JPH0573075A - Active type noise controller - Google Patents

Abstract

(57) [Summary] [Object] To provide an active noise control device capable of forming a comfortable acoustic space by suppressing the generation of abnormal noise in a divergent state. [Structure] While the engine is rotating with the ignition switch turned on, the LMS algorithm of the steepest descent method is used to sequentially update the filter coefficients in the adaptive filter processing based on the residual noise detection signals e _{1 to} e ₃ and the crank angle signal X. By performing the adaptive filter processing, a drive signal for suppressing noise is output to the loudspeaker to exert a noise damping effect. At this time, the divergence tendency is determined by whether or not the number of times that the W power becomes the divergence prevention threshold A or more before the reset timer T times out is a preset set value C or more (steps S8 to S1).
6) When there is a divergence tendency, the divergence suppression coefficient β is doubled to decrease the filter coefficient, and the output sound from the loudspeaker is decreased to suppress the generation of abnormal noise.

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 device for suppressing noise by interfering noise from a noise source with a noise suppressing sound generated by a control sound source, and more particularly to divergence of noise attenuation control. The present invention relates to an active noise control device for preventing the noise.

[0002]

2. Description of the Related Art Conventionally, as this type of active noise control device, for example, there is one disclosed in Japanese Patent Application Publication No. 1-501344. This conventional example is an active noise control device that reduces a noise transmitted from a noise source into a vehicle compartment by emitting a control sound from a loudspeaker. The residual noise in the vehicle compartment is detected by a plurality of microphones. And adaptively filtering the reference signal by generating at least one reference signal containing arbitrarily selectable harmonics of the noise source and supplying the residual noise detection signal and the reference signal to a processor / storage unit. Then, the drive signal of the loudspeaker is formed. Here, in the adaptive filter processing, the i-th filter coefficient W _mi (n + 1) is set to minimize the evaluation function J set as the sum of mean squares of the detection signals V _ek of the plurality of microphones according to the LMS algorithm. In addition, the update is performed sequentially according to the following equation (1).

[0003] Here, W _mi (n) is the filter coefficient at the previous time, that is, the nth sample, μ is the convergence coefficient, V _ek is the kth microphone output, x is the reference signal correlated to the noise source, and C _km ′ is m.
It is a base filter coefficient corresponding to a model space transfer function that models a real space transfer function between the th speaker and the kth microphone.

[0004]

However, in the above-mentioned conventional active noise control device, the actual spatial transfer function C _km itself between the loudspeaker and the microphone is controlled by the control algorithm for updating the filter coefficient of the adaptive filter. Rather than using a model space transfer function C _km ′ that models this _,
When the actual space transfer function changes due to the deterioration of the microphone, the temperature change of the closed space, the opening and closing of the door, etc., evaluation is performed by causing an error in the filter coefficient W _mi (n + 1) of the adaptive filter calculated by the equation (1). There is an unsolved problem that the function J may diverge without converging to generate abnormal noise and it may be difficult to maintain a comfortable acoustic space.

Therefore, the present invention has been made by paying attention to the unsolved problem of the above-mentioned conventional example, and outputs the control sound from the control sound source when the active noise control device tends to diverge. An object of the present invention is to provide an active noise control device capable of suppressing abnormal noise generation and maintaining a comfortable acoustic space by suppressing the noise.

[0006]

In order to achieve the above object, an active noise control device according to a first aspect of the present invention, as shown in FIG. 1, generates a reference signal according to the noise generation state of a noise source. Reference signal generating means, adaptive filter means for adaptively filtering the reference signal of the reference signal generating means and supplying it to the control sound source, residual noise detecting means for detecting residual noise at the observation position, and the reference signal generating means. An active noise control device comprising: a filter coefficient updating means for updating a filter coefficient of the adaptive filter processing by a steepest descent method based on a reference signal and a residual noise detection value of the residual noise detecting means, wherein the filter coefficient updating means is provided. Is provided with a filter coefficient correction means for reducing the filter coefficient according to the filter output of the adaptive filter means.

