JPH0683368A - Active noise controller for vehicle - Google Patents

Abstract

PURPOSE:To reduce the calculation quantity and speed up the processing without causing a decrease in noise reduction effect. CONSTITUTION:On the basis of the crank angle signal X of an engine, a reference signal generation part 11 generates and outputs a reference signal (x) consisting of an impulse sequence having the same period with a filling sound generated in the engine. Then a noise signal generation part 13 outputs the filter coefficient of an adaptive digital filter Wm as a driving signal in order at specific sampling clock intervals from the point of time when the latest impulse of the reference signal (x) is generated. Further, an adaptive processing part 17 updates filter coefficients whose number correspond to the period N of the reference signal (x) among the respective filter coefficients of the adaptive digital filter Wm on the basis of a reference processing signal r1m obtained by convoluting the reference signal (x) and a transfer function filter CLAMBDA1m and residual noises e1-e8 so that the noise in the cabin is reduced.

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 for a vehicle, in which a noise transmitted from a noise source to a vehicle interior is interfered with a control sound emitted from a control sound source to reduce the noise. In an active noise control device for reducing noise transmitted from a noise source that generates periodic noise, the amount of calculation is reduced and high-speed processing is possible.

[0002]

2. Description of the Related Art Conventional active noise control devices include those described in British Patent No. 2149614 and Japanese Patent Publication No. 1-501344. These conventional devices are noise reduction devices applied to aircraft cabins and similar closed spaces, and a single noise source such as an engine located outside the closed space has a fundamental frequency f ₀ and its harmonics. It operates under the condition that noise including the waves f _{1 to} f _n is generated.

[0003] More specifically, includes a microphone for detecting a plurality of the installed sound pressure to a location within the closed space, and a plurality of loudspeakers for generating a control sound to the closed space, the frequency f ₀ of the noise source - Based on the f _n component, those frequencies f ₀ ~
The loudspeaker is driven by a signal having a phase opposite to that of the f _n component,
Therefore, a control sound having a phase opposite to that of the noise transmitted to the closed space is generated from the loudspeaker to cancel the noise.

Then, as a method of generating the control sound emitted from the loudspeaker, PROCEEDINGS OF THE IEEE, VOL.
63 PAGE 1692,1975, “ADAPTIVE NOISE CANSELLATION:
PRINCIPLES AND APPLICATIONS ”
An algorithm that applies the DROW LMS 'algorithm to multiple channels is applied. The content of the paper is "A MULTIPLE ERROR LMS ALGORITHM AND
ITS APPLICATION TOTHE ACTIVE CONTROL OF SOUND AND
VIBRATION ”, IEEE TRANS.ACOUST., SPEECH, SIGNAL PRO
CESSING, VOL.ASSP −35, PP. 1423−1434, 1987.

That is, the LMS algorithm is one of the algorithms suitable for updating the filter coefficient of the adaptive digital filter, and is, for example, a so-called Filter.
In the ed-X LMS algorithm, a transfer function filter that models the transfer function from the loudspeaker to the microphone is set for all combinations of the loudspeaker and the microphone, and a reference signal representing the noise generation state of the noise source is set. The filter coefficient of the variable filter digital filter provided for each loudspeaker is updated so that the value of the predetermined evaluation function based on the value processed by the filter and the residual noise detected by each microphone is reduced. ing.

However, in such a conventional active noise control device, since the reference signal representing the noise generation state is taken in as a continuous signal such as a sine wave,
At the time of the convolution operation of the reference signal and the transfer function filter and the convolution operation of the reference signal and the adaptive digital filter, each value of the sequence formed by sampling the continuous signal at a predetermined interval, and the transfer function filter and the adaptive digital filter Since it is necessary to integrate each filter coefficient and further add the result of the integration, there is a problem that the calculation amount becomes large.

[0007] As a conventional technique for dealing with such a problem, 51 of "Proceedings of Acoustical Society of Japan, March 1992"
There are active noise control devices described on pages 5 to 516,
In this device, a sine wave is not used as a reference signal representing a noise generation state, and a pulse is used as it is, so that a convolution operation is simplified or omitted by applying an impulse train synchronized with a fundamental frequency of noise. I was trying.

[0008]

Certainly, according to the conventional active noise control device described in the above-mentioned paper, the convolution calculation is simplified, and as a result, the amount of calculation is reduced and the operation speed is increased. However, if such an active noise control device is applied to a vehicle where the period of the generated noise changes in a relatively wide range, such as in the passenger compartment of a vehicle, the noise with the longest period is generated. In such a situation, it is necessary to set the number of taps of the adaptive digital filter based on long-period (low-frequency) noise so that the control sound is generated over the entire period. In a situation where short (high frequency) noise is generated, it means that unnecessary calculation was performed when updating the filter coefficient of the adaptive digital filter.

The present invention has been made by paying attention to the unsolved problem of the conventional technique, and simplifies the convolution calculation by taking in the reference signal as an impulse train synchronized with noise. At the same time, an object of the present invention is to provide an active noise control device for a vehicle, which can further reduce the amount of calculation by omitting unnecessary calculation when updating the filter coefficient of the adaptive digital filter.

