JP2000230884A - Sound evaluation device, judgment device for operating state of engine and waveform data processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To evaluate the influence of a specific sound out of sounds which are generated from various sound sources. SOLUTION: A sound storage means 2 by which sounds, for a set period, generated from one or a plurality of sound sources are converted into digital data and which stores sound data is provided. A specific-frequency-sound extraction means 4 by which sound data at a specific frequency is extracted from the sound data stored in the sound storage means 2 is provided. A sound- pressure-ratio calculation means 6 which calculates a ratio with reference to the sound pressure of sound data as a whole regarding the sound pressure data at the specific frequency extracted by the specific-frequency-sound extraction means 4 is provided. A data-for-evaluation output means 8 by which the ratio of the sound pressure calculated by the sound-pressure-ratio calculation means 6 is output as data for the evaluation of the sounds generated from the sound sources is provided.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sound evaluation device, and more particularly to a sound evaluation device for evaluating the influence of a specific sound on the whole sound in terms of time or frequency. As a field of application of the present invention, for example, it is preferably used for analysis of sound emitted from an engine, but is not limited to this, and is preferably used for analysis of sound and vibration generated by mechanical vibration.

[0002] The present invention also relates to the evaluation of mechanical noise (mechanical noise including engine noise, exhaust noise, etc.) of two-wheeled vehicles and four-wheeled vehicles. The present invention can also be applied to a device that outputs a signal at a fixed timing simultaneously with the sound of a rotating object or the like, and can be applied to evaluation of a motor-driven structure. Also,
Not only for sound but also for analyzing vibration of a mechanism that outputs a timing signal at a constant timing. As for the product field, the sound quality of all models such as two-wheel, four-wheel, and special equipment can be evaluated as needed. As for the production technology, sound quality evaluation such as completion inspection of a factory or a production line can be performed.

[0003] The present invention further relates to a technique for determining an engine state from an engine timing pulse. The determination of the engine state is useful for evaluating sound and vibration, but is not limited to this, and can be applied to a field in which the state of the engine is to be determined for various uses.

[0004]

2. Description of the Related Art Conventionally, evaluation of mechanical sound of a motorcycle or the like has been performed by hearing including development and shipping inspection. Therefore, there is no absolute scale for sound evaluation, making the evaluation unstable and uncertain. Some are personal computers (P
C) was introduced to digitize the sound, but it was difficult to obtain a final score of the sound from the numerical value.

For example, Japanese Patent Laid-Open Publication No. Hei 8-122140 discloses a neural network in which sensory evaluation values for a plurality of sound pressure values from an FFT analyzer are learned in order to absolutely evaluate gear noise of an automobile transmission or the like. A gear noise evaluation device is disclosed.

[0006]

However, in the above-mentioned conventional example, there is a disadvantage that intermittently and intermittently generated sounds cannot be evaluated. Further, with respect to the continuous sound, there is a disadvantage that it is not possible to evaluate, for example, the effect of the gear noise on the overall sound. In addition, there is an inconvenience that it is not possible to generate a target sound for recognizing the sound in the current device and to change the sound to the desired sound.

[0007]

SUMMARY OF THE INVENTION The object of the present invention is to provide a sound evaluation apparatus which can solve the disadvantages of the prior art and, in particular, can evaluate the influence of a specific sound among sounds emitted from various sound sources. With that purpose.

[0008]

Accordingly, in the present invention, there is provided a method for controlling
Or, a sound storage means for converting sound of a certain period emitted from a plurality of sound sources into digital data and storing the sound data, and a specific frequency for extracting sound data of a specific frequency from the sound data stored in the sound storage means The sound extracting means, the sound pressure ratio calculating means for calculating the ratio of the sound pressure of the specific frequency sound extracted by the specific frequency sound extracting means to the sound pressure of the entire sound data, and the sound pressure ratio calculating means Evaluation data output means for outputting the sound pressure ratio as evaluation data of sound emitted from the sound source. This aims to achieve the above-mentioned object.

In the present invention, the original sound is digitized and stored in the sound storage means. And the specific frequency extracting means is:
A sound of a specific frequency is extracted from the sound data. For example,
A specific order sound is extracted using an order filter. Further, the sound pressure ratio calculating means calculates the ratio of the sound pressure between the specific sound and the whole sound. Then, the evaluation data output means outputs the ratio itself or an evaluation value correlated with the ratio as evaluation data. When the evaluation data is output according to this ratio, the influence of the sound extracted by the specific frequency extracting means on the whole sound is displayed.

In the present invention, a sound storage means for converting a sound emitted from one or a plurality of sound sources for a predetermined period into digital data and storing the sound data, and specifying the sound data stored in the sound storage means , A change rate calculating means for calculating a change rate of the sound data sound pressure in which the sound of the specific section is masked by the timing filter means, and a change rate calculating means for calculating the change rate of the sound data. Evaluation data output means for outputting the rate of change as evaluation data of the sound emitted from the sound source.

In this example, the timing filter means
The sound in the specific section is masked out of the sound data stored in the sound storage means. The specific section is, for example, a predetermined section or a section determined according to a timing pulse externally input. For example, a sound having a large sound pressure that periodically appears before and after the problem sound is masked. Then, the change rate calculating means calculates a change rate of the sound pressure of the sound data in which the sound in the specific section is masked. Then, the evaluation data output means outputs the evaluation data according to the change rate. Then, the change rate of the problem sound is obtained while masking sounds before and after the problem sound.

[0012]

Embodiments of the present invention will be described below with reference to the drawings. <Overall Configuration> Next, an embodiment of the present invention will be described with reference to the drawings. As shown in FIG. 1, a sound evaluation device according to the present embodiment converts sound for a certain period of time emitted from one or a plurality of sound sources into digital data, and stores the sound data. A specific frequency sound extracting means for extracting sound data of a specific frequency from the sound data stored in the means; and a sound pressure of the specific frequency sound extracted by the specific frequency sound extracting means with respect to a sound pressure of the entire sound data. A sound pressure ratio calculating means for calculating a ratio; and an evaluation data output means for outputting a sound pressure ratio calculated by the sound pressure ratio calculating means as evaluation data of a sound emitted from a sound source. I have.