Further, in the active noise control device according to a second aspect of the present invention, in the above structure, the filter coefficient correcting means counts the number of times the power based on the output of the adaptive filter exceeds the divergence prevention threshold, and the counting means. And a decrease amount increasing means for increasing the decrease amount of the filter coefficient each time the count value of is greater than or equal to a predetermined value. Further, in the active noise control device according to claim 3, the filter coefficient correction means counts the number of times that the power based on the output of the adaptive filter exceeds the divergence prevention threshold and is cleared at every predetermined time, and the counting means. And a decrease amount increasing means for increasing the decrease amount of the filter coefficient each time the count value of the means exceeds a predetermined value.

[0008]

According to the present invention, when the filter coefficient of the adaptive filter means for forming the drive signal for the control sound source is updated by the filter coefficient updating means, the filter coefficient correcting means adjusts the filter coefficient according to the filter output of the adaptive filter means. Reduce. Therefore, when the filter output has a diverging tendency and the filter output has a large value, the decrease amount of the filter coefficient is large, so that the filter output of the adaptive filter unit is suppressed to be small and the control sound output from the control sound source is suppressed. It

In particular, the filter coefficient correction means counts the number of times the power based on the output of the adaptive filter exceeds the divergence prevention threshold, and the reduction amount of the filter count each time the count value of the count means exceeds a predetermined value. And the decrease amount increasing means for increasing the decrease amount of the filter count increases as the divergence state approaches, and accordingly the filter output of the adaptive filter means decreases to suppress the generation of abnormal noise. You can

Further, by clearing the counting means at every predetermined time, there is a tendency that a sporadically generated door is closed with a loud noise or a disturbance such as passing through a railway guard is diverged. Prevent false detection.

[0011]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 2 is a schematic configuration diagram showing an embodiment in which the present invention is applied to a vehicle equipped with a 4-cylinder engine. In FIG. 2, reference numeral 1 denotes a vehicle body, and a four-cylinder engine 3 as a noise source is arranged in front of the vehicle interior 2. A front seat 4F and a rear seat 4R are arranged in the vehicle compartment 2, and for example, a lower part of the dashboard and a rear seat 4R are provided.
Loudspeakers 5a and 5b which also serve as a control sound source for outputting an audio signal as a control sound source are disposed on the rear side of the vehicle, and microphones 6a, 6b and 6c as residual noise detecting means are respectively provided at the front, center and rear of the ceiling. Are arranged.

Further, a crank angle sensor 7 as a reference signal generating means is attached to the engine 3, and the crank angle sensor 7 outputs 1 each time the crankshaft rotates 180 degrees from the crank angle sensor 7.
A crank angle detection signal X, which is a cycle sinusoidal signal, is output. Then, the residual noise detection signals e _{1 to} e ₃ output from the microphones 6a to 6c are the controller 1
5, the crank angle detection signal X of the crank angle sensor 7 and the switch signal of the ignition switch 10 are also input to the controller 15.

The controller 15, as shown in FIG.
A crank angle detection signal X is input, an A / D conversion circuit 21 that performs A / D conversion of the crank angle detection signal and outputs the signal, and a microphone 6a.
An amplifier 22a~22c for amplifying the residual noise detecting signals e ₁ to e ₃ of ~6c, 23a~ A / D conversion circuit for the amplified output is A / D converted outputs of these amplifiers 22a~22c
23c, a microcomputer 26 to which the conversion outputs of the A / D conversion circuits 21, 23a to 23c and the switch signal of the ignition switch 10 are input, and the loudspeaker 5 output from the microcomputer 26.
The D / A conversion circuits 27a and 27b for D / A converting the drive signals y ₁ and y ₂ of a and 5b, and the outputs of the D / A conversion circuits 27a and 27b are input to the microcomputer 26. From the analog switch 28a, 28b controlled by the control signal CA from the amplifier, and an amplifier 29a, which amplifies the analog signal output from the analog switch 28a, 28b and supplies it to the loudspeakers 5a, 5b.
29b.

