JPH07325587A - Device for reducing noise in vehicle - Google Patents

Abstract

PURPOSE:To provide a stable device for reducing noise in a vehicle room suppressing the effect due to a random signal and without giving a sense of incompatibility to the crew. CONSTITUTION:An Ig(ignition) pulse is shaped/thinned, and is made to a primary source Ps, and the primary source Ps is combolution product-sum operated with a filter coefficient of an adaptive filter 3, and is outputted as a cancel signal to be outputted from a speaker 6 as a cancel sound for a vibration noise in a reception point. A noise reduction condition is detected by a microphone 10 as an error signal to be inputted to an error signal process circuit 5. Then, in the error signal process circuit 5, an exponential mean value is obtained from the last time data and the error signal based on a noise vibration source signal, and the exponential mean value is processed to the value within the range set beforehand to be outputted to an LMS(least mean square) operation circuit 6 as this time process data. In the LMS operation circuit 6, a coefficient correction amount is obtained from the primary source Ps inputted through a CMN0 (compensation coefficient) circuit 4 and the this time process data, and the filter coefficient is updated.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a vehicle interior noise reduction device for reducing noise in a vehicle interior, which is mainly caused by engine vibration noise, by interfering with a canceling noise.

[0002]

2. Description of the Related Art In contrast to vehicle interior noise, which is mainly caused by engine vibration noise, a sound source that produces a sound having the same amplitude and opposite phase (cancellation sound) to reduce vehicle interior noise. Technology is proposed.

Recently, for example, Japanese Unexamined Patent Publication No. 3-178 has been used.
As disclosed in Japanese Patent Publication No. 845, etc., LMS (Least M
ean Square) algorithm (the theory that the correction formula of the filter is recursive to simplify the formula for calculating the filter coefficient of the adaptive filter, and is approximated by the mean square error), or this LMS algorithm MEFX-LMS (Multiple
A vehicle interior noise reduction device using the Error Filtered X-LMS) algorithm has been proposed and is partially put into practical use. In the vehicle interior noise reduction device using the LMS algorithm, when the vehicle interior noise generated mainly due to engine vibration is silenced, a signal highly correlated with the engine vibration is detected as a noise vibration source signal (primary source), and The canceling sound for the noise is synthesized by the adaptive filter from the primary source and generated from the speaker. Then, the noise reduction state at the listening point is detected by the microphone as an error signal, and the filter coefficient of the adaptive filter is updated by the LMS algorithm from this error signal and the primary source so that the noise reduction at the listening point becomes an optimum value. It has become.

[0004]

By the way, the above-mentioned LM
In a vehicle interior noise reduction device using the S algorithm or MEFX-LMS algorithm, the noise reduction state at the listening point is detected as an error signal by a microphone or the like, and the filter of an adaptive filter is detected from the instantaneous error signal and the primary source by the LMS algorithm. In order to update the coefficient, if the error signal contains noise components that are not subject to muffling (random signals; human speech, music, sudden sounds, etc.), the filter coefficient is updated due to the effect of this noise component. However, there is a problem in that the sound is emitted as a canceling sound from the muffling speaker, and the passenger feels uncomfortable.

The present invention has been made in view of the above circumstances. Even if a random signal such as a human voice, music, or a sudden sound is generated, the influence of the random signal is suppressed and the occupant feels uncomfortable. It is an object of the present invention to provide a stable vehicle interior noise reduction device that does not give noise.

[0006]

In order to achieve the above object, a vehicle interior noise reduction apparatus according to the present invention comprises a cancel signal synthesizing means for synthesizing a noise vibration source signal having a high correlation with engine vibration as a cancel signal by an adaptive filter. Cancellation sound generation means for generating the cancellation signal as a cancellation sound for noise, error signal detection means for detecting a noise reduction state at the listening point as an error signal, previous processing data based on the noise vibration source signal and the error signal An error signal processing means for obtaining an exponential average value from the above, processing this exponential average value to a value within a preset range, and outputting it as the current processed data, the noise vibration source signal, and the current processed data Filter coefficient updating means for updating the filter coefficient of the adaptive filter based on the above.