In the example shown in FIG. 1, the specific frequency extracting means 4 extracts a sound of a specific frequency from the sound data stored in the sound storage means 2. For example, specific frequency sound extraction means 4
When a timing pulse is added to sound data, an order filter frequency calculating unit 10 that calculates a frequency of a sound to be extracted for each order with respect to the cycle of the timing pulse
And functions as an order filter. As described above, the specific frequency sound extracting means calculates the frequency of the sound in question, and extracts only the specific frequency component with a predetermined bandwidth. Then, the sound pressure ratio calculating means 6 calculates the ratio between the sound pressure of the sound data of the original sound stored in the sound storage means 2 and the sound pressure of the sound extracted by the specific frequency extracting means. When this ratio is high, the sound of the specific frequency is heard loudly, while when this ratio is low, the sound of the specific frequency in question is masked by other whole sounds and becomes hard to hear. Therefore, when the ratio of the sound pressure is output as the evaluation value, it is possible to know the timing at which the sound of the specific frequency becomes a problem. Also, even if a specific frequency sound is extracted, even if only the sound of this frequency is reproduced, the sound will be far from the original sound. , The problem sound can be confirmed aurally. With the configuration shown in FIG. 1, a continuous problem sound is evaluated.

In the embodiment shown in FIG. 2, sound for a certain period of time emitted from one or a plurality of sound sources is converted into digital data, and sound storage means 2 for storing the sound data is stored in the sound storage means 2. Filter means 14 for masking a sound in a specific section of the sound data, and a change rate calculating means 1 for calculating a change rate of the sound pressure of the sound data in which the sound in the specific section is masked by the timing filter means 14.
2 and an evaluation data output means 8 for outputting the change rate calculated by the change rate calculation means 12 as evaluation data of a sound emitted from the sound source.

In the example shown in FIG. 2, an intermittent problem sound is evaluated. Here, the timing filter means 1
4 masks a sound in a specific section. For example, a sound other than the problem sound is masked. Then, the change of the problem sound appears favorably. In the example shown in FIG.
Has a function 16 for defining a specific section in response to a timing pulse input according to the operation of the sound source. In this example, it is possible to evaluate a problem sound that periodically appears from a sound from a periodically rotating sound source such as an engine or a motor. That is, when there is a problem sound and a sound other than the problem sound output at a specific timing, by masking the sound other than the problem sound, the change in the problem sound itself can be evaluated.

In the example shown in FIG. 2, after masking sounds other than the problem sound, the change rate calculating means 12 calculates the change rate of the sound pressure of the sound after the masking. Then, it is possible to specify at what timing the problem sound itself changes greatly. Then, the frequency of the sound at that time can be obtained, and the component serving as the sound source can be specified based on the frequency and the timing. In addition, since a sound that greatly changes generally tends to be aural noise, the problem sound can be evaluated by examining the change rate of the sound pressure. In this case, an evaluation value is determined in advance for each predetermined range of the change rate.

In the example shown in FIG. 3, by defining a section for masking sound data and vibration data based on a timing pulse, data suitable for evaluating the influence of the target vibration or sound is generated. That is, the waveform data processing device shown in FIG. 3 converts the waveform of the vibration of the object to be measured or the sound waveform due to the vibration into digital data and stores the waveform data, and the waveform data storage device 18. And a timing pulse storage means 20 for storing a timing pulse indicating the operation timing of the measuring object synchronized with the waveform of a certain period.

Further, based on the timing pulse stored in the timing pulse storage means 20, a mask section specifying means 22 for specifying a mask section for masking the level of the waveform for a certain period, and the mask section specifying means 22 specifies the mask section. A mask processing unit for masking the waveform data of the section stored in the waveform data storage unit to a predetermined level; Further, the example shown in FIG. 3 includes a waveform data output unit 30 that outputs the waveform data masked by the mask processing unit 24.

The waveform data output means 30 includes, for example,
The masked sound data is converted to analog data and output from a speaker, or the waveform itself is displayed on a display. Further, in a configuration having vibration analysis software, voice analysis software, or the like, the waveform data output means 30 inputs the waveform data after the mask processing to the analysis software. The analysis software can analyze the problematic part using the waveform data in which the vibration or sound in question is emphasized. In the example in which the vibration data is masked, the vibration caused by a specific part among various parts of the vibration source can be extracted and analyzed by the mask processing.

In the example shown in FIG. 3, the mask processing means 24 includes a target level calculating section 26 for calculating a target level such as a level of a low frequency component of the waveform data, and a waveform data in the mask section on a time axis. It is preferable to include a subdivision masking section 27 that subdivides and changes the level of each subdivided waveform data to the target level calculated by the target level calculation section 26. Then, mask processing other than the problem sound can be performed with higher accuracy.

FIG. 4 is a block diagram showing a configuration of an engine operating state determining device for determining an operating state of an engine from a timing pulse input from the engine, for example, in order to mask an exhaust sound among engine sounds. In the example shown in FIG. 4, a timing pulse storage unit 20 that stores a timing pulse output according to a piston position of the engine during operation of the engine, and an engine based on the timing pulse stored in the timing pulse storage unit 20. A rotation speed calculation unit 28 for calculating the rotation speed of the motor, and an explosion top dead center for determining whether each timing is an explosion top dead center based on a temporal change in the rotation speed calculated by the rotation speed calculation unit 28 A determination unit 29;

