JPH04505221A - Active sound/vibration cancellation device - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention] This invention relates generally to systems for controlling sound or vibration, and 7J to generate specially controlled sound or vibration fields. Multiple actuators are used and multiple actuators are used to measure the residual field. The present invention relates to an activity control system using sensors.

In contrast to previous systems directed to controlling periodic sound or vibrations, The system of this invention can be used even when the fundamental period of vibration is rapidly changing. can be used. For example, it controls engine noise inside a vehicle. can be used for

The improved method according to this invention is a series of methods for reducing multiple channel control systems. using the orthogonal transformation to a single-channel system and thereby providing a method. The output of each such system is vibration or the fundamental frequency of the sound source is changing. can be adapted to maintain high performance of the control system even when

The burden of invention The principles of active acoustics and vibration control have been known for many years and there is a wealth of Published literature exists. Most patent specifications in this field are applicable to specific situations. Regarding the method. The methods and systems described herein are periodic or near-periodic. Concerning advanced sound and vibration control. The two main strategies for this problem are: (1) Waveform shaping or 7 wave, such as U.S. Patent No. 4.506.380 and In UK Patent Application No. 25,201.858, issued by A reference signal containing one or more frequencies of sound and vibration is converted into a p-wave signal generate (7) and send it to the actuator, which in turn generates the desired sound or vibration. Occur.

(11) Waveform synthesis, in which the waveform generator is, for example, ．． 577.322 by a signal derived from a source.

The two methods are equivalent only when the vibration source is exactly periodic. If the source property is If changes are made in time, it is customary to use adaptive control systems and their Here, in the area to be controlled, the sensor detects residual sound or vibration I7 and provides information. is passed to the processor, which changes the f-wave coefficients or summation to give better control. change the created waveform.

Published UK Patent Application No. 2.201.858 describes a method for adapting filtering coefficients. Explain how. Need for compliance in British Patent Specification No. 1.577.322 was recognized and in a later patent specification, UK Patent Policy No. 2.107.960, was simply A simple technique for such a system using one actuator and sensor is The law is explained. This latter patent specifies how to shake the vibration when the period is changing. In this case, the transformation technique at the lowest expected frequency rather than at the frequency corresponding to the harmonic of the source. except that we propose that frequency components should be generated from That's right.

A further patent specification no. 2.122.052 is directed to vibration control. Skin shape synthesis technique use. In this method, sensors and actuators are placed at each of a number of locations. It is located in various places. This applies to systems with equal numbers of sensors and actuators. The method for yielding stems and adapting the waveforms in this particular case is presented against. In most applications, however, the source and sensor are are not located relative to each other and make better measurements of the resulting sound or vibrations More sensors than sources are typically used in the effort.

invention The theoretical background to this invention will be fully explained.

The referenced numbered mathematical equations are presented in the accompanying drawings.

The signals from each of the plurality of sensors are associated with the position of the source during that cycle. By using an analog-to-digital converter (ADC) driven by sampled. The data can be averaged over several cycles and the accuracy improve. This gives an almost periodic sequence, with discrete (disc+eH) There is an orthogonal transformation such as Fourier transformation. may be applied.

This method is well known for the analysis of periodic signals and is known as "ordinal ratio analysis" or "ordinal ratio analysis". This is called "order-locked analysis."

The sample signal from the i-th sensor is given by equation (3,1), where Then I++(nT) is the sensor due to the impulse at the jth controller output. is the response at i, x+(m) is the mth value of the jth controller output, and 7 + (n) is the nth value of the sense signal when there is no control, and T is is the sampling interval.

J is the number of controller outputs. If the length of the impulse response is smaller than the sampling period A slightly more complex formula should be used if it can be contrasted with a significantly changing time. It is. If r+ is sampled 8 times per cycle, then XI is periodic Since, equation (3,2) can be applied, where NT is the fundamental period.

Equation (3,1) can then be written as equation (3,3) , where equations (3, 4) define a periodic impulse response.

Orthogonal transformations can be used to simplify equation (3,3).

This example is a discrete Fourier transform defined by equations (3, 5). , where f=1/NT is the fundamental frequency.・ Equation (3,3) then becomes equation (3,6).

Since R+, Yt and xl are sampled exactly a number of times per cycle, It will be noted that they are independent of frequency f. Equation (3,6 ) indicates that each harmonic of the system, k, can be considered separately.

The control problem is to find the component x1(k) that produces the desired value of R+(k). be. All of the control outputs Xj (k) are interconnected to generate their respective sensor signals. This problem is complex because However, the combined equation (3,6) It is possible to use techniques to transform the set of into a set of independent equations. technique uses the single value decomposition of the transfer function matrix A++(kf) for each kf. Ru.

