST
The characteristic of F changes depending on the engine speed and the rotation state.
This is the rotation speed sensitive STF changing method.

Further, the phase of the output value of the signal output from the ADSG is inverted by the phase inverting section 11. And
Output from the speakers SP1 to SP4. The above is the basic operation of the ANC mode.

Next, the Filtered-X LMS will be described.

In the normal Filtered-X LMS, when realizing ANC in a three-dimensional space, a signal having a correlation with a signal (noise) to be silenced is taken in as a reference signal. The signal processing algorithm in this case is shown in FIG. This is a general Filtered-X LMS algorithm.

The coefficient updating formulas of this Filtered-X LMS are shown in (Equation 1) and (Equation 2).

[0029]

[Equation 1]

[0030]

[Equation 2] Here, w is an adaptive filter coefficient, i is a filter coefficient number, μ is a step size, e is an error symbol, r is a transfer function correction filter output signal, C is a transfer function between a microphone and a speaker, and X is Is the reference input signal, j is the impulse response number of C, and k is the number of taps of C.

(Equation 2) is a convolution calculation of the transfer function C of the space (between the speaker and the microphone) and the input data X, which is a feature of Filtered-X.

Here, it is assumed that an impulse represented by the following equation is input to the reference input. However, this impulse is repeated in synchronization with a certain periodic noise that is to be silenced.

X (0) = 1 X (i) = 0 i <0 i> 0 If the condition of the above equation is satisfied, the following equation holds.

N = j → X (n−j) = 1 n ≠ j → X (n−j) = 0 Then, (Equation 2) becomes the following equation.

R (n) = cn That is, r (n) is to sequentially output the impasl response cn of the transfer function C, and no convolution operation is required. As for the filter W, similarly, the filter coefficient wi may be sequentially output without performing the convolution operation, the amount of calculation can be significantly reduced, and the normal Fil as shown in FIG.
The calculation result is equivalent to the case where a pulse is input to the reference input of tered-X. This is Synchronized
Filtered-X (SFX) algorithm.

Next, the batch update type SFX-LMS will be described.

In SFX, (Equation 1) and (Equation 2) become (Equation 3).

[0038]

[Equation 3] In the conventional algorithm, (Formula 3) is updated by k + 1 times for one sample as shown in FIG. Therefore, in order to complete the update of the coefficient w i of the filter W, the tap number of C must be updated k + 1 times. This explanation will be given with reference to FIG. FIG. 4 is a diagrammatic representation of (Equation 3). Vertically aligned e and c indicate multiplication of the second term on the right side of (Equation 3). This is an example in which C is 8 taps (k = 7) and the reference input pulse is input to the 22nd sample, where the vertical axis is the sample number (n) and the horizontal axis is the tap number (i ). P1, P2, and P3 indicate pointer positions. These pointers will be described later.

For example, consider a case where the 11th sample wi is updated. In the conventional SFX-LMS algorithm, the eight coefficients w19 to w12 are updated according to (Equation 3) (operation in the part surrounded by the dotted line in FIG. 4). With this processing, the update of w12 is completed.

Next, focusing on w12, this tap is
The update starts from the 4th sample and ends at the 11th sample. The updated w12 'is shown as in (Equation 4).

[0041]

[Equation 4] It should be noted that (wi ′) indicates that updating of wi is completed.

When the equation (4) is converted into a general formula, it is represented by the equation (5).

[0043]

[Equation 5] The second term on the right side of the equation (5) is a convolution operation of the error signal and the coefficient in which the coefficient cj of the transfer function C is inversely arranged. (Fig. 4
The calculation of the part surrounded by the solid line is shown. ) (Equation 5) can be reduced by k + 1 times for both multiplication and subtraction, as compared with the case where the conventional (Equation 3) is operated k + 1 times.

Further, as a feature of the DSP, since the architecture which is good at the convolution operation is not adopted, in reality,
The calculation result can be shortened more than the reduced number of calculations.

This is the batch update type SFX-LMS algorithm. The UPF of FIG. 2 is the batch update type SFX-
The LMS is an update coefficient calculation unit, and the processing is an FIR filter having a coefficient of a sequence of reversely arranged Chat impulse responses in which the transfer function C is estimated.