[0010]

In order to achieve the above object, the invention according to claim 1 is capable of generating a control sound in a vehicle interior of a vehicle to which noise is transmitted from a noise source which emits periodic noise. A control sound source, a reference signal generating means for generating a reference signal composed of an impulse train having the same cycle as the noise emitted from the noise source, a residual noise detecting means for detecting residual noise at a predetermined position in the vehicle interior, and the control A transfer function filter that models a transfer function between the sound source and the residual noise detecting means, a reference processing signal generating means that convolves the transfer function filter and the reference signal to generate a reference processing signal, and a filter coefficient variable The adaptive digital filter and the filter of the adaptive digital filter at a predetermined sampling clock interval from the time when the latest impulse of the reference signal is generated. Drive signal generating means for sequentially outputting the output coefficient as a signal for driving the control sound source, and each filter coefficient of the adaptive digital filter so as to reduce noise in the vehicle compartment based on the reference processed signal and the residual noise. And adaptive processing means for updating a predetermined number of filter coefficients.

According to a second aspect of the present invention, in the first aspect of the invention, the adaptive processing means updates the number of filter coefficients of the adaptive digital filter according to the cycle of the reference signal. I decided. According to a third aspect of the present invention, in the first aspect of the invention, the adaptive processing means updates the number of filter coefficients corresponding to the cycle of the reference signal and its change rate among the filter coefficients of the adaptive digital filter. I decided to do it.

Further, in the invention described in claim 4, in the invention described in claim 1, the adaptive processing means generates the drive signal while executing a predetermined number of processes from the present time among the filter coefficients of the adaptive digital filter. The filter coefficient output from the means as a signal for driving the control sound source is updated. Further, the invention according to claim 5 is the invention according to any one of claims 1 to 4, wherein the noise source is an engine, and the reference signal generating means generates a reference signal based on the rotation of the engine. And

[0013]

According to the first aspect of the invention, the drive signal generating means performs predetermined sampling from the time when the latest impulse of the reference signal generated by the reference signal generating means is generated.
Since the filter coefficient of the adaptive digital filter is sequentially output to the control sound source as a drive signal at the clock interval, the control sound source generates a control sound corresponding to the filter coefficient of the adaptive digital filter.
Since the filter coefficient of the adaptive digital filter does not always converge to the optimum value, it cannot be said that the noise in the vehicle interior is necessarily reduced.

However, based on the reference processing signal generated by the reference processing signal generating means by convoluting the transfer function filter and the reference signal and the residual noise detected by the residual noise detecting means, the adaptive processing means makes the interior of the vehicle Since the filter coefficient of the adaptive digital filter is updated so that the noise of
As the control progresses, the filter coefficient of the adaptive digital filter converges to the optimum value. Therefore, the control sound emitted from the control sound source cancels the noise, and the noise in the vehicle interior is reduced.

The reference signal generated by the reference signal generating means is an impulse train having the same period as the noise emitted from the noise source, and the response of the transfer function filter to each impulse forming the reference signal is an impulse response. Therefore, it matches each filter coefficient of the transfer function filter. Therefore, the reference processed signal generating means does not need integration, and the convolution operation can be performed only by adding the filter coefficients.

Since the adaptive processing means updates only a predetermined number of filter coefficients among the filter coefficients of the adaptive digital filter, the amount of calculation is reduced as compared with the processing of updating all the filter coefficients. The number of filter coefficients to be updated in this adaptive processing means is, for example, in claim 2.
Thru | or it determines like invention of Claim 4.

That is, according to the second aspect of the invention, since the reference signal is an impulse train having the same cycle as the noise transmitted to the vehicle interior, the number of filter coefficients corresponding to the cycle of the reference signal is set. When updated, among the filter coefficients of the adaptive digital filter, only the filter coefficient output as the drive signal by the drive signal generation means within one cycle of the current noise is updated, so that unnecessary update calculation is omitted. become.

Since the period of the reference signal is known from the time from the generation of one impulse until the generation of the next impulse, strictly speaking, the noise currently being generated is It does not represent a cycle. Therefore, according to the second aspect of the present invention, in a situation where the noise cycle varies greatly, the number of updated filter coefficients does not correspond to the current noise cycle. There is a risk that the number of filter coefficients to be updated will decrease, and an unnecessary update operation may be performed.

Therefore, when the number of filter coefficients corresponding to the cycle of the reference signal and the rate of change thereof are updated as in the third aspect of the invention, the updated filter coefficients are obtained even in a situation where the cycle of noise fluctuates greatly. Since the number of the filter coefficients is determined in consideration of the variation, the number of updated filter coefficients may not correspond to the current noise period.

According to the fourth aspect of the present invention, the fact that the filter coefficient output as the drive signal from the drive signal generating means is updated while the process is executed a certain number of times from the present point in time is always This means that a fixed number of filter coefficients used in the previous process are updated, so that each filter coefficient is updated a fixed number of times until the time when it is used. Therefore,
If the fixed number is set so that the filter coefficient is appropriately converged by such a fixed number of update processes, the update calculation amount can be reduced without lowering the control accuracy.