In the example shown in FIG. 4, a timing pulse is output at the top dead center and the bottom dead center of the two-cycle engine, and the explosion top dead center and the exhaust top dead center of the four-cycle engine are output. When the timing pulse is output in (1), it is determined which timing pulse is the top dead center of the explosion.
The rotation acceleration is obtained by counting the time interval of the timing pulse, and it is determined whether or not each timing is the top dead center of the explosion based on the change in the acceleration. This gives 2
The timing of the top dead center of a four-stroke engine in which one of the top dead centers is an explosion top dead center, or of a two-stroke engine in which a misfire has occurred and not exploded at all the top dead centers, Can be obtained exactly. Specifically, in the case of a four-cycle engine, the timing at which the rotational acceleration increases can be determined as the top dead center of the explosion. In the case of a two-cycle engine, when the acceleration continuously increases with two timing pulses, it can be determined that the top dead center has occurred. If the acceleration continuously decreases in two timing pulses, it can be determined that a misfire has occurred, and if the acceleration has decreased in the first timing pulse and increased in the next timing pulse, it can be determined that incomplete combustion has occurred. . The state of the engine can be determined from the timing pulse based on such an acceleration / deceleration pattern. In particular, even in the case of a multi-cylinder engine, it is possible to satisfactorily determine the state of the engine by obtaining a pattern of acceleration / deceleration of rotation. In the present invention, this is used, for example, to specify the timing at which exhaust noise is emitted, but is not limited to this, and may be used for engine control.

FIG. 5 is a block diagram showing a configuration for evaluating sounds using the engine and its surroundings as sound sources.
In the example shown in FIG. 5, a timing pulse storage unit 20 that stores a timing pulse output according to the piston position of the engine during operation of the engine, and converts a sound of a certain period emitted from the engine into digital data and converts the sound into digital data. A sound storage means 2 for storing sound data and a sound evaluation means 8 for converting the sound data stored in the sound storage means 2 into evaluation data.

The sound evaluation means 8 includes an explosion top dead center specifying section 29 for specifying the timing of the explosion top dead center of the engine based on the timing pulse stored in the timing pulse storage means, A mask section setting unit 22 that sets a section predetermined in relation to the reference time with the explosion top dead center specified by the specifying unit 29 as a reference time, and a mask set by the mask section setting unit 22 A mask processing unit for masking the sound data stored in the sound storage means for the section; and an evaluation data output unit for outputting the sound data masked by the mask processing unit as sound evaluation data. I have.

In the configuration shown in FIG. 5, the engine operation state determination device and the sound evaluation device shown in FIG. 4 are combined, a predetermined sound is masked according to the engine operation state, and only the problem sound is extracted and evaluated. Create data for use. Here, the change rate of the sound pressure is not necessarily calculated, for example, when the sound in a masked state is converted to analog data and reproduced as it is,
Also, a case where the sound pressure of the problem sound is increased or decreased to reproduce the sound in order to search for a target sound is included. Of course, when the problem sound is evaluated, the change rate of the sound pressure may be calculated.

In the example shown in FIG. 5, the explosion top dead center specifying unit 29 corresponds to the configuration of the operation state determining device shown in FIG.
Is a rotational speed calculating function 34 for calculating the rotational speed of the engine based on the timing pulse stored in the timing pulse storage means 20, and each timing based on the time change of the rotational speed calculated by the rotational speed calculating function 34. Top dead center determination function 3 for determining whether the object is at the top dead center
6 is provided.

For this reason, for example, the timing at which the exhaust sound, which is the sound immediately after the top dead center of the explosion is output, and the operation state of the parts are specified in accordance with the operation of the engine. And the like. Further, the mask processing unit 24 illustrated in FIG. 5 reduces the sound pressure level of the sound data to a predetermined sound pressure level while maintaining the frequency ratio at each time interval obtained by subdividing the mask section along the time axis. A subdivision mask function 38 is provided. This makes it possible to calculate the mask ratio up to this reference, for example, based on the sound pressures before and after the problem sound section in each of the subdivided ranges, and perform the masking process. However, even if the sound to be masked has a fluctuation in sound pressure, it is possible to generate an auditory natural sound.

The evaluation data output section 8 shown in FIG. 5 calculates a fluctuation rate of the sound pressure of the sound data masked by the mask processing section 24 and outputs the fluctuation rate as sound evaluation data. 40 is provided. Thereby, a sound unnecessary for the evaluation of the problem sound is masked, and subsequently, a fluctuation rate of the sound pressure of the whole sound is calculated. Since a high change rate means that a loud sound is output suddenly, the change rate of the sound pressure is closely related to the auditory evaluation. Therefore, it is possible to automate the evaluation of the sound based on the fluctuation rate of the sound pressure.
In other words, an evaluation point may be determined for each predetermined sound pressure range and output, or a timing at which the fluctuation rate of the sound pressure becomes equal to or higher than a predetermined threshold may be output. You may. When the timing at which this fluctuation rate is large is known, it becomes possible to specify the component that has become the sound source and the operation state of the component, and then it is easy to specify the target of a design change or the like to reduce the sound. it can.

If the sound source is a device that performs a periodic operation, for example, if it is an engine or the like, the evaluation data output unit 8 outputs the mask processing unit based on the timing pulse stored in the timing pulse storage unit 20. It is preferable to provide an averaging function 42 for averaging the sound data masked by the averaging in units of engine cycles. Thereby, smoothing is performed and noise is reduced. Then, a more accurate fluctuation rate can be obtained. In this processing, for example, in the case of an engine, the data of the same rotation speed may be added and averaged, or the intervals between the data of different rotation speeds may be standardized and then added and averaged.

FIG. 6 is a block diagram showing a configuration of a sound evaluation device for evaluating a continuous sound. In the example shown in FIG. 6, a timing pulse storage unit 20 that stores a timing pulse output according to the piston position of the engine during operation of the engine, and converts a sound for a certain period emitted from the engine into digital data and converts the sound into digital data. Sound storage means 2 for storing sound data, order calculation means 44 for calculating a frequency for each order based on the number of gear teeth of the engine, the number of poles of the magnet rotor, and the timing pulse, and the order calculation means 44 Sound evaluation means 6 for evaluating sound for each frequency component of the calculated order.

