EP1920230A1

EP1920230A1 - Method and device for evaluating the annoyance of squeaking noises

Info

Publication number: EP1920230A1
Application number: EP06764096A
Authority: EP
Inventors: Fawzi Attia; Michael Fischer; Winfried Keiper; Thomas Zoeller; Uwe Bretz; Mats Karlsson; Uwe Martin; Michael Raquet
Original assignee: Robert Bosch GmbH
Current assignee: Robert Bosch GmbH
Priority date: 2005-08-25
Filing date: 2006-07-06
Publication date: 2008-05-14
Also published as: DE102005052630A1; US20090296945A1; JP2009504507A; WO2007023017A1

Abstract

The invention relates to a method for evaluating the annoyance of squeaking noises within an audible signal generated during the operation of a motor vehicle or during the operation of one of its components. According to the invention, the existence of at least one squeaking noise is detected, this at least one squeaking noise is evaluated with regard to at least two predetermined characteristics, and a quantity characteristic of the annoyance of this at least one squeaking noise is determined from the at least two evaluations of this at least one squeaking noise.

Description

Method and device for assessing the annoyance of squeaking noises

State of the art

Noises that are produced during the operation of a motor vehicle or its components frequently prove to be annoying for the driver and the environment and should be detected as far as possible in order to then enable noise reduction measures.

From DE 102 60 723 Al a method for the suppression of switching noise in the test control of valves and pumps in the hydraulic system of a brake circuit is known. The control takes place over such a short time interval that no mechanical and noise-causing reaction of the controlled component takes place.

Advantages of the invention

The invention relates to a method for assessing the annoyance or disturbance or the degree of disturbance of squeaking noises or the annoyance of substantially montonal noise within a generated during operation of a motor vehicle or the operation of one of its components sound signal in which the presence of at least one squeaking noise is detected, this at least one squeaking noise is evaluated in terms of at least two predetermined features and from the at least two assessments of this at least one squeaking noise a the annoyance of this at least one squeaking noise characterizing size is determined.

The knowledge of an objective measure of the annoyance of squeaking noises makes it possible to decide whether the squealing noises are acceptable or whether Countermeasures are required. To assess annoyance, a size is determined which indicates how strongly or to what extent the squeaking area is perceived by the human ear as disturbing or unpleasant.

An advantageous embodiment of the invention is characterized in that the squeaking noises are brake squeaking noises. Squeaky brakes prove to be a major noise annoyance for both the environment and the driver.

An advantageous embodiment of the invention is characterized in that

- That the squeaking noise is detected by finding a maximum in an amplitude spectrum of the sound signal and that the at least two features are taken from the following feature list, which as characteristics - the length of time of the squeaking noise,

the maximum level value of the weighted amplitude spectrum,

the maximum level value of a weighted and smoothed spectrum obtained from the amplitude spectrum by smoothing and weighting,

- the specific loudness of the signal, - the product of the length of time of the squeak signal and the maximum

Level value of the weighted and smoothed spectrum,

the product of the time length of the squeak signal and the maximum level value of the weighted spectrum, and

- contains the product of the time length and the specific loudness of the signal.

An advantageous embodiment of the invention is characterized in that for each of the at least two features in the at least one squeaking noise, a single evaluation number is determined - from the at least two determined individual evaluation numbers the annoyance of

Squeaking characterizing size is determined.

The individual score indicates, for each squeal sound, how significant it is to identify annoyance for each feature.

An advantageous embodiment of the invention is characterized in that that the at least one squeaking noise is at least one brake squeaking noise occurring during a single braking operation,

- That is formed for each of the at least two features by linking the determined for the respective feature for each brake squeal noise ratings a feature number and

- From the at least two determined feature evaluation numbers a the annoyance of the squeaking noise occurring during the braking process characterizing Einzelunästigkeitsgröße is determined.

This summarizes all squeal events of a braking event and an objective measure, i. the solitary size, for the total annoyance of the

Squeaking during braking determined. In particular, it is taken into account that multiple brake squeal occurs during the same braking operation.