Further, according to the invention of claim 5, since the reference signal is generated based on the rotation of the engine as the noise source, the noise generated in synchronization with the order component of the engine rotation, for example, the reciprocating four cylinders. In the case of an engine, the muffled noise and the like generated in synchronization with 1/2 rotation of the crank can be reduced.

[0022]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing an overall configuration of a first embodiment of the present invention. This embodiment is an active vehicle type for reducing muffled noise transmitted from an engine 4 as a noise source into a vehicle interior 6. The present invention is applied to the noise control device 1.

First, the structure will be described. The vehicle body 3 is supported by front wheels 2a, 2b, rear wheels 2c, 2d, and suspensions interposed between the wheels 2a to 2d and the vehicle body 3. The vehicle shown in FIG. 1 is a so-called front-mounted engine front-wheel drive vehicle in which the front wheels 2a and 2b are rotationally driven by an engine 4 arranged in the front part of the vehicle body 3. A crank angle sensor 5 is attached to the engine 4,
The crank angle sensor 5 supplies the crank angle signal X synchronized with the rotation of the crank angle of the engine 4 to the controller 10.

In the passenger compartment 6 of the vehicle body 3, loudspeakers 7a, 7b, 7c and 7d as control sound sources face front seats S ₁ and S ₂ and rear seats S ₃ and S ₄ , respectively. It is located at the door. Furthermore, each seat S ₁
The head restraint position of the to S ₄ is a microphone 8a~8h as residual noise detecting means, have been respectively disposed two by two, the residual noise signal e ₁ to e these microphones 8a~8h was measured as a sound pressure ₈ Are supplied to the controller 10.

The controller 10 includes a microcomputer, a necessary interface circuit, and the like, and the crank angle signal X supplied from the crank angle sensor 5 and the residual noise signal supplied from the microphones 8a to 8h. based on the e ₁ to e _8, and performs arithmetic processing to be described later, so that control sound to cancel the muffled sound to be transmitted to the passenger compartment 6 is emitted from the loudspeaker 7a to 7d, they loudspeaker 7a and outputs a drive signal y ₁ ~y ₄ to ～7D.

FIG. 2 is a block diagram showing the functional configuration of the controller 10. The controller 10 is based on the crank angle signal X and has an impulse train having the same cycle as the vibration generated in the engine 4 which causes a muffled noise. (For example,
In the case of four reciprocating cylinders, the cause of the muffled sound is based on the reference signal generator 11 that generates and outputs the reference signal x consisting of one impulse every 180 degrees rotation, and the interval between the impulses of the reference signal x. The cycle determination unit 12 that determines and outputs the cycle N of the vibration generated in the engine 4.

Further, the controller 10 includes the number of loudspeakers 7a to 7d (M: in the present embodiment,
M = 4) adaptive digital filter W _m (m = 1 to M)
Each filter coefficient W _mi, a drive signal generation unit 13 with the latest impulse of the reference signal x is output as the drive signal y _m from the time it is generated in order at a predetermined sampling clock interval, each loudspeaker 7a~ A transfer function filter C ^ _lm (l = 1) in which the transfer function between 7d and the microphones 8a to 8h is modeled in the form of a finite impulse response function.
˜L: L is the number of microphones 8a to 8h, and in this embodiment, L = 8) and the reference signal x are convoluted to generate and output the reference processed signal r _lm.
And a transfer function filter storage unit 15 that sets the transfer function filter C ^ _lm in the reference processing signal generation unit 14 based on the cycle N, and an r register 1 that temporarily stores the reference processing signal r _lm.
6, and the filter coefficient W _mi of the adaptive digital filter W _m in the drive signal generator 13 so that the muffled sound in the vehicle interior 6 is reduced based on the reference processed signal r _lm and the residual noise signals e _{1 to} e _8. And an adaptive processing unit 17 for updating the.

In this embodiment, the adaptive processing section 17 is
Based on the LMS algorithm, which is one of the suitable algorithms for updating the filter coefficient of the adaptive digital filter, among the filter coefficients W _mi of the adaptive digital filter W _m , the number of filter coefficients W _m0 corresponding to the cycle N is set.
~ Update W _mN . 3 and 4 show the controller 10
3 is a flowchart showing an outline of the processing executed in FIG. 3, FIG. 3 shows an interrupt processing executed every time the reference signal x (one impulse) is generated, and FIG. The interrupt processing executed at the timing of outputting the signal y _m is shown.

That is, when the reference signal x is generated, the interrupt processing of FIG. 3 is started. First, at step 001, other interrupt processing is disabled, and then step 00.
Move to 2. In step 002, cycle information N (P) to N (1) (FIG. 5) that stores past values of the cycle N of noise.
(A) and FIG. 6 (a)) are shifted according to the following equation (1) so that the latest information is fetched. Further, the process proceeds to step 003 and the transfer function filter C
^ Filter information C ^ that stores past values of _lm
lm (P) to C ^ _lm (1) (see FIGS. 6B and 6C)
Is also shifted to the latest information according to the following equation (2).