Moreover, the sound evaluation means 6 includes the order calculation means 4
And a sound extraction unit 46 for extracting the sound having the frequency calculated by the step S4 from the sound data, and the sound pressure of the sound extracted by the sound extraction unit 46 and the sound of the whole sound stored in the sound storage unit 2. A sound pressure ratio calculating section 48 for calculating a pressure ratio and an evaluation data output section 8 for outputting the sound pressure ratio calculated by the sound pressure ratio calculating section 48 as sound evaluation data are provided.

In the example shown in FIG. 6, the degree calculating means 44
The order for the timing pulse is calculated from the timing pulse. When the order is specified, the order-specific sound extraction unit 46 extracts a sound of this order frequency from the sound data stored in the sound storage unit 2. Further, the sound pressure ratio calculation unit 48
Calculates the ratio between the sound pressure of this specific order sound and the sound pressure of the whole sound, and the evaluation data output unit 8 outputs the ratio itself or an evaluation point corresponding to the ratio. In the example shown in FIG. 6, since the order is calculated from the timing pulse and the sound is extracted for each order, it is suitable for evaluation of, for example, a continuous sound emitted from an engine.

[0033]

Next, an embodiment of a sound evaluation apparatus according to the present invention will be described with reference to the drawings. FIG. 7 is a block diagram showing the configuration of the present embodiment. As shown in FIG. 7, the sound evaluation device includes a tachometer 32 for capturing a timing pulse and a sound level meter (microphone) 50 for capturing a sound with respect to the car 1 to be evaluated.
And an interface box 51 for connecting them to an A / D converter 53 added to the PC 52, and a PC 52. In the example shown in FIG. 7, a PC is used, but this capturing unit may be dedicated hardware.

The personal computer 52 includes, as the data processing unit 54, analysis software, timing filter software, and evaluation point calculation software for calculating evaluation points from numerical values obtained by the software. The analysis software has, for example, functions such as FFT used in sound analysis and a frequency filter. The present embodiment relates to timing filter software and evaluation point calculation software that are part of the data processing unit 54.

The evaluation object in this embodiment includes a continuous sound and an intermittent sound. The continuous sound is a sound generated by a large number of meshes or the like for one rotation of a crank, such as a gear or a chain, and is also called a beat sound. If there is a timing pulse, these sounds can be relatively easily extracted by an order filter (which determines the frequency from the number of rotations and the number of gear teeth and applies a filter at the moment). In order to obtain an actual evaluation point of the extracted sound, it is necessary to separately convert it into a numerical value. This is a part of the part proposed in the flowchart for calculating the correlation between the whole sound and the problem sound and calculating the evaluation score.

The intermittent sound is a sound that is generated about once or twice in one cycle and almost at a specific timing. The intermittent noise includes, for example, gear rattle caused by fluctuation of crank rotation immediately after combustion, and clutch noise. Some of these sounds are emitted at a limited frequency, while others are emitted over a wide frequency band. Conventionally, for such a sound, a frequency filter was used to extract only the problem sound and compare the loudness of the sound.However, in this method, the sound after extraction is a sound that differs from the original sound. In addition, sounds generated at other timings are included. If the frequency of the problem sound is wide, the extraction itself is difficult. A configuration for solving this is a timing filter, and calculates an evaluation score based on a numerical value obtained thereby.

FIG. 8 is a flowchart showing an outline of a processing example of the data processing section 54. When the sound data is captured (step S1), the type of the problem sound is confirmed (step S1).
2). This may be input to the PC from input means (not shown) in advance, or may be stored at the start of the sound data as instruction data for selecting one of the sound data. In the case of continuous sound evaluation (step S
3) Extract sound using an order filter (step S)
4). Then, the correlation between the whole sound and the extracted problem sound is calculated (step S5). For example, the sound pressure ratio is calculated. Then, an evaluation score is calculated based on the correlation.
For this reason, the auditory evaluation score of the sound can be obtained for each order.

On the other hand, if the sound is an intermittent sound (step S
7) First, the problem sound itself or a sound other than the problem sound is masked or amplified by the timing filter (step S8). After thus shaping the sound data for each timing, the sound pressure fluctuation of the whole sound (problem sound) is calculated (step S9). In the evaluation of intermittent sound, based on this sound pressure fluctuation,
Calculate the evaluation score.

<Continuous Sound> A continuous sound can be relatively easily extracted by using an order filter. However, it is difficult to calculate an evaluation point using only the extracted sound. This is because no matter how loud the extracted sound is, if other background noise is louder,
It does not matter and the evaluation score is better. Conversely, even if the extracted sound is small, if the other sounds are small, a problem arises, and the evaluation score deteriorates. Conventionally, it has not been possible to evaluate a sound having a frequency extracted in relation to the background noise.

For this reason, in this embodiment, the relative problem sound including the background sound is quantified by calculating the ratio of the problem sound to the entire sound. FIGS. 9A to 9C show examples of this technique. In FIG.
When the accelerator is turned on and off in the engine neutral state, the first half turns on the accelerator, the rotation increases, and the overall sound also increases. Then, turn off the accelerator,
This shows a case where the whole sound decreases as the rotation decreases.

FIG. 10 is a flowchart of the continuous sound evaluation. In the evaluation of the continuous sound (step S11), first, FIG.
As shown in (A), a sound of the target order 60 is cut out with a specific bandwidth 58 (step S12). Next, as shown in FIG. 9B, the sound pressure 64 of the cut out sound is obtained (step S13). Further, the sound pressure 63 of the entire sound is obtained (step S14). Further, as shown in FIG. 9C, the ratio of the sound pressure 67 of the problem sound to the entire sound pressure 66 is calculated (step S15). Then, the order sound ratio increases.
The portion indicated by reference numeral 62 in (C) is a numerical value indicating an unfavorable state of the sound, while the other portions are masked by hearing other than the problem sound, so that the evaluation score is relatively good. (Step S16).