An advantageous embodiment of the invention is characterized in that the linkage is an addition, in particular a weighted addition.

An advantageous embodiment of the invention is characterized in that

- That from the at least two feature evaluation numbers by a weighted addition, a first intermediate size is determined and

- From the first intermediate size of the annoyance of the squeaking noise characterizing size is determined.

An advantageous embodiment of the invention is characterized in that the Einzelunästigkeitsgröße from the first intermediate size according to the relationship

bonisqueal = where a, b and c are selectable parameters, OV is the first intermediate size, and bonisqueal is the single nuisance size. Due to the selectability of a, b and c, there are 3 degrees of freedom available for the most objective and meaningful determination of bonisqueal. - A -

An advantageous embodiment of the invention is characterized in that the values a = 0.016, b = - 23.64375 and c = 2.6327 are selected for the selectable parameters a, b and c. These values have proven to be particularly suitable in experiments.

An advantageous embodiment of the invention is characterized in that

- the single lump size is rounded to the nearest integer value and

- the single lump size is set to the value 1, if for the lump size a value less than 1 is determined and

- that the single lump size is set to the value 10, if a value greater than 10 is determined for the lump size.

Thus, a classification of the individual annoyance sizes, i. a classification into discrete classes instead.

An advantageous embodiment of the invention is characterized in that - at least two braking operations are performed,

that a total annoyance variable characterizing the annoyance magnitude of the squeaking noises occurring during the braking processes is determined,

- Wherein the Gesamtlästigkeitsgröße received the mean value of the determined for each braking Einzelallästigkeitsgrößen, which meet a predetermined condition. This makes it possible to determine an objective measure of the annoyance of the squeaking noises of a number of braking operations.

An advantageous embodiment of the invention is characterized in that the predetermined condition is the undershooting of a threshold value, in particular of the threshold value 9.5, by the particular individual annoyance variable.

This clearly shows that extremely slight squeaking noises, which are almost imperceptible, are not taken into account. Concerning. the number 9.5 is referred to Fig. 1, where the annoyance of squeaking numbers are assigned.

An advantageous embodiment of the invention is characterized in that in the

Gesamtlästigkeitsgröße additional additively enters a term in which in turn the number of performed squeaky braking operations based on the total number of braking operations, ie the percentage of squeaky braking processes received. An advantageous embodiment of the invention is characterized in that the term is determined from the number of performed squeaky braking operations based on the total number of braking operations by means of a predetermined characteristic.

An advantageous embodiment of the invention is characterized in that it is a monotonically decreasing characteristic in the characteristic.

An advantageous embodiment of the invention is characterized in that

- the total noise magnitude is rounded to the nearest integer value and - the total noise magnitude is set to the value 1, if for the

Gesamtlästigkeitsgröße a value less than 1 is determined and

- that the Total Impact Size is set to the value 10 if the Total Impact Size is determined to be greater than 10.

Discrete numbers are available for the degree of annoyance.

Furthermore, the invention comprises a device containing means for carrying out the methods of the preceding claims.

The advantageous embodiments of the method according to the invention are also expressed as advantageous embodiments of the device according to the invention and vice versa.

drawing

The drawing consists of Figures 1 to 5.

FIG. 1 shows an evaluation scale in which the relationship between an evaluation number indicating the annoyance of a squeaking noise and the degree of annoyance is indicated.

Figure 2 shows in a schematic way the extraction of features from a

Squeaking.

3 shows by way of example a correction term in the ordinate direction. In this case, the size NP is shown in the abscissa direction, ie the ratio of squeaky braking processes and the total number of braking operations in%. FIG. 4 shows the frequency response of different evaluation filters as a function of the frequency.

Fig. 5 shows the basic sequence of the method according to the invention.

embodiments

The invention consists in a method for the objective evaluation of the annoyance of in particular brake-excited squeaking noises. The evaluation is carried out on a ten-part scale with discrete levels from 1 to 10. Mean

1 = very unpleasant squeaking,

10 = no squeaking perceptible.