[0030] The cycle information N (p) and the filter information C ^ _lm (p)
The number in parentheses attached to indicates the period N or the transfer function filter C ^ _lm set in the previous process with respect to the current process.

Next, the process proceeds to step 004, and the latest noise cycle N is stored as cycle information N (0). It should be noted that the noise cycle N specifically uses the latest value of the counter k set in step 116 of FIG. 4 described later. Then, the process shifts to step 005 to clear the counter i, and shifts to step 006 to make the determination of the following formula (3).

N (0) <T (i) (3) When the determination in step 006 is “YES”, the process proceeds to step 007, the counter i is incremented,
The process moves to step 006 again and the determination of the above formula (3) is performed. Here, T (i) used in the determination in step 006
Is a value of each step obtained by dividing the range that the cycle N can take into predetermined steps. Therefore, the determination of the above formula (3) is performed in step 0
By sequentially performing from i = 1 in the loop of 06,007, when the determination of step 006 is “NO”,
It becomes clear how long the cycle stored in the latest cycle information N (0) is.

Therefore, the determination in step 006 is "NO".
If so, the process proceeds to step 008, and the filter information C ^ _lm (0) is set according to the following equation (4). _{C ^ lm (0) = C} ^ mem.lm (i) ...... (4) That is, the above _{(4) C ^ mem.lm (i} ) the expression set in advance corresponding to the length of period N A stored transfer function filter having a filter length that is not too long for the period N.

That is, in this embodiment, the transfer function filter C ^ _lm is set so that it is not too long with respect to the cycle N. Then, the procedure goes to step 009 to clear the counter k used in the processing shown in FIG. 4, and the procedure goes to step 010 to release the interrupt prohibition state, and then the processing shown in FIG. 3 this time is ended.

On the other hand, at the timing of outputting the drive signal y _m (that is, every sampling clock), the interrupt processing shown in FIG. 4 is executed. First, in step 101, other interrupt processing is disabled, Then, the process proceeds to step 102. In step 102, the value of the drive signal y _m is initialized according to the following equation (5).

Y _m = W _mk (5) Here, W _mk is the k-th filter coefficient of the adaptive digital filter W _m . Therefore, the initial setting value of the drive signal y _m set in this step 102 is the present time k of the adaptive digital filter W _m for the immediately preceding impulse input.
Is the response. Since this counter k is cleared at step 009 of the interrupt processing of FIG. 3, it is set to 0 at the time when the reference signal x (n) is generated, and the interrupt processing of FIG. 4 is executed many times from that time. It indicates whether or not it has been done (see FIGS. 5A and 6A).

Then, the process proceeds to step 103, and the drive signal y _m set in step 102 is output. Next, in step 104, the residual noise signals e _{1 to} e
₈ is read, and the routine proceeds to step 105, where the registers r _lm (I) to r _lm (1) storing the reference processed signal r _lm obtained in the past processing are shifted according to the following expression (6). .

[0038] Then, the process proceeds to step 106 and the register r _lm (0)
Is initialized according to the following equation (7).

R _lm (0) = C ^ _lmk (0) (7) where C ^ _lmk (0) is the latest transfer function filter C ^ set in step 008 of the interrupt processing of FIG. _lm
It is the k-th filter coefficient of (0) and is a transfer function filter filter C ^ _lm (0) for the immediately preceding impulse input.
Is the response at the present time k.

Then, the process proceeds to step 107, the counter u is set to 1, and then the process proceeds to step 108 to judge the following equation (8). len (C ^ _lm (u)) is the filter length (the number of taps) of the transfer function filter C ^ _lm (u) for the _uth previous impulse input. Therefore, when the determination in step 108 is "YES" Can determine that the response of the transfer function filter C _lm (u) to the u-th previous impulse input continues to the present time, and conversely, if the determination in step 108 is “NO”, it has already disappeared. You can judge that

Therefore, the determination in step 108 is "YE
In the case of “S”, the process proceeds to step 109 and the following (9)
The register r _lm (0) is accumulated according to the formula. r _lm (0) = r _lm (0) + C ^ _lmq (u) (9) where Is.

Here, C ^ _lmq (u) is the response of the transfer function filter C ^ _lm (u) to the _uth previous impulse input. Then, after the accumulation in step 109 is performed, the process proceeds to step 110, the counter u is incremented, the process returns to step 108 again, and the above-described processing is repeatedly executed until the determination in step 108 becomes “NO”.

When the determination in step 108 is “NO”, it can be determined that the response of the adaptive digital filter W _m to the uth previous impulse input from that time point does not continue until the present time, so that the drive signal generating section 14 It is assumed that the convolution calculation is completed, and the process proceeds to step 111. Then, in step 111, the filter coefficient W _mi of each adaptive digital filter W _m is updated according to the following equation (10) based on the LMS algorithm. However, the number of filter coefficients W _mi updated here is the number of filter coefficients W _m0 corresponding to the latest noise cycle N (0).
~ W _{mN (0)} only.