[0007]

[Operation] In the above configuration, first, when noise is generated in the vehicle compartment due to engine vibration noise as a main factor, the cancel signal synthesizing means synthesizes a noise vibration source signal having a high correlation with engine vibration as a cancel signal by an adaptive filter. Then, the canceling sound generating means generates the canceling signal as a canceling sound for noise. Then, the error signal detection means detects the noise reduction state at the listening point as an error signal, and the error signal processing means obtains an exponential average value from the previous processing data and the error signal based on the noise vibration source signal, This exponential average value is processed into a value within a preset range and output as the current processed data. Then, the filter coefficient updating means updates the filter coefficient of the adaptive filter based on the noise and vibration source signal and the current processed data.

[0008]

Embodiments of the present invention will be described below with reference to the drawings. 1 to 4 show a first embodiment of the present invention, FIG. 1 is a system schematic diagram of a vehicle interior noise reduction device, FIG. 2 is an explanatory diagram of an input signal conversion circuit, and FIG. 3 is an explanatory diagram of an error signal processing circuit. ,
FIG. 4 is an explanatory diagram showing an example of signal processing in the error signal processing circuit.

In FIG. 1, reference numeral 1 indicates a four-cycle engine, and an ignition pulse signal (Ig pulse signal) to an ignition coil (not shown) of the engine 1 is also output to an input signal conversion circuit 2.

As shown in FIG. 2, the input signal converting circuit 2 is composed of a waveform shaping circuit 2a and a thinning circuit 2b. The Ig pulse signal input to the input signal converting circuit 2 is The noise vibration source signal (primary source Ps is generated by forming and thinning out a signal having a frequency of 0.5 × n (n: integer) order component of the engine rotation in one pulse in two rotations of the engine in synchronization with the engine rotation. ), An adaptive filter 3 as a canceling signal synthesizing means, a speaker / microphone transfer characteristic compensation circuit (hereinafter abbreviated as "CMN0 circuit") 4
And output to the error signal processing circuit 5 as the error signal processing means.

This is because vibration noise associated with a 4-cycle engine requires two revolutions of the engine to complete one cycle in order to complete four strokes of intake, compression, explosion and exhaust in two revolutions (720 ° C.A) of the engine 1. Vibration noise is generated, and in the frequency domain, the 0.5th order component of engine rotation is the fundamental wave and the higher order component is the main spectrum (0.5 ×
This is because it is composed of n (n: integer) order components). Therefore, by shaping and processing the Ig pulse signal as described above, it is possible to obtain the primary source Ps having a very high correlation with the vibration noise to be silenced.

The adaptive filter 3 has a FIR (Finite Impulse) having a filter coefficient W (n) that can be updated by the LMS operation circuit 6 as a filter coefficient updating means.
Response) filter, which is formed with a predetermined number of taps (for example, 512 taps). The primary source Ps input to the adaptive filter 3 is convolution product summed with the filter coefficient W (n), and is output as a cancel signal to the D / A converter 7, and a filter circuit and an amplifier circuit (AMP) not shown The canceling sound is generated from the speaker 9 as the canceling sound generating means via the circuit 8.

The speaker 9 is arranged, for example, at a front door or the like in a vehicle (not shown), and an error signal detecting means is provided at a listening point in the vehicle (for example, a position close to an ear position of an occupant in the driver's seat). Is provided as a microphone 10.

The error signal indicating the noise reduction state detected by the microphone 10 (the signal indicating the result of the interference between the canceling sound and the vibration noise related to the engine; the error signal) is supplied to the amplifier circuit (AMP circuit) 11 and the filter. It is adapted to be inputted to the error signal processing circuit 5 via a circuit (not shown) and the A / D converter 12.

As shown in FIG. 3, the error signal processing circuit 5 includes a trigger signal generating section 5a, an acceleration / deceleration determining section 5b, and
The signal storage unit 5c, the exponential average calculation unit 5d, the signal determination unit 5e, and the signal setting unit 5f are mainly configured.

In other words, the trigger signal generating section 5a is a circuit section which receives the primary source Ps and outputs the pulse of the primary source Ps to the exponential average calculating section 5d as an exponential average trigger pulse. .

The primary source Ps is input to the acceleration / deceleration determination section 5b, the acceleration / deceleration of the engine is determined based on the input signal, and the exponential average calculation section 5d will be described later according to the acceleration / deceleration. It is a circuit that sets the constants of the calculation formula.

Further, the signal storage unit 5c is a RAM circuit, and a circuit for storing the processing data obtained this time as "previous processing data" for each tap in accordance with the output from the signal setting unit 5f. It is a division.