In FIG. 4, P1 is a coefficient output pointer. The pointer is incremented on the filter coefficient (w) for each sample. Since it is SFX, the convolution operation is not performed and the data wi indicated by P1 may be output. Also, this pointer returns to w0 when a pulse is input to the reference input. That is, the number of taps of the adaptive filter is variable. P2 is a coefficient updating pointer indicating a coefficient w to be collectively updated. P3 is a pointer for storing the STF output in ADSG.

Next, the ADSG coefficient stabilizing FIR filter (STF) will be described.

The following conditions can be considered as conditions under which the adaptive filter of this system becomes unstable.

1) Set the frequency band of noise to the speakers SP1 to SP1.
When the characteristic of 4 cannot be covered, the filter coefficient diverges. For example, the noise frequency is 50 Hz, and the speaker SP
When 1 to 4 cannot generate a low frequency band such as 50 Hz, the adaptive filter generates a signal of 50 Hz, but the error signal does not decrease, so that the adaptive filter diverges.

2) As a characteristic of Filtered-X LMS, the high frequency range is likely to diverge.

3) If an offset is applied to the microphones MIC1 to MIC4 input for some reason, the DC component is also superimposed on the adaptive filter.

4) When the engine speed changes, a discontinuity occurs in the output of the adaptive filter.

As a method of solving these problems, a filter coefficient stabilizing filter (STF) is adopted.

The principle of this filter coefficient stabilizing filter will be described.

The value updated by the batch update SFX-LMS algorithm is filtered by a linear phase FIR filter. The characteristic of this FIR filter is basically a band pass filter, and cuts the updating of the low sound range that cannot be reproduced by the speakers SP1 to SP4 and the high sound range that cannot be muted in space. By this filter, it is possible to prevent the divergence of the low sound range and the high sound range and reduce the generation of the discontinuous sound when the engine speed changes.

By using a linear phase FIR filter for STF, it becomes possible to perform frequency filtering while maintaining the phase state. Since the delay due to this STF is the number of samples that is half the number of taps, the adaptive filter must be updated after that time has elapsed.

This explanation will be given with reference to FIG. When the STF is a 7-tap FIR filter, the filtered result is output after 3 samples. Therefore,
The value obtained by filtering wi ′ (n + 1), which is the result of (Equation 5), may be updated after three samples. P3 is the address to be updated.

Next, the rotation speed, the rotation state sensitive step size, and the STF changing method will be described.

As a drawback of the SFX algorithm, the influence of the error signal is directly reflected in the output signal due to its structure.
In other words, when a signal having no correlation with the reference input is input to the error signal, the signal having no correlation is superimposed on the output signal.

In the case of ANC such as this system, when a voice is uttered toward the error microphone,
Echo may occur from the speaker. As a countermeasure against this, there is a method of suppressing the echo by reducing the step size and the update amount, but this processing deteriorates the system performance.

Conventionally, the step size has been determined in consideration of these balances. As a result, the performance of following the acceleration / deceleration of the engine could not be obtained.

Another problem is that due to the structure of the algorithm, a discontinuity may occur in the cancel signal during acceleration / deceleration. As a countermeasure, STF was added, but discontinuous sound could not be effectively reduced with a single characteristic.

Therefore, in this embodiment, the engine speed and the rotation state are monitored, and the step size, S
Changing the TF alleviated the above problems.

The principle of this system will be described.

The engine speed and rotation state were monitored using a DSP, and the step size and STF were controlled accordingly. The number of rotations is divided into three stages of low rotation, medium rotation and high rotation, and the rotation state is divided into four stages of steady rotation, acceleration, deceleration, and rapid deceleration, and a total of 12 cases of these combinations are divided, and each is divided into four cases. The optimum step size and STF coefficient are stored in a memory, which is a storage unit (not shown), and each is selected from the memory.

FIG. 5 shows an example of the engine speed, the step size with respect to the rotation state, and the STF, and FIG. 6 shows the frequency characteristic of the STF. For example, during acceleration / deceleration, a discontinuity is likely to occur, so the cutoff frequency of the low-pass filter is set low. The cutoff frequency at this time is 200-4
00 Hz is suitable. In addition, the step size is set to be large in order to improve the followability to the change in noise.

On the other hand, during steady rotation, the cutoff frequency of the low-pass filter is increased in order to mute the sound as high as possible, and the step size is decreased in order to suppress echo. At the time of sudden deceleration such as gear change, it is determined that the adaptive processing cannot follow, and all the ADSG coefficients are cleared. Note that FIG. 5 and FIG. 6 are examples for explanation and are not limited to this value.