[0044] Note that α is a coefficient called a convergence coefficient, and is involved in the speed at which the filter converges optimally and its stability.

Filter coefficient W _mi in step 111
When the update of (1) is completed, the process proceeds to step 112 to increment the counter k, the process proceeds to step 113 to release the interrupt disabled state, and then the interrupt process shown in this FIG. 4 is terminated. Next, the function and effect of this embodiment will be described. The vibration of the engine 4 is radiated as a muffled sound in the vehicle interior 6 through the frame and the like. On the other hand, when the crank angle signal X is supplied from the crank angle sensor 5 to the controller 10, the reference signal generation unit 11 is generated in the vehicle interior 6 based on the crank angle signal X due to the driving of the engine 4. A reference signal x composed of an impulse train having the same cycle as the muffled sound is generated and output.

Then, the drive signal generator 13 causes the filter coefficient W _1i of each of the adaptive digital filters W _{1 to} W _{4 at} the sampling clock interval from the time (k = 0) when the latest impulse of the reference signal x is generated. ~ W _4i are sequentially supplied to the loudspeakers 7a to 7d as drive signals y _{1 to} y ₄ . Then, from the loudspeakers 7a to 7d to the passenger compartment 6
Although the control sound is generated in the interior of the vehicle, the filter coefficient W _mi of the adaptive digital filter W _m does not always converge to the optimum value immediately after the control is started. Therefore, the muffled sound transmitted to the vehicle interior 6 is not always required. Cannot be said to be reduced.

However, the reference signal x is transferred to the transfer function filter C.
When the reference processing signal r _lm processed by ^ _lm is supplied to the adaptive processing unit 17, the residual noises e _{1 to} e ₈ in the vehicle interior 6 output by the microphones 8a to 8h are supplied to the adaptive processing unit 17. , Above (10) based on LMS algorithm
Since each filter coefficient W _mi of the adaptive digital filter W _m is updated according to the equation, those filter coefficients W _mi
Converges toward the optimum value.

As a result, the muffled sound transmitted to the passenger compartment 6 is canceled by the control sound emitted from the loudspeakers 7a to 7d, and the noise in the passenger compartment 6 is reduced. Then, the reference signal x output from the reference signal generation unit 11
Is an impulse train having the same period as the noise period N as shown in FIG. 5A, the response of the adaptive digital filter W _m to each reference signal x is as shown in FIG. 5B. It is equal to the filter coefficient W _mi of the adaptive digital filter W _m .

Therefore, in step 102 of FIG. 4, the drive signal y _m is set to the filter coefficient W _mi corresponding to the current time k,
Even if the process of outputting this in step 103 is executed, it is substantially equivalent to outputting the result of the convolution of the reference signal x and the adaptive digital filter W _m as the drive signal y _m . However, since the reference signal x is an impulse train, if the filter length of the adaptive digital filter W _m is short (the number of taps is small), the drive signal y _m is not generated in the latter half of the period when the noise period N becomes long. Therefore, it is necessary to set the filter length of the adaptive digital filter W _m to be appropriately long. But,
As will be described later, since the number of updates of the filter coefficient W _mi of the adaptive digital filter W _m is limited to the number corresponding to the noise cycle N, there is no problem caused by increasing the filter length.

Further, the reference signal x and the transfer function filter C ^
_Even in the convolution operation with _lm , the reference signal x is shown in FIG.
Since the impulse train has the same period as the noise period N as shown in (a), the response of the transfer function filter C _lm to each reference signal is as shown in FIG. 6 (b) or (c). It is equal to the filter coefficient C ^ _lmj (p) of those transfer function filters C ^ _lm (p).

Therefore, by the processing of steps 106 to 110 in FIG. 4, each filter coefficient C ^ _lmk (p) at the sampling time k, which is the response of each transfer function filter C _lm (p) continuing up to the present time n. By simply accumulating, the convolution operation of the reference signal x and the transfer function filter C ^ _lm can be performed. As described above, according to the configuration of this embodiment, the drive signal y _m can be generated by reading the filter coefficient W _mi of the adaptive digital filter W _m , and the convolution operation of the reference signal x and the transfer function filter C ^ _lm can be performed. Since only the addition can be performed, an extremely small amount of calculation is required as compared with the conventional device that needs the addition and addition for the convolution calculation, and the processing speed of the entire noise reduction processing can be improved.

For example, in a conventional device that directly takes in a signal representing the noise generation state as a reference signal, the sampling clock of the reference signal is 1 kHz (sampling period 1 msec), and the filter length of the transfer function filter C ^ _lm . 20 taps (the number of taps J), the number of channels (L
XM) is 8 (L = 4, M = 2) and the filter length (the number of taps I) of the adaptive digital filter W _m is 6, the convolution operation of the transfer function filter C ^ _lm and the reference signal is J × L × M = 20 × 4 × 2 = 160 calculations are required. Further, the convolution calculation of the adaptive digital filter W _m and the reference signal requires I × M = 6 × 2 = 12 calculations. Therefore, the total number of calculations required for the convolution calculation is 160 + 12 = 172.

