JP2929565B2 - Synchronous filter device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention is used for canceling noise synchronized with an engine such as vibration or noise in the interior of an automobile, and eliminating the periodic noise while synchronizing with the noise. The present invention relates to a synchronous filter device.

[0002]

2. Description of the Related Art Synchronous filter devices are designed to eliminate periodic noise while synchronizing it with the period of the noise. Therefore, even if the period of the noise changes, the filter responds adaptively to the change. , DSP), a normal FIR (Finite Respo)
(nse Filter) does not require many product-sum operations in the filter operation and can be configured very simply. Therefore, the filter is widely used particularly as a filter for canceling periodic noise.

Hereinafter, a conventional synchronous filter device will be described with reference to the drawings. FIG. 4 is a schematic block diagram of a conventional synchronous filter device. In FIG. 4, 41 is a pulse generator, 42 is a synchronous filter,
43 is a filter coefficient correction means.

Next, the operation of the conventional synchronous filter device will be described. In FIG. 4, a pulse generator 41
Represents the period of the input periodic noise by the synchronous filter 4
The pulse is divided by the number N of taps of 2 and a pulse is generated so as to perform exactly N samplings within the period of the input noise. Even when the period of the noise changes, the pulse period also changes according to the change of the period. The synchronous filter 42 is a filter that sequentially outputs the tap values of the filter in synchronization with the pulse generated by the pulse generator 41. For example, if it is realized by a DSP, it is constituted by a memory having only a tap length. After outputting to the end of the filter, return to the first tap of the filter and output again.
The output of the synchronous filter 42 is added to the periodic noise, and based on the addition error, the filter coefficient correction means 43 reduces the addition error by the filter coefficient correction means 43.
The value of the corresponding tap is updated. For example, when the time is k and the number of taps of the filter is N, the position j of the corresponding tap coefficient can be expressed by the following equation (1).

J = k mod N (mod N is an integer modulo N) (1) Here, a tap coefficient at time k is Wk, and an addition error is e.
(K) Assuming that a step size parameter corresponding to a gain is u, the tap coefficient Wk can be expressed by the following equation (2).

Wk (j) = Wk−N (j) + 2ue (k) (2)

With the above operation, a synchronous filter device that cancels out periodic noise can be realized. (For example, CAI Digital Signal Processing by Corona Co., Hidefumi Obata, pp. 134-136, or Transactions of the Society of Instrument and Control Engineers, Vol. 19, No. 3, page 34-
(Refer to “Synchronous adaptive filter” on page 39.)

[0008]

However, in the above configuration, when a disturbance that disturbs the periodicity such as random noise is added to the input and enters, the value of each coefficient of the synchronous filter is affected by the influence. Is disturbed and errors accumulate,
The step size parameter, that is, the gain of one filter tap coefficient update, cannot be set large.
Conversely, if the gain is suppressed, for example, when the amplitude of the periodic noise changes, there is a problem that the response of the filter is slow because the gain is small, and the convergence performance in the transient state deteriorates.

SUMMARY OF THE INVENTION In view of the above problems, it is an object of the present invention to provide an excellent synchronous filter device capable of minimizing a malfunction caused by disturbance and improving transient characteristics with respect to periodic noise. .

[0010]

In order to solve the above-mentioned problems, a synchronous filter device according to the present invention focuses on the periodicity of noise, and calculates the error between the waveform to be eliminated in that period and the output waveform of the synchronous filter. A means for changing the gain of the entire filter to one ratio so as to minimize the sum of squares is newly added.

[0011]

According to the present invention, it is possible to realize a synchronous filter which is less affected by disturbance and has improved transient characteristics with respect to periodic noise.

[0012]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a schematic block diagram of a synchronous filter device according to a first embodiment of the present invention. In FIG. 1, 11 is a pulse generator, 12 is a synchronous filter,
13 is a filter coefficient correction means, 14 is a periodic pulse generation means, and 15 is a filter gain correction means.