As shown in FIG. 9A, the order of the continuous sound has a higher frequency as the rotation increases, and as the rotation decreases with the accelerator off, the frequency also decreases. When a problem sound of such an order is extracted with a certain bandwidth and its sound pressure is obtained, the sound pressure becomes as shown in FIG. 9B. At this time, the overall sound pressure is also obtained as shown in the figure. Taking the ratio between the sound pressure of the extracted problem sound and the sound pressure of the whole sound, FIG.
As shown in (C), when the problem sound can be heard loudly, the ratio of the problem sound to the whole sound increases. By comparing the magnitude of this ratio, the problem sound including the background sound can be evaluated.

FIGS. 11A and 11B show the 8th order and 17th order.
The ratios of the next problem sound evaluated as being low and the sound evaluated as large are obtained by the above-described processing. The portion where the problem sound can be best heard is surrounded by a circle 62. This is the part where the accelerator is turned off and the overall sound decreases as the rotation decreases, but the problem sound remains and can be heard loud just before idling. As shown in FIG. 11, in the case of the evaluation with a small problem sound, the eighth sound is about 10%, the 17th order is about 20%, and the other sounds are about 70% of the whole sound. On the other hand, those with a high problem sound
The peak of the eighth order is as large as 60%, and the peak of the 17th order is as large as about 35%. By digitizing the ratio of the problem sound to the whole sound obtained in this way using evaluation point calculation software, it is possible to obtain a numerical value equivalent to a human auditory evaluation point. As described above, in the present embodiment, when a sound of a specific frequency is extracted, the sound becomes far apart from the original sound, and it is difficult to evaluate even if the actual sound is reproduced.
Since the influence of the problem sound is evaluated by focusing on the sound pressure ratio, useful evaluation data with high accuracy can be obtained with a relatively simple sound evaluation.

<Intermittent Sound> Next, the evaluation of the intermittent sound will be described using an engine as an example. Here, a method for evaluating the intermittent sound itself and a method for determining the operating state of the engine in order to evaluate the intermittent sound using this method will be described.

It is difficult to extract intermittent sounds with the above-described order filter or frequency filter. Therefore, the idea of the timing filter is to extract the problem sound at the timing when the sound is generated. Even with commercially available software,
There is a filter that filters sound data at a specific time to increase or decrease it, but there is no filter that uses a timing pulse and applies a cycle filter at each specific timing.

FIG. 12 is a flowchart showing an example of a process for evaluating an intermittent sound. It is the angle or time from the top dead center of the explosion that determines the timing of the intermittent sound. Normal data is arbitrary data during engine operation, and it is not known which pulse among many timing pulses is the top dead center of the explosion. This means that in the case of four cycles, the timing pulse is output at the top dead center every time, so that one cycle is indicated by two pulses, and there is an explosion top dead center and an exhaust top dead center. For this reason, it becomes impossible to determine which pulse among many timing pulses is the top dead center of the explosion.

In the case of two cycles, if the pulse is once per rotation, the position of the pulse always becomes the top dead center. In rare cases, a pulse is output even at the bottom dead center. You need to determine which pulse is the top dead center. For this, first, the time between pulses is obtained, converted into the number of rotations, and the fluctuation of the number of rotations is obtained. The pulse in which the rotation fluctuation becomes positive is the top dead center of the explosion of four cycles. In the case of 4 cycles, the explosion is relatively stable every time,
As shown in FIG. 13 (A), acceleration and deceleration are alternately repeated regularly every time. In other words, since the fluctuation of the rotational speed 70 is stable, the rotational fluctuation 72 is also stably repeated, and the acceleration and deceleration are alternately repeated regularly every time. For this reason, it can be determined that the timing pulse at which the rotation fluctuation 72 is positive is the top dead center of the explosion. Therefore, it is found that either the even-numbered pulse or the odd-numbered pulse is the top dead center of the explosion, and the other is the top dead center of the exhaust gas.

In the case of two cycles, unlike in the case of four cycles, an explosion often does not occur regularly every time. An example is shown in FIG. In this example, the timing at which the engine speed increases greatly in the idling state (the timing at which the acceleration is large) indicates a cycle in which the engine explodes, and the other engine speeds decrease indicating a cycle in which a misfire occurs. ing. As described above, in the case of two cycles, since the number of rotations increases due to the explosion, it is possible to determine the top dead center of the explosion by checking at what timing pulse the rotation increases.

In this way, it is possible to determine whether the top dead center of the explosion in the case of two pulses in one cycle is an even number or an odd number. Once the top dead center of the explosion is known, it is next determined when to apply the filter. This varies depending on the sound in question, and the way of applying the filter is also changed depending on the target.

As an example of use of the timing filter, a filter adapted to two cycles of sound emitted at a certain timing after the top dead center of the explosion will be described. FIG. 14A shows an original sound waveform and timing pulses, and this engine has two pulses per cycle. The problem sound 74 is a sound periodically emitted immediately after the top dead center (TDC) of the explosion as shown in the figure. Fig. 1 shows the sound pressure fluctuation at this time.
5 is shown in STD.

As shown in FIG. 14, the sound pressure fluctuation is caused by the exhaust noise 7
6 In a certain cycle, the problem sound is buried, and the difference in sound pressure between the cycle in which combustion occurred and the cycle in which misfire occurred is large. Have difficulty. Therefore, the exhaust sound is masked by a timing filter to extract a problem sound. For this purpose, it is necessary to estimate the combustion state from the timing pulse. The flowchart is shown in FIG. FIG. 16 shows the case of two pulses per cycle, but the combustion state can be estimated based on the same concept with one pulse per cycle. The specific processing shown in FIG. 16 will be described later.

The rotation for one cycle (interval of two pulses) obtained from the timing pulse is accelerated in the case of normal combustion, and decelerated in the case of misfire. There is a state of illegal combustion in the middle, and although small, you can hear the exhaust noise.