This gives equation (3,7), where asterisks indicate compound connections.

A matrix with composite components UIm and Vml represents an orthogonal transformation and the equation (3,8) and (3,9), where M is and 61m is the Kronecker force - CKtonecker) delta. . The term Dm(kf) is the mth single value at frequency kf. That is the actual scene.

The decomposition method was published in 1986 by Cambridge University Press (Camb+i). WH blur on pages 52 to 64 of dge Uniyersit7 Ptes+) “The secrets of numerical values – techniques of scientific calculation” (Nume “1” by Press et al. cal recipes-the ar tol 5 cienji c a II + p II + ing”). Equations (3, 6) include U ``tL multiplied and summed by i to give equation (3,10), equation (3,1 0,1) and (3,10,2) and (3,10,3) are applied to it It is possible.

These quantities are called the principal components of the corresponding signal. Equation Inode Standard Compatibility Algorithm can be solved iteratively using the system. If the obvious dependence on L and kf disappears, Then, equation (3, 10) reduces to equation (3, 11).

If the goal is to make R as small as possible, one algorithm results in equation (3°12) at the nth step, where μ is real convergence It is a factor.

Equation (3,1) by using equations (3,11) and (3,12) 2,1) is given, and from equation (3,11), the result is equation (3, 12,2) occurs.

These can be combined to give equation (3,13), and this If equations (3, 14) can be applied, the algorithm is stable; This shows that optimal convergence is obtained when μD=1.

Therefore, equations (3, 15) are applied, i.e. different convergence factors are frequency and each principal component.

To realize this algorithm, the transfer function A + It is necessary to measure i, (kf). This is the initial start-up or calibration phase and, if necessary, as described in British Patent Application No. 8825074゜1. Use the parameter estimator (estimate+) as explained. can be adapted. Transformation matrix U (k f) Oyo UV (k f) and A single value Dt (kf) is calculated from the measured transfer function and is Stored against During operation, appropriate transformation matrices and single values are used. The frequency f (or its equivalent period T) is measured so that the frequency f (or its equivalent period T) can be measured. Against that Since kf is unlikely to exactly match the value of the transfer function measured using new value is used. Alternatively, interpolation between nearby values can be used to gain further accuracy can be used. To maintain a given accuracy, the former method uses more memory and the latter uses more computation time.

Once X(kf) is found, equations (3,9) and (3,10,3) become can be used to give equation (3,16).

It is possible to apply the inverse discrete Fourier transform to obtain XI(n) It is Noh. These control signals are sent to a digital-to-analog converter (DAC), It is then passed on and amplified to provide a drive signal to the actuator.

In some applications it is desirable that the actuator not be overdriven. , and that the signal to the DAC signal is within the correct range. Drive One particular method to limit the dynamic amplitude is the minimization constraint, λ, according to equation (3, j-7). can be used in the given algorithm.

The constraint λ can be adapted after each interaction, i.e. if either of the drive signals, If they are too high, λ is increased or decreased if they are all within the desired range. .

Description of examples The invention is illustrated with reference to the accompanying drawings, in which a single The drawing shows one embodiment of an apparatus for implementing the method.

The digital values are stored in a memory device (1), which is for example a FIFO device. could be. These values are sent to a set of digital-to-analog converters (DACs) (2) , they are stimulated 8 times per cycle by successive electrical pulses from Senna (3). Driven. These pulses are related to the position of the source in that cycle. D.A. The analog signal from C is passed through a signal conditioning device (4) and is applied to a number of actuators. - one resulting acoustic or vibratory filter that provides a drive signal for the motor (,5); The field is measured by a sensor (6).

The signals from these sensors are stored in a memory device so that the sensor signals approach the desired values. used to match the value stored in location (1). The sensor signal is passed through the joint device (7) and driven by the signal from the position sensor (3). The signal is sampled synchronously with the source using an analog-to-digital converter (8).

, : these sampled values are located in the memory device (9) and are stored in a large number of complete 1, which reduces the effects of source-irrelevant signals that are averaged over multiple cycles. Joule (10) uses di-Zucreet Fourier transform to obtain 12, and each 1. The main component of the sensor signal, which generates the harmonic frequencies of the source and related components to the sensor. The components from the different sensors are then passed to a conversion module (11) to generate Hey tL is combined. Each of these independent components has a matching module 2, - rule (12 ) to generate the principal components of the new drive signal. These are conversion modules (13) to generate the frequency components of each drive signal, which are is converted into a time value via an inverse conversion module (15). The new time value is Return them to Mori Soso W(]). For conversion module (IJ) and (13) and adaptation module (12) requires knowledge of the period or frequency of the source. this This can be obtained from the position signal via a frequency counter (4) containing a real-time clock. Ru. This method may be used in aircraft configurations, where the source of the noise is Pella or prop fan (p+opfan) tF.