Next, the operation of the synchronous filter device according to the first embodiment will be described. In FIG. 1, a pulse generator 11 divides the period of the input periodic noise by the number of taps N of the synchronous filter 12, and generates pulses so that exactly N samplings can be performed within the period of the input noise. I do. Even when the period of the noise changes, the pulse period also changes according to the change of the period. The synchronous filter 12 sequentially outputs the tap values of the synchronous filter 12 in synchronization with the pulse generated by the pulse generator 11. For example, if it is realized by a DSP, it is constituted by a memory having only a tap length. When the output to the end of the synchronous filter 12 is completed, the process returns to the first tap again and outputs. The output of the synchronous filter 12 is added to the periodic noise, and based on the addition error, the value of the corresponding tap of the synchronous filter 12 is updated by the filter coefficient correction means 13 so that the addition error is reduced. Go. For example, when the time is k and the number of taps of the filter is N, the position j of the tap coefficient is determined as in the above equation (1).
Let Wk be the tap coefficient, e (k) be the addition error, and u be the step size parameter corresponding to the gain, and then update according to the relationship expressed by equation (2).

Next, the value of the periodic noise, which is the input waveform, and the value of the output of the synchronous filter 12 are stored for one cycle, and are generated by the periodic pulse generating means 14 once in one cycle of the periodic noise. One cycle of the square of the difference between the output waveform of the filter and the input noise waveform when it is assumed that the N-tap of the synchronous filter 12 is multiplied by a certain rate of gain g using a prescribed algorithm every time a pulse is input. For the sum of
The gain g is determined so that the sum is minimized. The filter gain correction means 15 determines the gain g as described above, multiplies each tap of the synchronous filter 2 by a factor,
A new tap coefficient is used.

As described above, according to the first embodiment,
Even if the amplitude of the input noise changes, as long as the periodic waveforms are similar, there is a one-cycle delay in the gain change, but it can respond steadily, thus improving the transient characteristics of the periodic noise. Can be. In addition, even when the step size parameter shown in the above equation (2), that is, the gain of one filter tap coefficient update is set to be small for the case where aperiodic noise is included, the gain g is equal to the step parameter. Since the convergence speed can be determined independently, the convergence speed can be increased.

Next, the algorithm for determining the gain used in the first embodiment will be described. The input noise at the discrete time i is Ai, and the output of the synchronous filter 12 is Bi. During one cycle, the synchronous filter 12 outputs N taps, and the difference between the waveform of the filter and the input noise waveform when the tap is multiplied by the gain g is 2
When the sum of one cycle of the power is calculated and the gain g is determined so that the average is minimized, the following equation (3) holds.

[0017]

(Equation 1) When g that satisfies the equation (3) is determined, the value of the tap of the synchronous filter 12 is updated every cycle using the g, as shown in the equation (4). Wi = Wi * g (i = 1,..., N) (4) Here, if the left side of Expression (3) is replaced with f, the following Expression (5) is obtained.

[0018]

(Equation 2) Then, when g satisfies the expression (5), f has a minimum value and a minimum value with respect to g.
Equation (6) holds. ∂f / ∂g = O (6) Next, equation (7) is rearranged by substituting equation (5) into equation (6).

[0019]

(Equation 3) Therefore, if g is determined according to equation (7), the sum of one cycle of the square of the difference between the waveform formed by the filter and the input noise waveform when it is assumed that the tap is multiplied by the rate gain g is: That value can be minimized.

Further, if the difference between Ai and Bi is ei,
Equation (8) holds. Ai = Bi + ei (8) This is substituted into Expression (7) to obtain Expression (9). Even if the input noise Ai cannot be measured and only Bi and ei can be measured, there is an advantage that the gain g can be determined by Expression (9).

[0021]

(Equation 4)

Next, a second embodiment of the present invention will be described with reference to FIG.
This will be described with reference to FIG. In FIG. 2, 21 is a pulse generator, 22 is a synchronous filter, 23 is a filter coefficient correction unit, 24 is a periodic pulse generation unit, 25 is a filter gain correction unit, and 26 has a transfer function P and oscillates an electric signal. An actuator for converting to an output, 27 is a control compensation filter, 28 is an actuator simulated transfer function filter having a transfer function P, 29 is a vibration-electric signal converting means for converting a mechanical vibration of a G sensor or the like into an electric signal, and 30 is a transmission. 9 is an actuator simulated transfer function filter having a function R.