Here, the microcomputer 26 always performs a convolution operation of the engine speed detection signal X as a reference signal based on the filter coefficient W _mi that is sequentially updated, and the drive signal y for the loudspeakers 5a and 5b.
Adaptive digital filter processing for calculating ₁ and y ₂ and filtering with a filter coefficient corresponding to the model space transfer function modeled according to the number of combinations of space transfer functions between the microphone and the speaker based on the engine speed detection signal X. Digital filter processing for generating a processed reference signal r _km (see the equation (5) described later), and adaptive digital filter processing based on the filtered reference signal r _km and residual noise detection signals e _{1 to} e _3. And a filter coefficient updating process for updating the filter coefficient W _mi in L.sub.m using the LMS (Least Mean Square) algorithm.

Here, the control principle of the microcomputer 26 will be described using a general formula. Now, the residual noise detecting signals the k-th microphone 6a~6c detects e _k
(n), the k-th microphone 6a to 6c when there is no control sound (secondary sound) from the loudspeakers 5a and 5b
Represents the residual noise detection signal detected by e _pt (n), and the transfer function H _km between the mth loudspeakers 5a and 5b and the kth microphones 6a to 6c is represented by a FIR (finite impulse response) function. J-th (j = 0, 1,
2, ──I _{c -} a first filter coefficient corresponding to the section) C
_kmj ', the engine speed detection signal X (n), and the i-th (i = 0) of the adaptive digital filter which inputs the engine speed detection signal X (n) and drives the m-th loudspeakers 5a and 5b. , 1, 2, --I _F -1) is defined as W _mi , the following equation (2) is established.

[0016] Here, the terms with (n) are all sample values at the sampling time n, and K is the microphone 6a.
6c (three in this embodiment), M is the number of loudspeakers 5a and 5b (two in this embodiment), and I _C is FIR
The number of taps (filter order) of the filter coefficient C _km ′ represented by the digital filter, I _F is the number of taps (filter order) of the filter coefficient W _mi represented by the adaptive digital filter.

[{ΣW _mi · X (nj
-i)} "(= the section y _m), the engine rotational speed detection signal X
Represents the output when adaptive digital filtering is performed,
The term “ΣC _kmj · {ΣW _mi · X (nji)}” means that the signal energy input to the m-th speaker 5a and 5b is output as acoustic energy from these speakers 5a and 5b, and the space transfer function in the vehicle interior It represents the signal when it reaches the kth microphone 6a to 6c via C _km , and the entire second term on the right side of "ΣΣC _kmj · {ΣW _mi · X (nji)}" is the kth microphone 6a. Since the arrival signals to ~ 6c are added for all speakers,
The total sum of secondary sounds that reach the k-th microphones 6a to 6c is shown.

Next, the evaluation function J is set as in the following equation (3). Then, in this embodiment, the LMS algorithm is adopted,
The filter coefficient W _mi that minimizes the evaluation function J is obtained, and each filter coefficient W _mi of the adaptive digital filter process is updated. The LMS algorithm, which is the steepest descent method, has the following (4)
As shown in the equation, the n-th value W _mi (n) is used as the filter coefficient of the adaptive digital filtering process, the mean square error gradient ∂J / ∂W _mi is calculated, and this is multiplied by α to obtain (n +
1) The second value W _mi (n + 1) is obtained, and calculation is performed so as to reduce the value of the evaluation function J.

[0019] However, Here, α is a convergence coefficient, which is a coefficient relating to the speed at which the filter converges optimally and the stability at that time.

Further, β is a divergence suppressing coefficient, which is a coefficient calculated as a negative term that determines the output of the control sound, and is an element for suppressing divergence. By increasing this value, the filter coefficient W _mi in the adaptive digital filter processing decreases, the values of the drive signals y _{1 and} y _{2 for} the loudspeakers 5a and 5b decrease, and the control sound output from the loudspeakers 5a and 5b decreases. Sound pressure of.