FIG. 14B shows an exhaust sound portion obtained by applying a timing filter to the exhaust sound portion of the obtained normal combustion cycle and masking the exhaust sound. As shown in FIG. 14B, only the exhaust sound portion 76 is masked, and the other portions remain the original sound. The sound pressure fluctuation at this time is represented by reference numeral 80 in FIG.
Indicated by As shown in FIG. 15, the sound pressure of the exhaust sound portion 76 decreases, and the sound pressure fluctuation of the problem sound that is periodically emitted can be seen. When the sound pressure fluctuation rate at this time is obtained, the fluctuation rate increases as the problem sound increases, and the fluctuation rate decreases as the problem sound decreases. By digitizing this fluctuation rate using evaluation point calculation software, a numerical value equivalent to a human auditory evaluation point can be obtained. Further, since this sound pressure fluctuation periodically appears for each rotation of the engine, the waveform of the sound pressure fluctuation shown in FIG. 15 may be averaged. Then, high frequency component noise can be removed.

Also, by masking at the timing of the problem sound and changing the mask ratio, a sample sound of only the problem sound having a different size can be easily obtained, and the score check and the sound check of the auditory evaluation can be easily performed. Can be. That is, when the engine and its surroundings are changed in design to remove the problem sound, a sound that aims to be such a sound can be generated as a sample sound.

Unlike an ordinary frequency filter, the specific sound is increased or decreased by increasing or decreasing only the amplitude without changing the period of the raw sound data at a specific time interval. In the frequency filter, sounds other than the target timing are also filtered, so that the sound is different from the original sound in terms of hearing. On the other hand, in the timing filter, the frequency does not change, and only the magnitude changes. Therefore, it is heard that only the target sound increases or decreases from the original sound.

The operation of evaluating the intermittent sound will be described with reference to FIG. First, the timing pulse time is checked, and the rotation speed is calculated from the timing pulse (step S2).
2). Subsequently, the position of the explosion pulse is detected by the processing shown in FIG. 16 (step S23). This is, for example, a 4-cycle explosion top dead center, a 1-cycle 2 pulse, or a 2-cycle explosion top dead center. Subsequently, one cycle interval is detected, and detailed setting of the filter mask is performed. This is, for example, whether the mask is a regular or irregular mask, a mask time, a mask shape, a delay time, and the like. Subsequently, a filtering process is performed (step S24). In the example shown in FIG.
The timing of 6 is specified from the top dead center of the explosion, and the exhaust sound portion 76 is masked. Then, as shown in FIG. 14B, the exhaust sound is masked, and a waveform in which the problem sound is extracted is obtained.

Further, the sound pressure fluctuation after the filtering is calculated (step S25). Then, for example, a waveform of the sound pressure fluctuation after the exhaust mask shown in FIG. 15 is obtained (step S25).
Further, after performing processing such as averaging, an evaluation point is calculated from the sound pressure fluctuation (step S26).

The characteristics of the mask of the timing filter include:
In addition to the mask time and the mask timing, there are parameters such as a mask shape and a mask ratio, and a mask pattern as shown in FIG. 17 is generally considered. Of course, T1,
A curve between T3 may be set. The method of selecting the mask ratio is also described in (1). A method of selecting a mask ratio so as to be equivalent to a cycle without sound,
(2). (3) a method of selecting a mask ratio by a coefficient according to the magnitude of the rotational acceleration due to combustion; (4) a method for fixing the mask ratio; There is a method of designating an absolute sound pressure value and selecting a mask ratio so that the sound pressure of a target sound becomes that value.

The method of selecting the mask ratio differs depending on how the target sound is handled. If the exhaust sound is to be masked as described above, (1). Is effective. When the generation timing of the target sound is constant, but the loudness of the sound depends on rotation fluctuation, (2). Is effective.

As a method of masking a sound generated at a specific timing as described above, a timing filter is very effective, and its adaptation range is wide. In addition, the mask rate can be selected in cycle units and can be masked at a specific timing in each cycle, so that the problem sound can be masked, and conversely, it is possible to mask other than the problem sound,
It is a very effective tool for sound evaluation.

The above method is advantageous in selecting a mask ratio and processing time. However, if there is a small sound within the mask width, the sound disappears and causes skipping. On the other hand, if the inside of the mask width is divided into minute sections and the mask processing is performed only for a sound pressure equal to or higher than a certain level, the mask processing is performed only for the loud sounds without changing the small sounds.
No skipping occurs.

<Judgment of Engine Operating State> Next, FIG.
A method for determining the state of the two-stroke engine will be described with reference to FIG. First, the rotation fluctuation shown in FIG. 13 is obtained from the timing pulse interval (step S31). Subsequently, it is determined whether the rotation fluctuation for one cycle is positive (Step S).
32). If the rotation fluctuation is positive, it is determined that the combustion is normal (step S33). On the other hand, if the rotation fluctuation is not positive, it is determined whether the rotation fluctuation for a half cycle is positive (step S34). If the rotation fluctuation for a half cycle is positive, it can be determined that the combustion is illegal (step S3).
5). And if the rotation fluctuation for one cycle is negative,
It is determined that a misfire has occurred (step S37). As described above, since the state of the engine can be determined from the timing pulse, it can be applied to various controls. In particular, since it is possible to determine the timing of the misfire and the timing of the explosion, evaluation of sound and vibration and feedback to the design are facilitated.

<Small Section Mask> A method of dividing the mask width into minute sections and masking will be described. When sound evaluation is performed based on fluctuations in sound pressure, it is necessary to determine how much the problem sound affects the overall sound. Here, when the filter scan is performed at a constant mask rate as described above,
As shown in FIG. 17B, the sound pressure is further reduced by the mask effect in the portion 83 where the sound is reduced, and the sound pressure fluctuation value there is increased. Even if the sound pressure fluctuation value of the problem sound part decreases, correct evaluation cannot be performed if the sound pressure fluctuation value of the other parts increases. Therefore, by dividing the inside of the mask width in a shorter time, calculating the target sound pressure assuming that there is no problem sound, and performing masking so as to achieve the sound pressure, it is possible to prevent such adverse effects. Can be.