Next, the operation of the second embodiment will be described. In FIG. 2, an actuator simulated transfer function filter 28 simulates a transfer characteristic P of an actuator 26, in which an impulse response (transfer characteristic P) sampled and converted into a pulse sampling period of the pulse generator 21 is generated. Is done. The filter coefficient correction unit 23 compensates for the addition error between the input noise obtained as the output of the vibration-electric signal conversion unit 29 and the output of the actuator 26 by using the output of the actuator simulated transfer function 28 to delay the actuator 26, and compensates for the delay. The tap coefficient of the synchronous filter 22 is corrected with the output. Specifically, at time k and the number of filter taps N, the position j of the tap coefficient is determined by equation (1), the m-th tap coefficient at time k is Wk (m), the addition error is ek, the actuator simulation is performed. The impulse response generated by the transfer function filter 28 is represented by c
(M) (m = 1,..., N), and assuming that a step size parameter corresponding to the gain is u, the filter updating equation is as shown in the following equation (10). Wk + 1 (m) = Wk (m) + 2 * u * ek * c (j−m + 1) (m = 1,..., N) (10) For example, an algorithm called Filtered-X is applied to this content. Equivalent to. In this case, the control compensation filter 2
The transfer characteristic R of 7 may be simply 1.

Instead of performing the correction as described in the actuator simulated transfer function filter 28, the control compensating filter 27 uses the inverse function (R = 1/1) of the actuator 26.
P) may be configured (actually, in order to prevent oscillation, R = Y / P, Y is often configured as a transfer function for phase compensation), in which case the filter coefficient correction is performed. The means 23 may correct the synchronous filter 22 by an equation, for example, in the same manner as in the first embodiment.

The addition error between the periodic noise as the input waveform and the output of the actuator 26 and the
As the value to be output to 6, a value for one cycle in the past is always stored. On the other hand, the impulse response characteristics of the actuator 26 are measured in advance by system identification or the like, and are realized as an FIR filter by the actuator simulated transfer function filter 30. Next, the actuator 2
The output after passing through the actuator 26 is estimated by calculating the product sum of the data for the past one cycle of the value output to 6 and the actuator simulated transfer function filter 30.

Here, every time a pulse generated once in the period of the periodic noise by the periodic pulse generating unit 24 enters, the filter gain correcting unit 25 uses the algorithm described below, for example, The gain g is determined and synchronized so that the sum of one cycle of the square of the difference between the output after passing through the actuator 26 and the input noise waveform when the N taps are multiplied by the linear gain g is minimized. Each tap of the expression filter 22 is multiplied by a factor to obtain a new tap coefficient.

Hereinafter, this algorithm will be described. The input noise at discrete time i is denoted by Ai, the filter output is denoted by Bi, and the difference between Ai and Bi is denoted by ei.
i, the addition error of the periodic noise and the output of the actuator 26 is stored for one cycle in the past, and the value is represented by Wi.
(I = 1,..., N). On the other hand, actuator 2
6 is measured in advance by system identification or the like, and F
It is realized as an IR filter. When the impulse response is Vi (i = 1,..., N), S (W
i * Vn-i + 1), the output Bi after passing through the actuator 26 can be estimated. When Bi is obtained, g can be determined by equation (9).

Even if the input noise Ai cannot be measured and the output Bi of the actuator 26 cannot be measured, and only the addition error and the value to be output to the actuator 26 are known, the output of the actuator 26 is estimated by estimating the output of the actuator 26. The gain g can be determined so that the sum or average of the sum or average of one cycle of the square of the difference between the output after passing and the input noise waveform is minimized.

As described above, according to the second embodiment,
Even if the amplitude of the input noise changes, as long as the periodic waveforms are similar, there is a one-cycle delay in the gain change, but it can respond steadily, thus improving the transient characteristics of the periodic noise. Can be. Also, even if the step size parameter u in the equations (2) and (10), that is, the gain of one filter tap coefficient update, is set small for the case where aperiodic noise is included, Since it can be determined independently of the parameters,
There is also an advantage that the convergence speed can be increased.