In this way, the filter coefficient W _mi (n + 1) in the adaptive digital filter processing is calculated by the microphone 6
residual noise detection signals e ₁ (n) to e output from a to 6c
₃ (n) and the engine speed detection signal X (n) from the crank angle sensor 7 and the output y _m (n) when the engine speed detection signal X (n) is subjected to the adaptive digital filter processing, the LMS algorithm By sequentially updating according to the following, drive signals y ₁ (n) and y ₂ (n) that minimize the input residual noise detection signals e ₁ (n) to e ₃ (n) are formed, and these are generated. Noises in the vehicle interior 2 are canceled by the control sounds supplied to 5a and 5b and output from them.

Next, the operation of the above embodiment will be described with reference to the flowcharts of FIGS. 4 and 5 showing the processing procedure of the microcomputer 26. When the key switch is turned on, the entire system is turned on and the microcomputer 26 executes the noise suppression control process shown in FIG. 4 as a timer interrupt process every predetermined time (for example, 1 msec). .. The control signal CA is set to "1" by the initialization processing by the main program when the power supply is turned on, thereby driving signal output analog switches 28a, by 28b from the microcomputer 26 y _1, y ₂
Is connected to the loudspeakers 5a and 5b, the β update counter N is cleared, the reset timer T is reset, and the divergence suppression coefficient β is set to “1”.

That is, first, in step S1, the switch signal of the ignition switch 10 is read and it is determined whether or not it is in the ON state. This determination is to determine whether the engine 3 is rotating. When the ignition switch 10 is in the off state, it is determined that the engine is stopped and the timer interrupt process is ended as it is, and the ignition switch 10 is turned off. When 10 is in the ON state, the process proceeds to step S2.

[0024] In step S2, the residual noise detecting signals e ₁ to e ₃ and the reference signal X (n) reads, then the processing proceeds to step S3, wherein (5) by performing digital filtering processing corresponding to the formula Filtered reference signal r _km
Is calculated, and then the process proceeds to step S4 and step S3
The filter coefficient W _mi (n + 1) is calculated by performing the calculation of the equation (4) based on the filtered reference signal r _km calculated in step ₁ and the residual noise detection signals e _{1 to} e _3, and then step S5 To the driving signals y ₁ and y ₂ for the loudspeakers 5a and 5b by executing the adaptive digital filter processing based on the filter coefficient W _mi (n + 1) calculated in step S6, and then to step S6. Calculated drive signal y
_After outputting ₁ and y ₂ to the D / A conversion circuits 27a and 27b, the process proceeds to step S7.

In step S7, the loudspeaker 5
W that represents the stability of the drive signals y _{1 and} y ₂ that drive a and 5b
The power W _P is calculated according to the following equation (11). Where W
The derivation of the power W _P will be described. The general formula for the output for each loudspeaker is: Here, W _m0 and W _m1 are the filter coefficients of the adaptive digital filter that controls the amplitude and phase of the m-th speaker output signal, and y _m is the output of the adaptive digital filter, which is the filter coefficient of the adaptive digital filter. And the reference signal X (n).

Z conversion of the equation (7) gives the following equation (8). Y _m (Z) = W _m0 · X (Z) + W _m1 · X (Z) · Z ⁻¹ = (W _m0 + W _m1 · Z ⁻¹ ) X (Z) ………………… (8) Here so,

[0027]

[Equation 1]

Here, ω = 2πf: input signal angular frequency T = 1 / f _sample : sample period f _sample : sample frequency When the powers of both sides are taken by the equation (9), the following equation (10) is obtained.

[0029]

[Equation 2]

In the equation (10), since the input power X _P is created by the program, the amount used to actually check the instability of the output is calculated from the filter coefficient of the adaptive digital filter as follows ( Equation (11) is used and this is called W power W _P. W _P = W _m0 ² + W _m1 ² +2 · W _m0 · W _m1 cos (2πf / f _sample ) (11) Next, the system proceeds to step S8 and is calculated in step S7. W power W _P representing the stability is determined whether a divergence prevention threshold a or more, when a W _P <a is the system will immediately terminated timer interrupt processing is judged to be stable predetermined Return to the main program of
When W _P ≧ A, it is determined that the system is approaching instability, and the process proceeds to step S9.