In order to perform such masking, as shown in FIG. 18, first, the sound pressure in the Δt section before and after the mask width is calculated, and the target sound pressure for masking the problem sound portion is obtained. Subsequently, the inside of the mask width is divided by Δt, and individual sound pressures are calculated. Further, the mask ratio is calculated individually so as to achieve the target sound pressure. Then, except for the final mask portion, the mask ratio is changed in the initial portion of the mask as shown in the right diagram to avoid discontinuity with the previous mask portion. In the final mask portion, both initial and end mask ratios are changed to avoid discontinuity before and after.

In the example shown in FIG. 18, the target sound pressure is selected based on the sound pressure before and after the mask width. However, if the same sound pressure is to be used in cycle units, the target sound pressures may all be the same. Has a great influence on the evaluation.
However, if this process is performed, the process takes longer time than the process in which the mask ratio is constant. If the width of the problem sound is substantially uniform and there is no small portion of the sound within the mask width as described above, processing with a constant mask rate can sufficiently cope with the problem.

In the present embodiment, the sound emitted from the two-wheeled vehicle has been evaluated.
As long as there is an angular correlation, it can be applied to all. By changing the final evaluation portion, various sounds can be evaluated. The same effect can be obtained not only for sound but also for vibration waveforms.

As described above, according to the present embodiment, since the timing pulse is used as data together with the sound, the time, rotation fluctuation, and combustion state of each cycle can be determined by the timing pulse as well as the timing filter. Furthermore, since the ratio between the sound pressure of the continuous sound (problem sound) and the sound pressure of the whole sound is obtained, the loudness of the problem sound can be quantified by a relative value considering the loudness of the whole sound. Can be obtained.

In the example of masking the sound other than the problem sound and obtaining the sound pressure fluctuation of the intermittent sound (problem sound), the magnitude of the problem sound is removed from the influence of the other sounds, and the sound pressure fluctuation is quantified as the magnitude of the sound pressure fluctuation. can do. In the example of masking a sound with a timing filter, a sound sample can be easily sampled by changing a mask ratio of the problem sound, and a problem sound can be extracted by masking other than the problem sound. it can.

Further, in the technique of arbitrarily selecting the mask method of the timing filter, the method of
Parameters such as a mask ratio, a mask shape, and a timing can be easily set.

Further, by determining the interval between the timing pulses, it is possible to determine which pulse is the top dead center of the explosion. In order to determine the rotation fluctuation in cycle units, normal combustion, illegal combustion, and misfire are detected. It can be determined in cycle units.

[0071]

According to the present invention, the specific frequency extracting means extracts a sound of a specific frequency from the sound data, and the sound pressure ratio calculating means extracts the specific sound from the specific sound. In order to calculate the ratio of the sound pressure of the whole sound, when the evaluation data output unit outputs the ratio itself or an evaluation value correlated with this ratio as evaluation data, the whole sound of the sound extracted by the specific frequency extraction unit is output. In particular, when the frequency component is extracted, the tone will be different from the original sound.However, by taking the ratio of the sound pressure, it is possible to know the timing when the effect of the sound of the specific frequency is large. Thus, by listening to the original sound at the timing, the original sound affected by the specific frequency can be evaluated. Furthermore, when the sound at the frequency is noise, the ratio of the sound pressure is large. That is, the evaluation is low, and therefore, it is possible to obtain an evaluation of the sound of the frequency by, for example, associating the range of the ratio value with the evaluation value. can do.

Further, in the configuration in which the change rate calculating means calculates the change rate of the sound pressure of the sound data in which the sound in the specific section is masked, the change rate of the problem sound is masked while the sound before and after the problem sound is masked. It is possible to obtain the sound pressure fluctuation of the problem sound itself, so that it is possible to determine the sound when the fluctuation is large, determine the frequency, and perform processing such as specifying the sound source of the problem sound. Also, in general, if the sound pressure fluctuation is large, the sound becomes harsh as noise. Therefore, if the evaluation value is determined for each range of the sound pressure fluctuation value, the evaluation value of the problem sound can be output. It is possible to provide an excellent sound evaluation device which is not in the above.

With these sound evaluation devices, continuous sound and intermittent sound can be evaluated.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment for evaluating continuous sounds according to the present invention.

FIG. 2 is a block diagram showing a configuration of an embodiment for evaluating an intermittent sound according to the present invention.

FIG. 3 is a block diagram showing a configuration of an embodiment for processing waveform data according to the present invention.

FIG. 4 is a block diagram showing a configuration of an embodiment for determining an operation state of an engine according to the present invention.

FIG. 5 is a block diagram showing a configuration for evaluating an intermittent sound using a timing pulse according to the present invention.

FIG. 6 is a block diagram showing a configuration for evaluating a continuous sound using a timing pulse according to the present invention.

FIG. 7 is a block diagram showing a configuration of an example of the present invention.

FIG. 8 is a flowchart showing an outline of processing of the embodiment shown in FIG. 7;

9A and 9B are waveform diagrams showing an example of continuous sound evaluation in the configuration shown in FIG. 7; FIG. 9A is a diagram showing extraction of a sound of a specific frequency, and FIG. FIG. 9C is a diagram illustrating an example of comparison with the extracted sound pressure, and FIG. 9C is a diagram illustrating a ratio of each sound pressure.

FIG. 10 is a flowchart showing an example of a continuous sound evaluation process in the configuration shown in FIG. 7;

FIG. 11 is a waveform diagram showing a ratio of sound pressure in each order;
FIG. 11A is a diagram illustrating an example in which the sound after the order filter does not matter, and FIG. 11B is a diagram illustrating an example in which the eighth-order sound does not matter.

12 is a flowchart illustrating an example of a process for evaluating an intermittent sound with the configuration illustrated in FIG. 7;

13A and 13B are waveform diagrams illustrating an example of engine rotation fluctuation. FIG. 13A illustrates an example of a four-cycle engine, and FIG. 13B illustrates an example of a two-cycle engine. FIG.

14A and 14B are explanatory diagrams showing a timing pulse and a waveform of a sound, wherein FIG. 14A shows a state including an exhaust sound to be masked, and FIG. 14B shows a state in which the exhaust sound is masked; FIG.