The exponential average calculator 5d reads the previous processed data e (n) from the signal memory 5c,
An exponential average value Px (n) is obtained from the previous processed data e (n) and the error signal E (n) input from the microphone 10,
The circuit section outputs the exponential average value Px (n) to the signal determination section 5e. Here, the formula of the exponential averaging process is given by Px (n) = ((N-1) e (n) + E (n)) / N (1). Further, N is a constant of 1 or more set by the acceleration / deceleration determination unit 5b, and as a result of the experiment, it is 4 in a steady state.
It is preferable to set 1 ≦ N ≦
When it is defined as 4, N = 4-α × | Psn-Psn-1 | (2) Here, α is a constant, Psn is the pulse interval of the current trigger pulse, and Psn-1 is the pulse interval of the previous trigger pulse.

As is clear from the above equation (1), if the value of the constant N is set too large, the influence of the error signal E (n) obtained this time is lowered too much, and the system is in a transient state. There is a possibility that the followability will be deteriorated, and it is necessary to set it within a range where sufficient stability can be secured. Therefore, in the first embodiment, the acceleration / deceleration determination unit 5b is used.
Is provided and the constant N is variable according to the engine operating state.

The signal determining section 5e is a circuit for determining whether the exponential average value Px (n) input from the exponential average calculating section 5d is within a range determined in advance by experiments or the like. is there. This determination is performed by the following formula. | Px (n) | ≧ P (3) P is a limit value obtained by experiments or the like.

Then, the signal setting section 5f has the equation (3) of the signal determining section 5e (| Px (n) | ≧
As shown in FIG. 4C, the data of each tap value (of P) is replaced with the previous processed data for each tap value, and if | Px (n) | ≧ P, the data is stored in the signal storage unit 5c. e '(n)
Is the previous processed data e (n), and if | Px (n) | <P, the data e ′ (n) is written as the current processed data e (n), and the current data is written to the LMS arithmetic circuit 6 at this time. It is a circuit unit that outputs processed data e ′ (n).

When | Px (n) | <P at all tap values in FIG. 4 (a), | Px (n) | ≧ P at the x1th tap in FIG. 4 (b). 4c, if
An example in which the data in FIG. 4B is replaced by the signal setting unit 5f is shown in FIG.

On the other hand, in the CMN0 circuit 4, a speaker / microphone transfer characteristic CMN is set in advance by approximation with a finite impulse response (as a compensation coefficient CMN0), and the input primary source Ps is compensated by the above-mentioned compensation. The circuit outputs the signal to the LMS operation circuit 6 after being compensated by multiplying by the coefficient CMN0 (convolution product sum).

In the LMS arithmetic circuit 6, the current processed data e '(n) from the error signal processing circuit 5 and the primary source Ps compensated (convoluted product sum) by the CMNO circuit 4 are obtained. From the above, the correction amount of the filter coefficient W (n) of the adaptive filter 3 is obtained by the well-known LMS algorithm, and the filter coefficient W (n) is updated.

In FIG. 1, reference character C represents a vehicle body transmission characteristic with respect to vibration noise of the engine 1.

Next, the operation of the embodiment having the above structure will be described. First, engine vibration noise is
From the inside to the inside of the vehicle through a mount or the like (not shown), and the sounds of intake and exhaust also propagate inside the vehicle. These engine-related vibration noises are all 0.
It is mainly composed of a frequency spectrum of a 5 × n (n: integer) -order component, and reaches the listening point (for example, a position close to the driver's ear) by being multiplied by the vehicle body transfer characteristic C for each vibration source.

On the other hand, the ignition pulse signal (Ig pulse signal) to the ignition coil (not shown) of the engine 1 is input to the input signal conversion circuit 2 and the waveform shaping circuit 2a and the thinning circuit 2b change the engine speed. Synchronized with one pulse for two engine revolutions, 0.5 engine revolutions
Shaped into a signal consisting of the frequency of × n (n: integer) component
Noise and vibration source signals (primary source Ps) are thinned out
Are output to the adaptive filter 3, the speaker / microphone transfer characteristic compensation circuit (hereinafter abbreviated as “CMN0 circuit”) 4, and the error signal processing circuit 5.

The primary source Ps input to the adaptive filter 3 has a filter coefficient W of the adaptive filter 3.
By a convolution product sum with (n), it is output to the D / A converter 7 as a cancel signal for canceling vibration noise, and is output to the speaker 9 via a filter circuit and an amplifier circuit (AMP circuit) 8 not shown. Is output from the speaker 9 as a canceling sound for the vibration noise at the listening point. At this time, the canceling sound reaches the listening point by receiving the speaker / microphone transfer characteristic CMN.

