JPH06149272A - Sound insulating panel - Google Patents

Abstract

PURPOSE:To provide the sound insulating panel which is increased in sound insulating performance to a sound wave disturbance of an unsteady signal by improving the precision of the mean signal power estimated value of the arithmetic circuit of a driving part for active vibration control. CONSTITUTION:The active sound insulating panel consists of a sensor 4 for noise detection, an actuator 3 and a vibration sensor 2 which are mounted on a panel 1, and a driving part 5 which drives the actuator 3 according to the outputs of the sensors 2 and 4; and the driving part 5 is provided with a digital filter which estimates current signal power by smoothing the mean value of the squares of a signal including a past estimated value and a current signal. Therefore, the convergence of a high-performance active filter is obtained and the sound insulating panel which provides high sound insulation for the sound wave disturbance of the unsteady signal is obtained.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound insulation panel which is lightweight and thin and has high sound insulation properties.
It is used to improve the sound insulation of walls, windows, etc. of musical instrument playing rooms and audio rooms.

[0002]

2. Description of the Related Art In recent years, noise has become an important issue in the living and working environments due to the increase in noise generated by the development of transportation facilities and production facilities in factories, and the increasing density of housing complexes. ing. For this reason, various sound insulation panels have been used as means for preventing noise, but at present, satisfactory sound insulation performance has not been obtained.

Therefore, as shown in FIG. 1, the sound insulation panel 1
A vibration sensor 2 for detecting vibration of the panel 1 and an excitation actuator 3 are attached to the noise detection sensor 4 such as a microphone for detecting sound or vibration emitted from a noise source.
Then, the drive unit 5 for driving the actuator 3 performs an arithmetic operation on the electric signal of the noise inputted from the vibration sensor 2 and excites the panel 1 in a phase opposite to the noise, whereby the panel 1 It is conceivable that the vibration of the panel 1 is actively controlled so as to cancel the vibration to enhance the sound insulation.

[0004]

The control system of the sound insulation panel in the prior art is represented by the model shown as a block in FIG. In the figure, 6 is a noise transmission system, 8 is a control force transmission system, 71 is a digital filter, 72 is a control force transmission system model, and 73 is a control algorithm. The portion surrounded by the broken line is the arithmetic circuit 7 of the drive unit 5. The noise transfer system 6 is a transfer function that describes how the sound insulation panel 1 vibrates in response to a sound wave disturbance of an unsteady signal. Since this is a physically existing system, the noise transfer system 6 This is a unique function that depends on the material, thickness, area, etc., and cannot be freely manipulated. The input of the noise transmission system 6 is the sound wave disturbance of the unsteady signal detected by the noise detection sensor 4 such as a microphone, and the output of the noise transmission system 6 is the vibration of the sound insulation panel 1 detected by the vibration sensor 2. A panel with high sound insulation even in a passive state is a transmission system that can reduce the output amplitude even if the input amplitude of the noise transmission system 6 is large, but in reality, the sound insulation in the high frequency range is good. However, the sound insulation in the middle / low frequency range is poor, and the noise transmission system 6 is a kind of middle / low-pass filter. The digital filter 71 is a circuit that inputs a sound wave disturbance of an unsteady signal detected by the noise detection sensor 4 such as a microphone and calculates a control signal of the actuator 3 for canceling it, and a DSP (digital signal Processor). The output of the digital filter 71 is given to the control force transmission system 8.
This control system transfer system 8 is a transfer function that describes the transfer system in which the output of the digital filter 71 is amplified by the amplifying means of the drive unit 5 and converted into the vibration of the sound insulation panel 1 by the actuator 3. Since the system is a physically existing system, it is a unique function that depends on the characteristics of the actuator 3 formed of a piezoelectric element or the like, the mounting state on the panel 1, and the like. If the digital filter 71 is properly configured, the error signal ε (t) detected by the vibration sensor 2
Is the smallest. In order to minimize this error signal ε (t), the digital filter 7 is controlled by the control algorithm 73.
Each coefficient of 1 is controlled. The control force transmission system model 72 is a transfer function that simulates the transfer function of the control force transmission system 8. This does not physically exist, but exists in the arithmetic circuit 7 of the drive unit 5. The control algorithm 73 controls the coefficient W (t) of the digital filter 71 based on the signal S (t) output from the control force transmission system model 72 so that the error signal ε (t) is minimized. Is a calculation algorithm for.

Conventionally, when controlling a sound wave disturbance of an unsteady signal, as a control algorithm of the arithmetic circuit 7 of the drive unit 5,
L for sequentially estimating the signal power according to the equations (3) and (4)
An MS (Least Mean Square) algorithm (hereinafter referred to as "sequential signal power estimation type LMS algorithm") is used. However, the vector W
(T) = {w (1, t), w (2, t), ..., w (N,
t)} is the filter coefficient at time t, and the vector S (t) = {s (t), s (t−1), ..., S (t−).
N)} is the output signal sequence of the control force transmission system model 72 at time t, ε (t) is the error signal at time t, P
(T) ² is an estimated value of the average power of the signal S (t) at time t. u and α are parameters, and 0 <U <
1, 0 ≦ α << 1. N is the order of the filter.