In step S9, the β update counter N is incremented and the process proceeds to step S10. Where β
The update counter N is a counter that accumulates the number of times the W power becomes equal to or higher than the divergence prevention threshold A during the noise suppression control.
In step S10, it is determined whether or not the flag 1 that is set when the reset timer T starts is set. If the flag 1 is set, the process directly proceeds to step S13, and the flag 1 must be set. For example, the process proceeds to step S11 to start the reset timer T, then the flag 1 is set in step S12, and then the process proceeds to step S13. Here, the reset timer T
Is set to a time (for example, 180 seconds) for accumulating the number of times the W power becomes equal to or higher than the divergence prevention threshold A during the noise suppression control.

In step S13, it is determined whether or not the reset timer T has timed out. When the reset timer T has timed out, the process proceeds to step S14 to clear the β update counter N, and then to step S15 to reset the flag 1. Then, the timer interrupt process is terminated and the process returns to the predetermined main program. When the time is not up, the process proceeds to step S16.

In step S16, it is determined whether or not the count value N of the β update counter is equal to or greater than the preset value C. This judgment is to judge whether or not the system tends to diverge from the stable state, and N <C
If it is, it is judged to be in a stable state, the timer interrupt processing is terminated, and a predetermined main program is restored.
When ≧ C, it is determined that there is a divergence tendency, the process proceeds to step S17, the value of the divergence suppression coefficient β is updated twice, and then the process proceeds to step S18 to clear the β update counter N and then step S19. Move to.

In this step S19, step S17
It is determined whether or not the value of the divergence suppression coefficient β updated by is equal to or larger than a preset value D. When β <D, it is determined that the divergence suppression coefficient β is still in the stable region, and the timer interrupt process is ended. Then, the process returns to the predetermined main program, and when β ≧ D, it is determined that the divergent state has been reached, the process proceeds to step S20, and the control signal CA having the logical value “0” is output to the analog switches 28a and 28b. Then, the analog switches 28a and 28b are turned off, and the loudspeaker 5
The system is stopped by stopping the sound emission of the control sounds from a and 5b.

Here, the processing of steps S3 and S4 of FIG. 4 corresponds to the filter coefficient updating means, the processing of step S5 corresponds to the adaptive filter means, and steps S7 to S15.
The processing of (1) corresponds to the counting means, and the processing of steps S16 to S18 corresponds to the reduction amount increasing means. Therefore,
Now, assuming that the key switch is off and the vehicle is parked, the power supply of the controller 15 is cut off and the execution of the noise suppression control process is stopped.

When the key switch is turned on from this state, energization of the controller 15 is started and the noise suppression control process of FIG. 4 is started to be executed by the controller 15. At this time, when the ignition switch 10 is maintained in the off state, the engine 3 is stopped and noise is not generated. Therefore, when the timer interrupt process of FIG. The noise suppression control process is not executed because the integration process is terminated.

In this state, when the ignition switch 10 is turned on and the engine is started, the noise suppression control process is started when the timer interrupt process shown in FIG. 4 is executed. In this noise suppression control process, step S1
Through, filtered reference signal r reference signal X from the residual noise detecting signals e ₁ to e ₃ and the crank angle sensor 7, which is the output of the microphone 6a~6c a (n) read by the digital filtering process in step S2 _{The km} is calculated, and the adaptive filter coefficient W _mi is calculated based on the _km according to the equation (4) (steps S3 and S4). At this time, the divergence suppression coefficient β is “1” set in the initialization process.
Is maintained and the term including the divergence suppression coefficient β that determines the reduction amount of the filter coefficient is the drive signal y in the normal state.
_{Since m} is a small value, it is a small value.

As described above, the filter coefficient W _mi in the adaptive digital filter processing according to the LMS algorithm
Are sequentially updated, the loudspeaker 5a, the sum of squares of the evaluation function J, that is, the residual noise detection signal and the sum of squares of the drive signals to the loudspeakers 5a and 5b, are minimized. Drive signals y ₁ and y for 5b
₂ is calculated, and this is supplied to the loudspeakers 5a and 5b via the D / A conversion circuits 27a and 27b and the analog switches 28a and 28b. Therefore, the control sounds corresponding to the driving signals y ₁ and y ₂ are emitted from the loudspeakers 5a and 5b, and the control sounds interfere with the noise, whereby the evaluation function J, that is, the residual noise detection signals e of the microphones 6a to 6c.
_The muffling control is performed so that the sum of the square sum of _{1 to} e _{3 and} the square sum of the drive signals y _{1 and} y ₂ is minimized.