Therefore, at the listening point, the vibration noise related to the engine and the canceling noise interfere with each other to reduce the vibration noise, and at the same time, the microphone 10 arranged near the listening point causes The result of the interference between the vibration noise and the canceling sound is detected, and the error signal processing is performed as an error signal E (n) via the amplifier circuit (AMP circuit) 11, the filter circuit (not shown) and the A / D converter 12. It is input to the circuit 5.

In the error signal processing circuit 5, the error signal E (n) is input to the exponential average calculation unit 5d, and the primary source Ps is input to the trigger signal generation unit 5a and the acceleration / deceleration determination unit 5b.

Then, the acceleration / deceleration of the engine is determined based on the primary source Ps input to the acceleration / deceleration determination section 5b (based on the change of the pulse interval), and the exponential average calculation section 5d of the exponential average calculation section 5d determines the acceleration / deceleration. Set the constant of calculation formula.

Thereafter, the exponential average calculator 5d is triggered by the pulse of the primary source Ps input to the trigger signal generator 5a, and the input error signal E (n) is stored in the signal storage 5c. An exponential averaging process is performed based on the read previous processed data e (n) to process a past error signal value into a compressed value and output to the signal determination unit 5e.

The signal determination section 5e determines whether or not the input exponential average value Px (n) is within a range determined in advance by experiments or the like. For example, as shown in FIG. To
When the | xx tap is | Px (n) | ≧ P, a signal is output to the signal setting unit 5f together with the information (| P
Even when x (n) | <P, a signal is output to 5f together with these pieces of information).

When | Px (n) | ≧ P in the signal setting section 5f as shown in FIG. 4B, the previous processing data e (n) from the signal storage section 5c is obtained. Is read, and the value at the x1 tap is the previous processed data e
The value is replaced with the value at the x1 tap of (n) and output to the LMS operation circuit 6. On the other hand, the replaced processing data e ′ (n) is newly written in the signal storage unit 5c as the previous processing data e (n). When | Px (n) | <P, the current processed data e ′ (n) is output to the LMS operation circuit 6 without replacement, while
The signal storage unit 5 is newly added as previously processed data e (n).
Write to c.

The primary source Ps input to the CMN0 circuit 4 is a speaker / microphone transfer characteristic CMN.
Is approximated by a finite impulse response (compensation coefficient CMN0) and the product is convoluted and output to the LMS operation circuit 6.

Then, in the LMS operation circuit 6, the processing data e '(n) from the error signal processing circuit 5 and the CMN are input.
0 From the primary source Ps compensated by the circuit 4, L
The correction amount of the filter coefficient W (n) of the adaptive filter 3 is obtained by the MS algorithm, and the filter coefficient W (n) is updated.

Thus, according to the first embodiment, for example,
Even if a random signal such as a person's voice, music, or sudden sound is generated, this random signal is limited by the error signal processing circuit, so that a stable silencing system that is not affected by these random signals can be realized. .

Further, since the error signal processing circuit is provided with the acceleration / deceleration judging section, the silencing system can flexibly cope with a transient state such as sudden acceleration / deceleration. Needless to say, the acceleration / deceleration determination unit can be omitted when the muffling system can function under stable conditions with few such transient states.

In the first embodiment, the Ig pulse is used as the primary source Ps, but a signal having a high correlation with other engine-related vibration noise (for example, fuel injection pulse Ti) is used. Primary source Ps
Also good.

In the first embodiment, the example of the noise reduction device using the LMS algorithm of one channel (one microphone, one speaker) has been described.
MEFX-LM with algorithm expanded to multiple channels
The present invention is also applicable to a vehicle interior noise reduction device (for example, a device having four microphones, four speakers, etc.) that uses the S (Multiple Error Filtered X-LMS) algorithm.

Next, FIGS. 5 and 6 show a second embodiment of the present invention, FIG. 5 is an explanatory diagram of an error signal processing circuit, and FIG. 6 is an explanatory diagram showing an example of signal processing in the error signal processing circuit. is there. The second embodiment is different from the first embodiment in that the method of setting the limit value of the exponential average value in the signal determination unit of the error signal processing circuit and the replacement of the signal setting unit of the error signal processing circuit are different. The part is the same as that of the first embodiment, and the same parts are denoted by the same reference numerals and the description thereof is omitted.