On the other hand, with the configuration of this embodiment, for example,
The engine 4 is an in-line 4-cylinder engine with an engine speed of 1
At 500 rpm, the noise cycle is 20 msec. Therefore, the convolution calculation of the transfer function filter C ^ _lm and the reference signal x is performed as compared with the conventional apparatus that performs the convolution calculation that is composed of 20 sum-product calculations every 1 msec. It becomes 1/20, and J × L × M / 20 = 8.

Further, the number of taps of the adaptive digital filter W _m is set so that the control sound is generated over the entire area within one cycle even if the engine speed is 1500 rpm.
It is necessary to divide 0 msec into 20 by dividing the sampling clock interval (interval for executing interrupt processing in FIG. 4) by 1 msec, but the number of taps of the adaptive digital filter W _m is 2.
Even if it is set to 0, the drive signal y _m only _needs to be read out from the filter coefficient W _mi of the adaptive digital filter W _m , and therefore the amount of calculation is 2, which is equal to the number M of loudspeakers.

That is, in the configuration of this embodiment, when the engine speed is 1500 rpm, the total calculation amount required for the convolution calculation is 8 + 2 = 10, which is about 1/17 of the conventional calculation amount. Further, when the engine speed is 4500 rpm, the noise cycle is about 6.7 msec. Therefore, the transfer function filter C ^ is smaller than that of the conventional device that performs the convolution operation that is the sum product operation 20 times every 1 msec. The convolution operation of _lm and the reference signal x is 1 / 6.7, which is J × L × M / 6.7≈24.

Since the drive signal y _m is generated only by reading the filter coefficient W _mi of the adaptive digital filter W _m , the amount of calculation is 2 which is equal to the number M of loudspeakers. That is, in the configuration of the present embodiment, when the engine speed is 4500 rpm, the total calculation amount required for the convolution calculation is 24 + 2 = 26, which is 1/6 or less of the conventional calculation amount.

Further, when the engine speed is 7500 rpm, the noise cycle is 4 msec. Therefore, the transfer function filter C ^ is compared with the conventional device which performs the convolution operation consisting of 20 sum-product operations every 1 msec. The convolution operation of _lm and the reference signal x is ¼, and J × L × M / 4 = 40.

Since the drive signal y _m is generated only by reading the filter coefficient W _mi of the adaptive digital filter W _m , the amount of calculation is 2 which is equal to the number M of loudspeakers. That is, in the configuration of the present embodiment, when the engine speed is 7500 rpm, the total calculation amount required for the convolution calculation is 40 + 2 = 42, which is 1/4 or less of the conventional calculation amount.

As described above, with the configuration of this embodiment, the noise reduction process can be executed in the entire normal traveling area with a smaller amount of calculation as compared with the conventional device. Here, as is apparent from the above calculation example, the calculation amount required for the convolution calculation of the reference signal x and the transfer function filter C ^ _lm tends to increase as the noise cycle N becomes shorter.

This means that when the noise period N becomes short as shown in FIG. 7A, the transfer function filter C ^ _lm necessary for the convolution operation becomes relatively small as shown in FIGS. 7B to _7F . This is because it is necessary to consider information in the distant past. Therefore, in this embodiment, the size of the noise cycle N is determined in steps 005 to 007 of the interrupt processing of FIG.
By setting the transfer function filter C _lm (0) in step 008 according to the size of the period N, the amount of calculation is prevented from extremely increasing even when the period N becomes short.

That is, the transfer function filter C ^ that accurately represents the acoustic transfer characteristics between the loudspeaker and the microphone.
If it is as shown in FIG. 8 (a), the transfer function filter C ^ should be always used, but the frequency characteristic of the transfer function filter C ^ (see FIG. 8 (b) ) of the reference), the frequency f _a ~f and characteristics of the band sandwiched by _b, the frequency characteristic shown in FIG. 8 (c) number of taps less transfer function of the filter as shown in C ^ '(FIG. 8 (d)
If the characteristics of the band sandwiched by the frequencies f _{a to} f _b match, and the frequency of the noise at that time exists in the frequency band f _{a to} f _b , the transfer function filter C ^ Even if C ^ 'is used instead of, the same noise reduction control can be performed.

That is, in step 008, the cycle N
By setting the transfer function filter C ^ 'as shown in FIG. 8 (c), which is not too long, the transfer function filter C ^ _lm necessary for the convolution operation can be calculated without degrading the control characteristic. As shown in 7 (b) to 7 (f), it is not necessary to consider even relatively far past information.
It is possible to prevent the amount of calculation from increasing extremely even when is shortened.

Further, in this embodiment, the number of updates of the filter coefficient _mi of the adaptive digital filter W _m in step 111 of FIG. 4 is determined by the noise period N (0 ), The number of filter coefficients to be updated is relatively large as shown in FIG. 9 in a situation where the noise period N is relatively long. In a relatively short situation, the number of filter coefficients to be updated is relatively small, as shown in FIG.