The calculated index, also called "Brake Objective Noise Index Squeal" or "BONI-Squeal", has, in relation to the perceived annoyance, a high degree of agreement with human perception. After extraction of physical and psychoacoustic

Characteristics from the time signal of a squeaking noise is formed by combining these characteristics of the weighting index.

Such an index can be used for example in the application or the final acceptance of motor vehicle brakes. Here, vehicles are often driven by test drivers on defined test tracks and brake noise, especially squeaking, assessed subjectively. This can lead to strong deviations between the ratings of different drivers as well as between the ratings of the same driver, although physically identical squeak signals are present. The invention makes it possible, by processing recorded airborne sound signals, to calculate an evaluation index which corresponds to the average perceived annoyance of the noise. This rating index already allows a reliable and objective statement about the noise quality of a brake during the application phase. The high correlation of the Evaluation index of the average annoyance sensation of humans was proven in extensive hearing tests.

The method provides a rating index for the annoyance of especially decelerated squealing noises, which can take values of 1 to 10 on an ordinal scale. The individual values have the meanings shown in FIG. 1, with a higher value indicating lower annoyance in each case.

In this case, possibly existing squeaking noises are determined in a recorded airborne sound signal x (t). In practice, x (t) may be e.g. a microphone signal from the

Be vehicle interior. First, the squeaking noises in x (t) must be recognized by a suitable method and described according to their frequency-time structure.

After the analysis of x (t) by such a method, the following quantities are present for each detected squeak event q with q = 1, 2,..., Nq: start time tqstart and end time tqend of the squeak signal q,

Center frequency fq of the squeak signal q, airborne sound level Lq for the squeak signal q

Nq is the number of squeaking events. Already during a single

Braking can occur several squeaking events.

For each squeak occurrence q found, M will have different characteristics Mqi for one

Section xq (t) calculated from the signal x (t). Mqi indicates the value of

Feature i for the squeak event q.

Such a squeaking event is shown in FIG. It is in the upper half of

Fig. 2 plotted in the ordinate an airborne sound signal x (t) over the pointing in the abscissa t time. Between times tqstart and tqend, the presence of a squeak signal was detected. Therefore, between these two times the signal is denoted by xq (t). During this time segment, i. during the presence of the squeaking event, various characteristics Mqi are calculated for the squeak signal. Some of these features result from linking already calculated features, e.g. through their multiplication.

For example, the following characteristics MqO, ..., Mq6 are calculated from xq (t): 1) MqO: duration of squeaking event q. The duration is denoted by dq and results from dq = tqend - tqstart

2) MqI: A-weighted third octave Lq (A)

3) Mq2: A-weighted maximum level Lqpeak (A) from peak spectrum 4) Mq3: specific loudness Ns according to ISO 532 B or DIN 45631

5) Mq4: product of duration and A-weighted third octave, i. dq * Lq (A)

6) Mq5: product of duration and A-weighted peak level, i. dq * Lqpeak (A)

7) Mq6: product of duration and specific loudness, i. dq * Ns

The term A-weighting is understood to mean the multiplication of a spectrum by the A-weighting curve shown in FIG. For this purpose, a relative sound pressure level in dB over the frequency in Hz is plotted in FIG. The A rating curve is marked A. This curve takes into account the frequency dependence of the human volume perception. For example, a low frequency such as 50 Hz is only noticeable at a much higher sound pressure level than a tone at 1000 Hz. If a spectrum is weighted with an A curve, then low and high tones are attenuated and frequencies around 1000-6000 Hz hardly changed. The levels in the A-weighted spectrum at different frequencies are then directly comparable in terms of their volume perception by humans. As a concrete example, consider an unweighted spectrum with the following two tones included therein:

Sound at 50 Hz with a sound pressure level of 50 dB and sound at 1000 Hz with a sound pressure level of 20db. The weighting with an A-curve leads to an attenuation of 30 dB at 50 Hz. This results in an A-weighted level of 20 dB at 50 Hz and an attenuation of 0 dB at 1000 Hz. This results in an A-weighted one Level of 20 dB at 1000 Hz.