In FIG. 5, reference numeral 5'denotes an error signal processing circuit corresponding to the error signal processing circuit 5 of the first embodiment.
This error signal processing circuit 5'includes a trigger signal generation section 5a.
An acceleration / deceleration determination unit 5b, a signal storage unit 5c, an exponential average calculation unit 5d, a signal determination unit 15e, and a signal setting unit 15f.
It is mainly composed of and.

The signal determining section 15e is a circuit for determining whether or not the exponential average value Px (n) input from the exponential average calculating section 5d is within a range determined in advance by experiments or the like. The determination is performed by the following formula. Px (n) ≧ e (n) + Q (4a) Px (n) ≦ e (n) -Q (4b) Here, Q is a limit value set so that it can be applied even in a transient state by an experiment, e (n) is the last processed data.

Then, the signal setting section 15f uses the equation (4a) or the equation (4b) of the signal determining section 15e.
As shown in FIG. 6C, the data of each tap value in the formula is converted into the formula (4a) (Px
In the case of (n) ≧ e (n) + Q), the data of this tap portion is set to e (n) + Q, and the equation (4b) (Px (n) ≦ e (n) −
In the case of Q), the data in this tap portion is replaced as e (n) -Q, and the processed data e ′ is stored in the signal storage unit 5c.
(n) is written as a new previous processed data e (n), and the processed data after replacement (current processed data e ′ (n)) is output to the LMS arithmetic circuit 6 as a circuit unit. There is. Other than the expressions (4a) and (4b), that is, e (n)
If −Q <Px (n) <e (n) + Q, the data storage unit 5c is not replaced and the processed data that is not replaced = e ′ (n) as a new exponential average value is replaced with a new e (n). With writing as
It outputs e ′ (n) to the LMS operation circuit 6.

That is, as shown in FIG. 6 (a), the restriction band according to the above equations (4a) and (4b) is formed with reference to the previous processed data e (n), As shown in FIG. 6, at the x2 tap, when exiting from this restricted zone, as shown in FIG.
Then, it is replaced with the value of the restriction band (e (n) + Q in FIG. 6 (c)).

As described above, in the second embodiment, since the exponential average value from the exponential average calculating unit 5d is limited with reference to the previous processed data, the random signal is not affected more effectively. It can be a stable silencing system.

[0048]

As described above, according to the present invention,
For example, even if a random signal such as a person's voice, music, or a sudden sound is generated, it is possible to suppress the influence of the random signal and obtain stable muffling performance that does not make the occupant feel uncomfortable.

[Brief description of drawings]

FIG. 1 is a system schematic diagram of a vehicle interior noise reduction device according to a first embodiment of the present invention.

FIG. 2 is an explanatory diagram of an input signal conversion circuit according to the first embodiment of the present invention.

FIG. 3 is an explanatory diagram of an error signal processing circuit according to the first embodiment of the present invention.

FIG. 4 is an explanatory diagram showing an example of signal processing in the error signal processing circuit according to the first embodiment of the present invention.

FIG. 5 is an explanatory diagram of an error signal processing circuit according to a second embodiment of the present invention.

FIG. 6 is an explanatory diagram showing an example of signal processing in an error signal processing circuit according to a second embodiment of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 engine 3 adaptive filter (canceling signal synthesizing means) 5 error signal processing circuit (error signal processing means) 6 LMS arithmetic circuit (filter coefficient updating means) 9 speaker (cancellation sound generating means) 10 microphone (error signal detecting means) e ( n) Previous processed data e '(n) Current processed data E (n) Error signal (error signal) Ps Primary source (noise and vibration source signal) Px (n) Exponential average value W (n) Filter coefficient

Claims (1)

[Claims] 1. A cancel signal synthesizing means (3) for synthesizing a noise vibration source signal (Ps) having a high correlation with engine vibration as a cancel signal by an adaptive filter, and a canceling sound generation for generating the cancel signal as a canceling sound against noise. Means (9), an error signal detecting means (10) for detecting the noise reduction state at the listening point as an error signal (E (n)), and the previous processing data (e ( n)) and the error signal (E (n)) to obtain an exponential average value (Px (n)) and process the exponential average value (Px (n)) to a value within a preset range. Then, the error signal processing means (5) for outputting the processed data (e '(n)) of this time, the noise vibration source signal (Ps) and the processed data of this time (e' (n)) are used for the adaptation. A filter that updates the filter coefficient (W (n)) of filter (3). A noise reduction device (6) for updating the vehicle interior noise.