[0006]

[Equation 3]

[0007]

[Equation 4]

However, in the conventional algorithm,
In the signal power estimation process in Equation 4, (t-1)
Estimated average signal power P (t-1) ² at time and time t
The average signal power estimation value s (t) ² obtained only from the signal s (t) at is internally divided by the parameter α and smoothed to obtain the average signal power estimation value P (t) ² at time t. Therefore, there is a problem that the estimation accuracy of P (t) ² is deteriorated.

The present invention has been made in view of the above circumstances, and improves the accuracy of the average signal power estimation value in the arithmetic circuit of the drive unit for active vibration control, thereby improving the accuracy of the acoustic wave of the unsteady signal. It is an object of the present invention to provide a sound insulation panel with improved sound insulation against disturbance.

[0010]

In the sound insulation panel of the present invention, in order to solve the above problems, as shown in FIG. 1, the vibration of the panel 1 and the panel 1 attached to the panel 1 is suppressed. A vibration sensor 2 for detecting, an excitation actuator 3 attached to the panel 1, a drive unit 5 for driving the actuator 3, a noise detection sensor 4 for detecting a sound or vibration emitted from a noise source, and the sensors 2, 2 shows a sound insulation panel for canceling the vibration of the panel 1 due to the noise by driving the panel 1 in a phase opposite to the noise by the driving unit 5 for the electric signal of the noise inputted from 4 to excite the panel 1 in a phase opposite to the noise. When the control algorithm 73 of the arithmetic circuit 7 of the drive unit 5 is composed of the sequential signal power estimation type LMS algorithm, the adaptive filter update algorithm sets the fill filter at the time t. Coefficient W (t) = {w (1, t),
w (2, t), ..., w (N, t)}, the output signal sequence of the control force transmission system model 72 at time t is S (t) = {s.
(T), s (t−1), ..., S (t−N)}, ε (t) is the error signal at time t, and signal S is at time t.
When the estimated value of the average power of (t) is P (t) ² , u, α are constants, and N is the order of the digital filter 71, it is a circuit that realizes the following equation.

[0011]

[Equation 5]

[0012]

[Equation 6]

[0013]

The time series signal s (i) as shown in FIG. 3 or FIG.
At (−∞ <i <+ ∞), the number of taps (N +
1) The number of digital filters is the time series signal s (i)
Of (-∞ <i <+ ∞), s (t-i) (i =
Since 0, ..., N) signals are handled, the time-series signals s (t−i) (i = 0, ...,) In the frame F (t) of length (N + 1).
When N) is cut out, “the cut out signal is a stationary ergodic signal, and the past and future signals of the cut out signal have the same probability distribution as s (t−i) (i = 0, ..., N). , Which is a stationary ergodic signal generated from [s (t)] (t−), where a signal generated from the same probability distribution as the time series signal s (t−i) (i = 0, ..., N) is [s (t)] (t− i)
(-∞ <i <+ ∞). However, [s (t)] (t
-I) = s (t-i) (i = 0, ..., N). At that time, the average power estimation value P ′ (t) ^{2 of the} signal shown in Expression 7 is obtained.
Is an unbiased variance value, so [s at time t
(T)] (t−i) (−∞ <i <+ ∞), which is an estimated value having the true average signal power unbiasedness, consistency, and validity, and is the signal s at time t in the conventional control algorithm.
From only (t), the signal [s (t)] (t
-I) Let s be the average power estimate for (-∞ <i <+ ∞)
This is a method with higher accuracy than the estimation method using (t) ² .

[0014]

[Equation 7]

Here, P '(t) ² is s (t-i)
(I = 0, ..., N) [s (t)] (t−
i) The signal power estimation value of (-∞ <i <+ ∞). However, when dealing with non-stationary signals, this estimate P '
(T) is not an estimated value that is time-invariant with respect to time t,
Since it varies, by using the smoothing filter shown in Equation 8, the signal s (i) (−∞ <i <+ ∞) is taken into consideration in consideration of the past estimated value P ′ (i) (i <t). The average signal power of is estimated.

[0016]

[Equation 8]

Therefore, in the control algorithm of the present invention, since the average signal power estimating means with high accuracy is provided, the convergence of the high-performance adaptive digital filter can be obtained, and the sound insulation panel with high performance can be obtained.