FIG. 15 is a waveform chart showing a sound pressure fluctuation of each waveform shown in FIG.

FIG. 16 is a flowchart showing an example of a process for determining the state of the engine at the time indicated by each timing pulse shown in FIG.

17A and 17B are waveform diagrams showing an example of mask processing in the configuration shown in FIG. 7; FIG. 17A is a diagram showing an example of a mask waveform; and FIG. It is a figure showing the example where a part occurred.

18A and 18B are waveform diagrams showing an example of a minute partial masking process in the configuration shown in FIG. 7; FIG. 18A is a diagram showing a shape of a problem sound, and FIG. FIG.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 evaluation object (evaluation vehicle, sound source, vibration source) 2 sound storage means 4 specific frequency extraction means 6 sound pressure ratio calculation means 8 evaluation data output means 12 change rate calculation means 14 timing filter means

Claims (12)

[Claims] 1. A sound storage means for converting sound for a certain period of time emitted from one or a plurality of sound sources into digital data and storing the sound data, and a sound of a specific frequency from the sound data stored in the sound storage means. A specific frequency sound extracting means for extracting data; a sound pressure ratio calculating means for calculating a ratio of a sound pressure of the specific frequency sound extracted by the specific frequency sound extracting means to a sound pressure of the entire sound data; An evaluation data output unit that outputs a sound pressure ratio calculated by the ratio calculation unit as evaluation data of a sound emitted from the sound source. 2. The specific frequency sound extracting means includes an order filter frequency calculating unit for calculating a frequency of a sound to be extracted for each order with respect to a period of the timing pulse when a timing pulse is added to the sound data. The sound evaluation device according to claim 1, wherein: 3. A sound storage means for converting sound for a certain period emitted from one or a plurality of sound sources into digital data and storing the sound data, and a specific section of the sound data stored in the sound storage means. Timing filter means for masking a sound, change rate calculation means for calculating a change rate of sound data sound pressure in which a sound in a specific section is masked by the timing filter means, and change rate calculated by the change rate calculation means An evaluation data output unit for outputting as evaluation data of a sound emitted from the sound source. 4. A waveform data storage means for converting a vibration of a measurement object or a sound generated by the vibration into digital data and storing the waveform data, and a waveform data storage means provided in the waveform data storage means for a predetermined period of time. A timing pulse storage unit for storing a timing pulse indicating a synchronized operation timing of the measurement object; and a mask for masking a waveform level during the certain period based on the timing pulse stored in the timing pulse storage unit. A mask section specifying section for specifying a section; and a mask processing section for masking waveform data stored in the waveform data storage section for a section specified by the mask section specifying section to a predetermined level. Data processing device. 5. A mask processing unit comprising: a target level calculation unit for calculating a target level such as a level of a low frequency component of the waveform data; and a waveform data in the mask section is subdivided on a time axis and each subdivision is performed. 5. The waveform data processing device according to claim 4, further comprising: a subdivision mask unit that changes a level of the digitized waveform data to a target level calculated by the target level calculation unit. 6. A timing pulse storage unit for storing a timing pulse output according to a piston position of the engine during operation of the engine, and a rotational speed of the engine based on the timing pulse stored in the timing pulse storage unit. And an explosion top dead center determination unit that determines whether each timing is an explosion top dead center based on a time change of the rotation speed calculated by the rotation speed calculation unit. An engine operating state determination device, characterized in that: 7. A timing pulse storage means for storing a timing pulse output in accordance with a piston position of the engine during operation of the engine, a sound for a certain period of time emitted from the engine being converted into digital data, and a sound data And sound evaluation means for converting sound data stored in the sound storage means into evaluation data, wherein the sound evaluation means stores a timing pulse stored in the timing pulse storage means. An explosion top dead center specifying unit that specifies the timing of the explosion top dead center of the engine based on the explosion top dead center specified by the explosion top dead center specifying unit, A mask section setting section for setting a section to be determined as a mask section; and a sound section for the mask section set by the mask section setting section. A sound processing unit for masking sound data stored in the storage means; and an evaluation data output unit for outputting the sound data masked by the mask processing unit as sound evaluation data. apparatus. 8. The engine according to claim 1, wherein the top dead center specifying unit calculates a rotation speed of the engine based on a timing pulse stored in the timing pulse storage unit. 8. The sound evaluation apparatus according to claim 7, further comprising: an explosion top dead center determination function for determining whether each timing is an explosion top dead center based on the time change of the rotation speed. 9. The mask processing section, wherein the mask section is subdivided along the time axis and the sound pressure level of the sound data is maintained at a predetermined sound pressure level while maintaining the frequency ratio at each subdivided time interval. 8. The sound evaluation apparatus according to claim 7, further comprising a subdivision mask function for reducing the number of subdivisions. 10. A fluctuation rate output function for calculating a fluctuation rate of sound pressure of sound data masked by the mask processing section and outputting the fluctuation rate as the sound evaluation data. The sound evaluation device according to claim 7, further comprising: 11. An averaging function in which the evaluation data output unit averages sound data masked by the mask processing unit based on a timing pulse stored in the timing pulse storage unit in cycle units of the engine. The sound evaluation device according to claim 7, further comprising: 12. A timing pulse storage means for storing a timing pulse output in accordance with a piston position of the engine during operation of the engine, a sound for a certain period of time emitted from the engine being converted into digital data and a sound data Sound calculating means for calculating a frequency for each order based on the number of gear teeth of the engine and the timing pulse; and a sound for each frequency component of the order calculated by the order calculating means. Sound evaluation means for evaluating the sound, the sound evaluation means extracts the sound of the frequency calculated by the degree calculation means from the sound data, and a sound extraction unit for each degree, and the sound extraction unit for each degree A sound pressure ratio calculation unit that calculates a ratio of the sound pressure of the sound to the sound pressure of the whole sound stored in the sound storage unit; The sound evaluation apparatus characterized by comprising an evaluation data output unit for outputting a sound pressure ratio calculated I as sound evaluation data.