An important application of the method of activation control described above is the noise-related effects on vehicles. This is for engine control. When the inside of a car "goon and when ηJ ( UK Patent Application No. 2.20 issued for a control system for controlling a boom ], No. 858. It uses the wave shaping or February wave technique described above. use. The system has a delay that is correlated to the acoustic propagation time from the actuator to the sensor. It is designed to fit on a comparable time scale. Inside of my car j2-o However, there are sounds from many sources, which the 'L engine is not related to. For example, road noise, wind noise, sounds from the entertainment system in the car. etc. This noise contaminates the sensor signal and degrades system performance. .

The inventive method uses a signal sampled synchronously through a number of cycles. using averaging. This reduces one novel of contamination and improves system performance. improve.

However, the time taken for averaging is limited by changes in engine speed and load. reduce the system's ability to detect changes in the acoustic field due to . Therefore, for a given level of contamination noise, the optimal number of averaging There will be a cycle and it will depend on the engine speed and rate of change of load. cormorant. The rate of change in engine speed can be obtained from the position signal and The load is applied to the injection (inle+) manifold or the pressure on the throttle position sensor. It can be obtained from additional sensors such as pressure sensors. This information is can be used to control the rate of adaptation so that optimal performance of the system is obtained. this child And is needed! “Buzzing” when the dense acoustics are much louder than the contaminating noise. Rather than "sounding", it is possible to obtain back performance over a range of operating conditions. make it possible to

Most modern car engines are computer controlled; engine management Use the system. Some of the systems include the activation control system and the engine. Can be used for both management systems. Furthermore, the same microprocessor Can be used to control the system.

x, (m) = x, (m + N) Jj a, (nT-mT) = E I, (nT-mT-kNT) (3-3) Rp( kf) = Yt(kf)” D2(kf) %(kf) (3, 10) = n -n, -L ~ n-1 (3, 12) international search report Axis +lIl11mles-mu @t”ll”N・PCT/CF+9010O1’i 17 International Search Report GB 9000617 S^　36458

Claims (10)

[Claims] 1. periodic sound produced by a source with a varying fundamental period of vibration or An active acoustic or vibration control system for controlling vibrations that uses distributed multiple Several noise or vibration sensors and multiple distributed noise or vibration compensation applications actuator and at least one analog for sampling the sensor signal. associated digital converter (ADC) and the position of the source in its cycle and signals used to drive at least one ADC multiple times per cycle a source sensor for generating a signal, and the output of at least one ADC is activated. A system used to generate drive signals for the controller. 2. The memory means is used to store digital values and the memory means is The stored value is extracted from the output of the noise or vibration sensor through at least one ADC. continuously adapted or modified by the sampled signal obtained by A system according to claim 1. 3. A system according to claim 2, wherein the memory means is a FIFO device. 4. The sampled values of the noise and vibration sensors are located in the memory device. and averaged over a plurality of cycles. stem. 5. The signals output from the memory device are utilized in signal analysis and processing means. Claim 2 or Claim 3 or Claims wherein the value given to the memory means is generated. A system that follows 4. 6. Signal analysis and processing means detect source harmonics for each noise and vibration sensor. a first conversion module that generates a frequency-related component and a main independent component of the sensor signal; a second conversion module that generates a main component of the sensor signal; and a second conversion module that is driven from the sensor signal main component. an adapted processor for generating a principal component of a signal and each from said driving signal principal component; a third conversion module for generating frequency components of the drive signal of the claim. A system according to section 5. 7. Signal analysis and processing means also perform inverse transformations to generate time values from frequency components. according to claim 6, comprising a conversion module, wherein said time value is provided to memory means. system. 8. The period or frequency of the source is determined by the frequency counter incorporating the clock of the source sensor. The source period or frequency signal obtained from the output is transmitted to a second and third conversion module. and to an adaptation processor for controlling generation of frequency components. or a system according to claim 7. 9. The output signal from the memory means is small for generating the actuator drive signal. 2 or 3, wherein the at least one digital-to-analog converter (DAC) A system according to any of claims 8. 10. The signal analysis and generation means are based on equations (3.1) to (3) mentioned above. ．． 16) or (3.17).