Next, the sampling timing adjusting function by the engine pulse interrupt will be described.

Algorithm of this system (collective update type S
FX-LMS) is a variable tap algorithm that changes the tap length of the adaptive filter according to the engine speed.
The tap length of the adaptive filter matches the cycle of two engine revolutions, which is discrete with the sampling cycle as a unit. Therefore, the conventional method cannot accurately match the cycle of the noise and the cycle of the cancel sound, which causes a problem that a beat sound is generated.

Therefore, as a countermeasure against this, the engine pulse is used as a reference of the sampling cycle, and the cycle of the noise and the cycle of the canceling sound are forcibly matched. This operation is shown in FIG.

Specifically, the engine pulse is input to the external interrupt of the DSP, and the sampling period (timing of AD / DA converter) is synchronized with the engine pulse. When the engine pulse is input, the processing of the Nth sample in FIG. 7 is interrupted and new sampling is started. As a result, the sampling frequency is artificially increased only when the engine pulse is input, and beat noise can be suppressed.

Further, as described above in the explanation of the principle of the filter coefficient stabilizing filter (STF), in any case, as shown in FIG. 4, the coefficient output pointer P1
And the coefficient update pointer P2 is equal to the number of taps of the FIR filter (UPF).
2 and STF output result is stored in ADSG
The interval with 3 is half the number of STF taps, and the causality in the number of STF and UPF taps in the interval between each pointer is unchanged.

As a result of outputting the coefficient from the coefficient output pointer P1, the coefficient of the coefficient updating pointer P2 is calculated, and as a result of inputting the coefficient of the coefficient updating pointer P2 to the STF, the calculation is performed by the pointer P3. Value is A
It is stored in the DSG.

However, when the engine speed increases,
The interval of engine pulses synchronized with the engine speed becomes short, and the coefficient output pointer P1 may overtake the pointer P3. In this case, there is a problem that the causality in the number of taps of STF and UPF in the interval between the pointers described above cannot be satisfied, accurate adaptation cannot be performed, and noise is generated.

Therefore, according to the active noise control system of the present invention, only in such a case, one engine pulse is ignored, that is, one engine pulse is skipped, and the coefficient output is performed. Since the occurrence of the situation where the pointer P1 overtakes the pointer P3 is eliminated, it is possible to prevent the generation of noise.

Since the noise is repeated at engine pulse intervals, the cancel sound is also repeated. Therefore,
By copying the cancellation signal for the first engine pulse interval after the skipped engine pulse,
The effect can be further enhanced for discontinuous sounds.

[0077]

According to the active noise control system of the present invention configured as described above, when the engine speed rapidly increases and the engine pulse is input at an interval shorter than the predetermined interval, the cancel signal is issued. It is possible to overcome the situation in which the process of generating is not disturbed and noise is generated due to the discontinuity point generated in the cancel signal.

[Brief description of drawings]

FIG. 1 is a perspective view showing a three-dimensional structure of the present embodiment.

FIG. 2 is a block diagram showing an outline of an ANC mode.

FIG. 3 is a block diagram showing a basic algorithm of SFX-LMS.

FIG. 4 is an explanatory diagram illustrating a batch update type SFX-LMS.

FIG. 5 is an explanatory diagram showing tuning results of engine speed and rotation state.

FIG. 6 is an explanatory diagram showing STF coefficients and characteristics.

FIG. 7 is a conceptual diagram showing an engine pulse interrupt state.

[Explanation of symbols]

SP1-4 Canceling sound output speaker MIC1-4 Error microphone C11-44 Transfer function Chat11-44 Transfer function ADSG1-4 Adaptive filter STF ADSG coefficient stabilizing filter UPF Batch update SFX-LMS coefficient update FIR filter MYU2 R Step size 2 ENGINE PULSE Engine pulse 10 Engine speed / rotation status judgment unit

Claims (2)

[Claims] 1. An adaptive filter for generating a cancellation signal with an engine pulse synchronized with engine noise, and canceling the engine noise in a closed space in a vehicle cabin or the like based on the generated cancellation signal. When the engine pulse is input at an interval shorter than a predetermined interval, the cancel signal is generated only for a specific engine pulse among a plurality of engine pulses input at the short interval. The active noise control system is characterized in that other engine pulses are skipped without generating the cancel signal. 2. The cancel signal generated immediately before the skipped engine pulse is copied to the skipped engine pulse.
Active noise control system as described.