That is, in the present embodiment, among the filter coefficients W _mi of the adaptive digital filter W _m , the filter coefficient W output as the drive signal y _m within one cycle of noise.
_m0 to _W-mN only ₍₀₎ is updated, in other words, the drive signal y not outputted as _m filter coefficients W _{m (N (0) +1)} ~
Since W _mI is not updated, unnecessary update operation is omitted.

As described above, according to the configuration of this embodiment, the amount of calculation in the controller 10 can be extremely small, so that noise reduction processing can be performed at high speed, and an expensive digital signal processing device or the like having a high calculation speed. It can be fully realized without applying. Here, in the present embodiment, the crank angle sensor 5 and the reference signal generation unit 11 constitute the reference signal generation means, and the reference processing signal generation unit 14 and the processing of steps 106 to 110 constitute the reference processing signal generation means. Drive signal generator 13 and step 102
Drive signal generating means is constituted by the processing of
Processing of the adaptive processing unit 17 and step 111 ((10) above)
Equation) constitutes adaptive processing means.

FIG. 11 is a diagram for explaining the second embodiment of the present invention. Since the basic structure is the same as that of the first embodiment, its illustration and description will be omitted. Here, in the first embodiment, the number of filter coefficients corresponding to the noise cycle N (0) is updated, but the noise cycle N (0) is shown in FIGS. As can be seen from (a), since it is set based on the interval between the latest impulse of the reference signal x and the impulse generated immediately before it, strictly speaking, the noise currently being generated is Does not match the cycle of.

Therefore, there is no particular problem in a situation where the noise cycle does not fluctuate, for example, in a situation where the engine speed is constant, but in a situation where the noise cycle fluctuates greatly, the number of updated filter coefficients is currently No., the number of updated filter coefficients is smaller than that of the noise cycle, and the filter coefficients that have not been updated are output as the drive signal y _m near the end of the cycle. There is a risk that an unnecessary update calculation may be performed.

Therefore, in the present embodiment, the number of filter coefficients to be updated is determined based on not only the period of the reference signal x, but also the change state thereof, so as to deal with the above-mentioned problem. Specifically, as shown in FIG. 11, at the time n when the latest impulse of the reference signal x is generated, the cycle N (0) of the noise currently generated is predicted based on the following equation (11). The updating process of the above equation (10) is performed based on the predicted cycle N (0).

N (0) = N (1) + (N (1) −N (2)) (11) However, N (1) is the latest impulse of the reference signal x and the immediately preceding impulse. It is the interval between the generated impulse and N (2) is the interval between the immediately preceding impulse and the immediately preceding impulse. As a result, the number of updated filter coefficients will correspond to the current noise cycle even when the engine speed rapidly changes, such as during downshifts or during idling. Therefore, the problem that the number of updated filter coefficients is smaller than that of the above is solved.

FIG. 12 is a diagram for explaining the third embodiment of the present invention. Since the basic structure is the same as that of the first embodiment, its illustration and description will be omitted. That is, in the first or second embodiment, the number of filter coefficients to be updated is determined according to the interval of the reference signal x, but in the present embodiment, as shown in FIG. have decided to update the predetermined number of times filter coefficients are output as the drive signal y _m while performing the processing of _{Q W m (k + 1)} ~W m (k + Q).

In other words, in the present embodiment, the Q filter coefficients used in the subsequent processing are always updated, so that each filter coefficient is maintained until the time when it is used. Will be updated Q times. Therefore, if the number of times Q is set so that the filter coefficient appropriately converges by such Q times of updating processing, the amount of calculation can be reduced without lowering the control accuracy.

Since the number of updated filter coefficients is a fixed number regardless of the tap length I of the adaptive digital filter W _m and the noise cycle N, even when the noise cycle N is relatively long. Even if the tap number I is set so that the control sound is generated over the entire area, the amount of calculation required for updating is constant, and therefore the calculation load at the time of updating can be suppressed to a low level.

In each of the above embodiments, the case where the active noise control system for a vehicle according to the present invention is applied to a system for reducing the muffled noise transmitted from the engine of the vehicle to the passenger compartment has been described. The noise that can be reduced by the present invention is not limited to this, and it is possible to generate periodic noise, and as long as it can generate a reference signal composed of an impulse train of the same period as the noise. Of course, it is applicable.