On the other hand, in the filter coefficient correction processing after step S7, in the normal state where the system is in the stable state, the evaluation function J is controlled so as to be the minimum as described above. _{Since P} has a small value, W _P <A, and the timer interrupt process is ended as it is from step S8, the divergence suppression coefficient β is not updated, and the state of β = 1 is held.

However, when the noise suppression control has a diverging tendency that is approaching an unstable state, the filter coefficient W _mi does not converge to a small value and becomes a relatively large value, so the W power W
_{When P} becomes equal to or higher than the divergence prevention threshold value A, the β update counter is incremented (step S9), and then, if the reset timer T is not counting, the timing is started (step S).
10 to S12). Then, it is determined whether or not the reset timer T has timed out (step S13). In this way, the reset timer T is provided because W power is temporarily increased due to disturbances other than divergence, for example, when the door is closed with a loud noise and when the train passes through the guard. This is because the time for accumulating the β update counter is divided for the purpose of preventing erroneous determination as divergence.

Therefore, if the reset timer T has timed out, the β update counter is cleared (step S14), and then the reset timer T resets the flag 1 which is set during timing, and then the timer interrupt processing is performed. finish. On the other hand, when the reset timer T has not timed out, it is determined whether the count value N of the β update counter is equal to or greater than the threshold value C as the number of times the W power W _P is equal to or greater than the divergence prevention threshold value A (step S16). .. That is, when N ≧ C, the system is considered to have become unstable. Therefore, if N <C, it is not unstable yet, so the timer interrupt processing is ended as it is, but if N ≧ C, it is determined that the system has become unstable, and the divergence suppression coefficient β is set. Update to twice the current value (β = β × 2) (step S17), and filter coefficient W _mi
By doubling the amount of decrease, smaller loudspeakers 5a, the output of 5b to reduce the driving signals y _1, y _2.

As described above, even if the divergence suppression coefficient β is doubled, if the number of times the W power W _P is equal to or more than the divergence prevention threshold A is equal to or more than the set value C, the divergence suppression coefficient β
The output of the loudspeakers 5a and 5b is suppressed by repeatedly doubling, but when the divergence suppression coefficient β reaches the set value D, that is, when the number of updates of the divergence suppression coefficient reaches a predetermined number,
In the divergent state, it is determined that it is difficult to suppress it, the control signal CA is set to the logical value "0", the analog switches 28a and 28b are turned off, and the loudspeaker 5 is turned on.
The generation of control sounds from a and 5b is stopped. As a result,
Generation of abnormal noise due to divergence can be suppressed.

By stopping the vehicle, turning off the ignition switch 10, and then turning on the ignition switch 10 again, the divergence suppression coefficient β is reset to "1" in the initialization process. As described above, in the divergence suppressing process, the reset timer T and the β update counter N are provided, and the β update counter N is cleared at regular intervals after the β update counter N is started. The divergence suppression coefficient β is not updated due to the increase in the filter output in the adaptive digital filter processing that occurs sporadically over time, and conversely, the divergence is suppressed when the filter output in the adaptive digital filter processing frequently increases. When it is necessary, the divergence suppression coefficient β can be effectively updated.

In the above embodiment, the case where the divergence suppression coefficient β is doubled as the filter coefficient correction means has been described, but the invention is not limited to this, and the divergence suppression coefficient β is set to a constant value or the number of updates. A value that increases accordingly may be added, and the point is to increase the decrease amount of the filter coefficient when the divergence tendency continues.

In the above embodiment, the case where the divergence suppressing coefficient β is maintained at the cumulative value while the ignition switch 10 is kept in the ON state has been described.
Not limited to this, the divergence suppression coefficient β may be decreased when the divergence suppression coefficient β is not updated for a predetermined time or longer. Furthermore, in the above-described embodiment, the case where the loudspeaker is applied as the control sound source has been described, but the present invention is not limited to this, and a vibrator may be applied, and the microphone as the residual noise detecting means is also accelerated. A vibration sensor can also be applied.