Further, even when the actuator output to be added to the input noise cannot be directly obtained, the gain g can be determined by estimating the actuator output.

In the above-described second embodiment, the sum of one cycle is calculated. However, the sum may be an integral multiple of the cycle, or the sum or the average value of a sufficiently long time so that the influence of the cycle may be sufficiently ignored. In this case, it is possible to operate similarly well.

Next, a third embodiment of the present invention will be described with reference to FIG.
This will be described with reference to FIG. In FIG. 3, 31 is a pulse generator, 32 _{1 to} 32 _M are synchronous filters, 33 _{1 to} 33 _M
Is a filter coefficient correcting means, 34 is a periodic pulse generating means,
35 is a filter gain correction means.

Next, the operation of the third embodiment will be described. This embodiment is an example in which M synchronous filters and filter coefficient correcting means in the first embodiment are provided, and the number of input noises to be eliminated increases to P. The basic operation is as follows. because it is as already mentioned in the description of the first embodiment, erasing the input noise P number by synchronous filter the M, the synchronous filter 33 _to 333
_M gain Gj (j =, ..., M ) is the filter gain correction unit 35, synchronous filter j (j
=,. . . , M), the sum of squares of the difference between the filter output and the above-mentioned input noise waveform when the tap is multiplied by the amplification factor Gj is taken at P points, and the sum of the P square sums becomes smaller. Is determined as follows.

Here, the algorithm for determining the gain used in the filter gain correction means 35 will be described. The k-th input noise at discrete time i is A _k (i), and the component for eliminating the k-th input noise by the j-th synchronous filter 32 _j is B _jk (i). 1 synchronous filter 32 _j between the period and outputs N taps, the output waveform and the input noise waveform of the filter when it is assumed that multiplied by the Ichiritsu gain g _j to the tap of the j-th synchronous filter 32 _j Difference 2
When the sum of one cycle of the power is calculated and the gain g is determined so as to minimize the sum, the following equation (11) is established.

[0035]

(Equation 5) If g that satisfies equation (11) can be determined, the g
, The tap value of the j-th synchronous filter 32 _j is updated every cycle as shown in Expression (12). W _j (i) = W _j (i) * g _j (i = 1,..., N) (12) When the left side of equation (12) is replaced by f, the following equation (13) is obtained. )become that way.

[0036]

(Equation 6)And now, the a-th synchronous filter 32_aGain g
_aAsk for. f is g_a( _a= 1,. . . , M)
Since it is a small value and a minimum value, g of f_aThe partial derivative to is
0, and equation (14) holds. ∂f / ∂g_a= 0 (14) Next, when equation (13) is substituted into equation (12), equation (15)
Become like

[0037]

(Equation 7) Here, when each A _k (i) is approximated by the following equation,

[0038]

(Equation 8) Substituting equation (16) into equation (15) gives equation (1)
It becomes like 7).

[0039]

(Equation 9) Here, if g _k other than g _a is close to 1 or a constant value, g _a is obtained by the following equation (18).

[0040]

(Equation 10) Therefore, if g _a is determined according to equation (18), the tap of the a-th synchronous filter 2 is multiplied by the percentage gain g _a to obtain the difference between the output waveform of the synchronous filter 32 _a and the input noise waveform. The sum of one square cycle can be minimized. Further, when g _k other than g _a is largely apart from 1, there may be a case where normalization is performed, for example, as in the following equation (19). g _k = 1 + (g _k −1) / M (k = 1,..., M) (19)

[0041]

As described above, according to the present invention, even if the gain of the step parameter of the synchronous filter is reduced so that the influence of disturbance is reduced as much as possible, it is possible to quickly cope with the gain fluctuation of the periodic noise, There is an effect that characteristics can be improved. In particular, it has a transfer characteristic P,
Actuator that converts the output of a synchronous filter into a vibration output
Error between input data, input noise and actuator output
A vibration-electric signal conversion means for converting
Vibration-to-electrical signal converter compensated for delay caused by tutor
Compensates tap coefficients of synchronous filter based on output of stage
Filter coefficient correction means and an
The product of the value output to the tutor and the data for the past one cycle
To obtain the sum and estimate the output after passing through the actuator.
Tutor simulated transfer function filter and one round of input noise
Each time a pulse is input, a synchronous filter
When taps are multiplied by a predetermined amplification factor.
Sum of squares of difference between output and input noise after passing through actuator
And a synchronous filter so that the sum of squares becomes smaller.
Filter gain correction means for determining the amplification factor of
Therefore, the influence of disturbance is small and periodic noise
To realize a synchronous filter device with excellent transient characteristics
Can be.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram of a synchronous filter device according to a first embodiment of the present invention.