Thus, both tones are perceived as equally loud at a respective A-weighted level of 20 dB.

The loudness Ns is another quantity that reflects the loudness perception of humans. In this standard standardized in the ISO 532 B many effects such as the masking of individual tones by others, louder sounds and the loudness sensation depending on the level are considered. The spectrum of a time signal can be calculated by dividing the time signal into sections of equal length, for each of which a spectrum is calculated. The sections may overlap and, to improve the results, they may be weighted with a window function before the spectrum is calculated. The total spectrum of the signal is then calculated by averaging the individual spectra by averaging all values at the same frequency. In contrast to this, one obtains a peak value spectrum from the individual spectra referred to, by searching for the maximum value for each frequency in each spectrum and then applying this accordingly in the resulting peak spectrum.

For the practical application of brake squeal detection, a smoothed spectrum is formed by arithmetic averaging the sound pressure level of the unsmoothed spectrum in frequency intervals of one third. The level of this smoothed spectrum, which is also referred to as a third octave spectrum, is also called

Called third octave level.

It is possible to obtain the values for these features by means of an FFT analysis (FFT = "Fast Fourier Transform"). For the FFT analysis, the following settings have proved suitable: FFT length = 4096 sampling points, overlap of the

Time window = 50%, weighting with Hanning window.

To arrive at an index describing a squeak event in the further calculations, all features Mqi, i. the features of the type i or the i-th features for the squeaking event q, those squeaking events q from the signal x (t) combined in each case, which occur simultaneously in time or overlap in time. Even non-overlapping squeaking events that originate from the same braking process can be added as an option.

The combining takes place in each case by adding all the features Mqi of the type i to a feature sum which is normalized with the feature-specific factor Ci and thus the normalized feature sum FSi

FSi = Ci * Σ _q (Mqi), results. Ci typically takes values between 0.01 and 1. In this case, Σq denotes a summation over all squeaking events q. Thus, for each feature of the type i, that is to say for MqO, MqI,..., Mq6, there is in each case a feature sum FSi, that is to say FSO, FS1,..., FS6. It should be emphasized here that the sum FSi can also extend only over a squeak event, ie the feature sum comprises only one

Summands.

Subsequently, all normalized feature sums are weighted with a feature-sum-specific factor Ki and added to give Σi Ki * FSi.

In the exemplary embodiment with the features MqO, MqI,..., Mq6, i = 0, 1,

..., 6 sums up.

After normalization with Σj Ki, one obtains an objective variable OV, which represents the summarized squeaking events q:

OV = Σi (Ki * FSi) / (Σi Ki).

By substituting the objective quantity OV into the equation

bonisqueal = b, / U | C - OV ₁ | (1)

the objective valuation index bonisqueal is calculated. Here, bonisqueal is defined for values of 1 to 10 so that the value calculated by Equation (1) is set to 1 if Equation (1) gives a result less than 1 and set to 10 if (1) on Result greater than 10 delivers. a, b and c are selectable parameters. Ki typically takes values between 1 and 10.

For further simplification, it makes sense in terms of average human rating accuracy to round the calculated value bonisqueal to integer values.

For the method described, the following values for the parameters a, b and c have proved particularly suitable: a = 0.016, b = - 23.64375 c = 2.6327

The size bonisqueal is the rating for the annoyance of a single squeaking noise or a series of squeaking noises.

In the practical vehicle test, a great deal of braking and / or stopping operations are undertaken, which can be combined into a measurement sequence or a so-called "session." The frequency of squeaking events is then determined for a measurement sequence or session a measurement sequence evaluation index or

Session rating index sessionbonisqueal considered. For example, all braking operations during a test period or test day can be considered.

First of all, the arithmetic mean over all unrounded valuation indices bonisqueal with values smaller than 9.5 determined during the test period or test day is formed. Only the detected squeaking events are included in this mean value, but not squeaky braking processes.