Furthermore, the number of loudspeakers and microphones to be installed is not limited to the number in the above embodiment, but may be any number of 1 or more. Further, in the above embodiment, the case where engine noise is suppressed has been described, but the present invention is not limited to this. Detecting vibration input from the road surface to the wheels to suppress road noise, and detecting vibration of window glass. Thus, the wind noise can be suppressed, and the composite sound of these can also be suppressed.

Further, in the above embodiment, the case where the microcomputer 26 performs the adaptive digital filter processing and the digital filter processing of the filter coefficient according to the model space transfer function has been described. However, instead of these, an independent adaptive digital filter processing is performed. Also, a digital filter can be applied. Furthermore, in the above embodiments, the case where the present invention is applied to a vehicle has been described, but the present invention is not limited to this, and can be applied to the case of reducing noise in the interior of an aircraft.

[0048]

As described above, according to the active noise control device of the first aspect, the filter coefficient updating means for updating the filter coefficient of the adaptive filter means filters according to the filter output of the adaptive filter means. Since the filter coefficient correction means for reducing the coefficient is provided, when the divergence tendency is generated, the filter coefficient is reduced to suppress the control sound output from the control sound source to a small level, and the abnormal sound is output from the control sound source. The effect that can prevent it can be obtained.

According to the active noise control device of the second aspect, the filter coefficient correction means sets the filter coefficient each time the number of times the power based on the output of the adaptive filter exceeds the divergence prevention threshold becomes a predetermined value or more. Since the decrease amount increasing means for increasing the decrease amount is provided, when the divergence tendency continues, the decrease amount of the filter coefficient sequentially increases, so that the abnormal sound is more reliably output from the control sound source. The effect that can be prevented is obtained.

Further, according to the active noise control apparatus of the third aspect, the counting means for counting the number of times the power based on the output of the adaptive filter exceeds the divergence prevention threshold is cleared every predetermined time, so that the door is large. It is possible to prevent the noise suppression effect from being lowered by falsely determining that a disturbance that occurs sporadically over a long period of time by closing with sound or passing through a railway guard as a divergence state.

[Brief description of drawings]

FIG. 1 is a functional block diagram showing a basic configuration of the present invention.

FIG. 2 is a schematic configuration diagram of an example.

FIG. 3 is a block diagram of a controller according to the embodiment.

FIG. 4 is a flowchart of an example.

[Explanation of symbols]

 2 Cabin 3 Engine 5a, 5b Loudspeaker 6a-6c Microphone 7 Crank angle sensor 10 Ignition switch 15 Controller 26 Microcomputer 28a, 28b Analog switch

Claims (3)

[Claims] 1. A reference signal generating means for generating a reference signal according to a noise generation state of a noise source, an adaptive filter means for adaptively filtering the reference signal of the reference signal generating means and supplying it to a control sound source, and observation. Residual noise detecting means for detecting residual noise at a position, and a filter for updating the filter coefficient of the adaptive filter processing by the steepest descent method based on the reference signal of the reference signal generating means and the residual noise detection value of the residual noise detecting means In the active noise control device including coefficient updating means, the filter coefficient updating means includes a filter coefficient correcting means for reducing a filter coefficient according to a filter output of the adaptive filter means. Noise control device. 2. The filter coefficient correcting means counts the number of times the power based on the output of the adaptive filter exceeds the divergence prevention threshold, and the filter coefficient decreases every time the count value of the counting means becomes a predetermined value or more. The active noise control device according to claim 1, further comprising: a decreasing amount increasing means for increasing the amount. 3. The filter coefficient correcting means counts the number of times the power based on the output of the adaptive filter exceeds the divergence prevention threshold and is cleared at every predetermined time; and the count value of the counting means is a predetermined value or more. 2. The active noise control apparatus according to claim 1, further comprising: a decrease amount increasing means for increasing the decrease amount of the filter coefficient each time.