[0018]

FIG. 1 is a perspective view of an embodiment of the present invention. 1
Is a panel, which is arranged between the noise source and the sound insulation environment, and includes a vibration sensor 2 and an actuator 3. The vibration sensor 2 is provided to detect the vibration of the panel 1, and the actuator 3 is provided to excite the panel 1 in a phase opposite to that of noise. Reference numeral 5 denotes a drive unit, which includes an amplifier circuit for driving the actuator 3,
An arithmetic circuit for calculating the control force applied to the actuator 3 is provided. Reference numeral 4 denotes a noise detection sensor, which is composed of a microphone or the like and detects noise generated from a noise source. The noise source generates a disturbance sound wave having an unsteady signal.
In this embodiment, the arithmetic circuit of the drive unit 5 for active vibration control of the sound insulation panel 1 has an adaptive filter coefficient updating algorithm.

The structure of the arithmetic circuit 7 of the drive unit 5 is shown in FIG.
ing. The arithmetic circuit 7 includes a digital filter 71 and
Equipped with control force transmission system model 72 and control algorithm 73
The noise of the noise detected by the noise detection sensor 4.
The air signal is passed through the control force transmission system model 72, and the vibration sensor
In order to minimize the error signal obtained by
The coefficient of the digital filter 71 is set to the control algorithm 73.
More control. With this control algorithm 73
Is shown in FIG. 5 output from the control force transmission system model 72.
From the signal sequence s (i) (-∞ <i <+ ∞)
Cut at frame F (t) of length (N + 1) at t
Average of the output signal series s (t-i) (i = 0, ..., N)
Signal power P '(t)²Is calculated based on the formula of Equation 7,
As shown in 6, P ′ (t)²And based on the formula of Equation 2
P '(t) in the past at the obtained time (t-1)²
Average signal power estimation value P (t-
1) ²The value that internally divides and into α: (1-α) is P ′ (t)
²Is the estimated value of the average signal power at time t.
It Below, for time t + i (i = 1, 2, ...)
The operations 1 and 2 are repeated. However, P (0)²= 0
And In this way, the signal power estimation process is performed at the time (t
-1) average signal power estimation value P (t-1)
²And (N + 1) past signals s (t−i) (i =
0, ..., N)
By slipping, the average signal power at time t
Estimated value P (t)²High-accuracy average signal power
-High-performance adaptability due to having an estimation algorithm
The high-speed shielding of the digital filter is obtained.
It becomes a sound panel.

[0020]

As described above, according to the present invention, in the arithmetic circuit of the drive unit for actively controlling the vibration of the sound insulation panel to suppress the vibration, the successive signal power estimation which can be applied to the unsteady signal is also performed. In order to further improve the estimation accuracy of the average signal power estimation process in the improved LMS algorithm, the signal power at the current time is estimated by smoothing the root mean square value of the signal including the past estimated value and the signal at the current time. As a result, there is an effect that a high-performance adaptive filter can be converged, and a sound insulation panel having a high sound insulation property against a sound wave disturbance of an unsteady signal can be realized.

[Brief description of drawings]

FIG. 1 is a perspective view showing an embodiment of a sound insulation panel of the present invention.

FIG. 2 is a block diagram showing a control system of an embodiment of the sound insulation panel of the present invention.

FIG. 3 is an explanatory diagram showing a first signal sequence for explaining the operation of the present invention.

FIG. 4 is an explanatory diagram showing a second signal sequence for explaining the operation of the present invention.

FIG. 5 is an explanatory diagram showing a third signal sequence for explaining the operation of the present invention.

FIG. 6 is an explanatory diagram showing a method of calculating an average signal power estimated value according to the present invention.

[Explanation of symbols]

 1 panel 2 vibration sensor 3 actuator 4 noise detection sensor 5 drive unit

 ─────────────────────────────────────────────────── ─── Continuation of front page (72) Kenichi Yaoi, 1048, Kadoma, Kadoma City, Osaka Prefecture Matsushita Electric Works, Ltd.

Claims (1)

[Claims] 1. A panel, a vibration sensor attached to the panel for detecting vibration of the panel, an excitation actuator attached to the panel, a drive unit for driving the actuator, and a sound or vibration emitted from a noise source. In the sound insulation panel that cancels the vibration of the panel due to noise by exciting the panel in a phase opposite to that of the noise by the noise detection sensor and the electric signal of the noise input from each of the sensors, the driving unit performs calculation. When the arithmetic circuit of the driving unit is configured by the sequential signal power estimation type LMS algorithm, the adaptive filter update algorithm sets the filter coefficient at time t to W (t) = {w (1, t), w
(2, t), ..., w (N, t)}, the output signal sequence of the control force transmission system model at time t is S (t) = {s
(T), s (t−1), ..., S (t−N)}, ε (t) is the error signal at time t, and signal S is at time t.
Letting P (t) ^{2 be} the estimated value of the average power of (t), u and α be constants, and N be the order of the filter, [Equation 2] A sound insulation panel that is a circuit that realizes