Furthermore, the ratio of all decelerating related braking events is determined based on the total number of braking operations during the test period or test day. The value in% determined for this ratio is called NP.

Since the non-squeaky braking processes have so far not found an input into the determination of the arithmetic mean, a correction term designated as CORRECTION is determined in the following, which is the arithmetic

Mean value is added.

The determination of this correction term is shown in FIG. Here, in the abscissa direction, the size NP, i. the ratio of squeaky braking and the total number of braking actions in%.

The value 100 means that squeaking noises occurred during all braking operations. In the ordinate direction, the value of the correction factor CORRECTION is plotted. The correction factor CORRECTION assumes in the example for 0 <NP <10% values between 1 and 8, for NP <10% the correction value CORRECTION = 0. In this case, in FIG. 3, the 6 supporting points entered as black dots in the diagram were determined on the basis of experimental experiments:

1) For NP = 0.001, the correction value is CORRECTION = 8

2) For NP = 0.01 the correction value is CORRECTION = 6 3) For NP = 0.1 the correction value CORRECTION = 3

4) For NP = 1 the correction value is CORRECTION = 1.5

5) For NP = 10, the correction value CORRECTION = 1

6) For values NP> 10, the correction value CORRECTION = 0.

For intermediate values, e.g. how drawn a linear interpolation can be used. Of course, other courses for the correction factor and the definition of other interpolation points and / or interpolation point values are possible.

The meaning of this correction term becomes plausible if one considers that, according to FIG. 1, the annoyance of the noise for increasing values of bonisqueal decreases.

Very small values of NP in Fig. 3 mean that squeaking noises occurred at all with only a very small fraction of the braking operations. Therefore, as the value of NP decreases, an ever larger value CORRECTION is added to bonisqueal. This clearly shows that the annoyance of noise decreases, the rarer the noises occur during a braking process.

This correction term is added to the not yet integerized or truncated at 1 or 10 size bonisqueal and then the sum is rounded to integer values. Further, the sum is set to 1, if it assumes a value less than 1 is set to 10, if this assumes a value greater than 10. This integer value truncated at 1 and 10, which is referred to as the session bonus, also represents an objective index that assesses the annoyance of squeaking noises. This index is shown in FIG.

The sequence of the method according to the invention is shown in FIG. 5. After starting the method in block 500, at least one braking operation is investigated in block 501 and evaluated with regard to the squeaking behavior. Subsequently, in block 502, each detected squeaking sound is evaluated with regard to 6 features MqI,..., Mq6. For each Feature, there is a single rating number for each squeaking noise. Subsequently, in block 503, a feature evaluation number FSi is formed by linking the individual evaluation numbers formed with each squeaking sound for this characteristic. In block 504, a first intermediate variable OV is determined from the feature evaluation numbers FSi by a weighted addition. Thereafter, the size bonisqueal is determined in block 505 according to the specified relationship (1). This represents a measure of the annoyance of the squeaking noise that has occurred during the braking process. In block 506, a query is then made as to whether the evaluation should only extend over one braking operation. If the answer, yes "(in Fig. 5 with" y "marked), ie it is considered only one braking operation, then directly to

End of the procedure in block 508.

If the answer is "no", (indicated by "n" in FIG. a plurality of braking actions are considered, then in block 507 the mean of the individual determined values of bonisqueal is determined and, in addition, a term f (NP) is added, in which the number of performed squeaky braking operations relative to the total number of braking operations becomes this ratio referred to as NP, is received.

In block 509, the size sessionbonisqueal is determined therefrom, which is the total noise size for the squeaking noises occurring during the braking processes. In block 508, the process ends.

Claims

claims

1. A method for assessing the annoyance of squeaking noise within a sound signal generated during operation of a motor vehicle or in the operation of one of its components, in which the presence of at least one squeaking noise is detected (501), which evaluates at least one squeaking sound with respect to at least two predetermined features (Mqi) becomes (502) and from the at least two evaluations of this at least one squeaking noise a characterizing the annoyance of this at least one squeaking noise

Size is determined (bonisqueal, sessionbonisqueal, 505, 509).