FIG. 2 is a schematic block diagram of a synchronous filter device according to a second embodiment of the present invention.

FIG. 3 is a schematic block diagram of a synchronous filter device according to a third embodiment of the present invention.

FIG. 4 is a schematic block diagram of a conventional synchronous filter device.

[Explanation of symbols]

 11, 21, 31 Pulse generators 12, 22, 32 Synchronous filters 13, 23, 33 Filter coefficient correction means 14, 24, 34 Periodic pulse generation means 15, 25, 35 Filter gain correction means 26 Actuator 27 Control compensation filter 28 Actuator simulated transfer function filter 29 Vibration-electric signal conversion means 30 Actuator simulated transfer function filter

──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Takeshi Okada 4-3-1 Tsunashima Higashi, Kohoku-ku, Yokohama-shi, Kanagawa Prefecture Matsushita Communication Industrial Co., Ltd. (56) References JP-A-2-238707 (JP, A JP-A-2-207170 (JP, A) JP-A-59-174013 (JP, A) JP-A-6-110469 (JP, A) JP-A-6-149269 (JP, A) JP-A-6-149269 204801 (JP, A) Special Table Hei 4-500144 (JP, A) (58) Fields investigated (Int. Cl. ⁶ , DB name) H03H 17/00-21/00 G10K 11/16

Claims (4)

(57) [Claims] 1. The period of a periodic component of input noise is divided into N.
Pulse generator that generates pulses that can be sampled
And sequentially outputs N taps in synchronization with the pulse,
After passing through the element of the transfer characteristic P, the periodicity of the input noise
A synchronous filter for eliminating components, and a transfer characteristic P,
An act of converting the output of the synchronous filter into a vibration output
Between the input noise and the output of the actuator.
Vibration-electric signal conversion means for converting the addition error into an electric signal
And the vibration compensated for the delay by the actuator
The synchronous filter based on the output of the electrical signal conversion means;
Filter coefficient correction means for correcting tap coefficients of
Exceeding the value output to the actuator for each sampling
The product sum with the data for the last cycle is calculated
Actuator transfer function for estimating output after passing through
Filter and one pulse input per cycle of input noise
Each time the tap of the synchronous filter
Actuator passage assuming multiplication by amplification factor
Calculate the sum of squares of the difference between the output after and the input noise.
The gain of the synchronous filter is increased so that the sum of squares becomes smaller.
A synchronous filter device comprising: a filter gain correcting means for determining the gain . 2. A method for compensating a delay caused by an actuator.
The sampling period of the pulse of the pulse generator.
Actuator that generates the impulse response after the pulling transformation
The synchronous filter device according to claim 1 , further comprising a data simulated transfer function filter . 3. A method for compensating for a delay caused by an actuator.
For this purpose, a transfer function equivalent to the inverse function of the actuator is constructed.
The synchronous filter device according to claim 1, further comprising a control compensation filter to be formed . 4. At a discrete time i, the tap value of the filter exceeds
Output of last m time or past m of control compensation filter
The output of time is Q = (Qm, Qm-1,..., Q1)
And the same or similar to the transfer characteristic P
FIR filter P '= (P1, P
2,. . , Pm-1), the Q and P '
The output Vi ′ after P is estimated by the sum of products v, and the output Vi ′ and the input
From the difference Ei between the force noise Ui and the true P output Vi, one cycle
Alternatively , when the sum of n cycles is represented by S (),
((Ei * v'i) / (V'i * V'i) +1)
4. The method according to claim 1, wherein the rate of increase in
The synchronous filter device according to any one of the above .