For example, in order to reduce the noise generated in the transmission, the reference signal may be generated based on the rotation signal of the shaft and the gear position of the transmission, and the noise generated in the final reduction gear is reduced. If so, a reference signal should be generated based on the rotation signal and gear position of the final reduction gear, and if noise generated on the drive shaft should be reduced, the reference signal should be generated based on the drive shaft rotation signal. Is required, or if the noise generated by the propeller shaft is reduced, the reference signal should be generated based on the propeller shaft rotation signal, and the noise generated by the compressor of the air conditioner is reduced. If so, the reference signal should be generated based on the rotation signal of the compressor. If the noise generated from the fan of the motor is reduced, the reference signal may be generated based on the rotation signal of the fan.If the noise generated from the supercharger is reduced, the supercharger is reduced. The reference signal may be generated based on the rotation signal of the pump.If the noise generated by the water pump or the oil pump is reduced, the reference signal may be generated based on the rotation signal of the pump. If you want to reduce the noise generated by the alternator, you can generate the reference signal based on the rotation signal of the alternator.If you want to reduce the noise generated by the rotation of the wheel, rotate the wheel rotation signal. The reference signal may be generated based on

In each of the above-described embodiments, the reference signal x having a correlation with the muffled sound generated by the engine 4 is generated based on the crank angle signal X, but the invention is not limited to this. , The reference signal may be generated in synchronization with combustion in the engine 4.

[0076]

As described above, according to the present invention,
The impulse train with the same cycle as the noise emitted from the noise source is used as the reference signal, and only the necessary filter coefficients among the filter coefficients of the adaptive digital filter are updated, so the noise reduction effect is not reduced and the amount of calculation is reduced. It is possible to obtain an effect that the processing speed is increased due to the decrease of the number.

[Brief description of drawings]

FIG. 1 is an overall configuration diagram of a first embodiment of the present invention.

FIG. 2 is a block diagram showing a functional configuration of a controller.

FIG. 3 is a flowchart showing an outline of processing executed in a controller.

FIG. 4 is a flowchart showing an outline of processing executed in the controller.

FIG. 5 is a diagram showing a relationship between an adaptive digital filter and a reference signal.

FIG. 6 is an explanatory diagram of a convolution operation of a transfer function filter and a reference signal.

FIG. 7 is a diagram illustrating a problem when a noise cycle is shortened.

FIG. 8 is a diagram illustrating a solution to a problem when a noise cycle becomes short.

FIG. 9 is a diagram showing the number of filter coefficients updated when a noise cycle is long.

FIG. 10 is a diagram showing the number of filter coefficients to be updated when the noise cycle is short.

FIG. 11 is an explanatory diagram of the second embodiment of the present invention.

FIG. 12 is an explanatory diagram of a third embodiment of the present invention.

[Explanation of symbols]

 1 Vehicle Active Noise Control Device 4 Engine (Noise Source) 5 Crank Angle Sensor 6 Cabin 7a to 7d Loudspeaker (Control Sound Source) 8a to 8h Microphone (Residual Noise Detection Means) 10 Controller 11 Reference Signal Generation Unit 12 Period Judgment Unit 13 drive signal generation unit 14 reference processing signal generation unit 15 transfer function filter storage unit 16 r register 17 adaptive processing unit

Continuation of the front page (51) Int.Cl. ⁵ Identification number Reference number within the agency FI Technical indication location H03H 17/04 A 7037-5J 21/00 7037-5J (72) Inventor Kenichiro Muraoka Takara-cho, Kanagawa-ku, Yokohama-shi, Kanagawa No. 2 Nissan Motor Co., Ltd. (72) Inventor Tsutomu Hamabe No. 2 Takaracho, Kanagawa-ku, Yokohama, Kanagawa Prefecture Nissan Motor Co., Ltd.

Claims (5)

[Claims] 1. A reference composed of a control sound source capable of generating a control sound in a vehicle interior of a vehicle to which noise is transmitted from a noise source that emits periodic noise, and an impulse train having the same period as the noise emitted from the noise source. Reference signal generating means for generating a signal,
A residual noise detecting means for detecting residual noise at a predetermined position in the vehicle compartment, a transfer function filter modeling a transfer function between the control sound source and the residual noise detecting means, and the transfer function filter and the reference signal. Reference processing signal generating means for convolving to generate a reference processing signal, adaptive digital filter with variable filter coefficient, and the adaptive digital filter at a predetermined sampling clock interval from the time when the latest impulse of the reference signal is generated. Drive signal generating means for sequentially outputting the filter coefficients as a signal for driving the control sound source, and each filter of the adaptive digital filter so as to reduce noise in the vehicle compartment based on the reference processed signal and the residual noise. Adaptive processing means for updating a predetermined number of filter coefficients among the coefficients,
An active noise control device for a vehicle, comprising: 2. The active noise control device for a vehicle according to claim 1, wherein the adaptive processing means updates the number of filter coefficients of the adaptive digital filter according to the cycle of the reference signal. 3. The active noise control device for a vehicle according to claim 1, wherein the adaptive processing means updates the number of filter coefficients corresponding to the cycle of the reference signal and the rate of change thereof among the filter coefficients of the adaptive digital filter. 4. The adaptive processing means updates, among the filter coefficients of the adaptive digital filter, the filter coefficient output as a signal for driving the control sound source from the drive signal generating means while executing the processing a certain number of times from the present time. Claim 1
An active noise control device for a vehicle as described. 5. The active noise control device for a vehicle according to claim 1, wherein the noise source is an engine, and the reference signal generating means generates a reference signal based on the rotation of the engine. .