2. Method according to claim 1, characterized in that the squeaking noises are brake squeaking noises.

3. The method according to claim 1, characterized

- That at least one squeaking noise is detected based on the Auffϊndens a maximum in an amplitude spectrum of the sound signal (501) and

- that the at least two features (Mqi) are taken from the following feature enumeration, which as features

the length of time the squeaking noise,

the maximum level value of the weighted amplitude spectrum,

the maximum level value of a weighted and smoothed spectrum obtained from the amplitude spectrum by smoothing and weighting, the specific loudness of the signal,

the product of the time length of the squeak signal and the maximum level value of the weighted and smoothed spectrum,

the product of the time length of the squeak signal and the maximum level value of the weighted spectrum, and the product of the time length and the specific loudness of the signal contains.

4. The method according to claim 3, characterized in that for each of the at least two features in the at least one squeaking noise, a single evaluation number (Mqi) is determined (502) and from the at least two determined individual evaluation numbers (Mqi) characterizing the annoyance of the at least one squeaking noise Size is determined (bonisqueal, sessionbonisqueal, 505, 509).

5. The method according to claim 4, characterized

that the at least one squeaking noise is at least one brake squeaking noise occurring during a braking operation,

- that for each of the at least two features (Mqi) by combining the determined for the respective feature for each brake squeal noise rating a number (FSi) is formed (503) and

- From the at least two determined feature evaluation numbers a the annoyance of the squeaking noise occurring during the braking process characterizing Einzelunästigkeitsgröße is determined (505, 509).

6. The method according to claim 5, characterized in that it is the addition is an addition, in particular a weighted addition.

7. The method according to claim 5, characterized

- That from the at least two feature evaluation numbers (FSi) by a weighted addition, a first intermediate size (OV) is determined (504) and

- From the first intermediate size (OV), the Einzelunästigkeitsgröße is determined (505, bonisqueal).

8. The method according to claim 7, characterized in that - the Einzelunästigkeitsgröße (bonisqueal) from the first intermediate size according to the

relationship

bonisqueal = where a, b, and c are selectable parameters, OV is the first intermediate size, and bonisqueal is the single annoyance size (505).

9. Method according to claim 8, characterized in that for the selectable parameters a, b and c the values a = 0.016, b = - 23.64375 and c = 2.6327 are selected.

10. The method according to claim 8, characterized in that

- the solubility size (bonisqueal) is rounded to the nearest integer value and - the solubility size (bonisqueal) is set to the value 1, if for the

Single lump size a value smaller than 1 is determined and

- that the solubility size (bonisqueal) is set to the value 10, if a value greater than 10 is determined for the individual annoyance size.

11. The method according to claim 8, characterized

- that at least two braking operations are performed,

- that the annoyance size of the occurring during the braking squeaking noises characterizing Gesamtlästigkeitsgröße (sessionbonisqueal, 509) is determined, - in the Gesamtlästigkeitsgröße (sessionbonisqueal, 509) the average value of each determined for each individual lessening behaviors (bonisqueal, 505) received, which is a predetermined condition fulfill.

12. The method according to claim 11, characterized in that it is the falling below a threshold value, in particular the

Threshold 9.5, which is the size of each individual (bonisqueal, 507).

13. The method according to claim 12, characterized in that

- in that the Gesamtlästigkeitsgröße (sessionbonisqueal, 509) additionally additively enters a term, in which in turn the number of performed squeaky

Braking occurs in relation to the total number of braking actions (NP).

14. The method according to claim 13, characterized in that the term is determined from the number of performed squeaky braking operations based on the total number of braking operations by means of a predetermined characteristic.

15. The method according to claim 14, characterized in that it is a monotone falling characteristic in the characteristic.

16. The method according to claim 11, characterized in that

- the total noise magnitude is rounded to the nearest integer value, and

- the total impact size is set to the value 1 if the value of the total turbidity is less than 1 and

17. Device containing means for carrying out the method of the preceding claims.