JP2021067605A - In-device abnormal coordinate detection system and in-device abnormal coordinate detection method - Google Patents

Abstract

To provide an in-device abnormal coordinate detection system and an in-device abnormal coordinate detection method for detecting the coordinate of an abnormal sound source from inside a device without having to decompose the device.SOLUTION: The in-device abnormal coordinate detection system comprises: a microphone for collecting sound waves; a storage unit 103 for storing the data of collected sound waves; and a calculation unit 105 for inputting the shape of an inspection object, the physical property value of each of materials constituting the inspection object and the position on inspection object of the microphone, building a CAE model from the inspection object and dividing the inspection object into lattice form, and, assuming that a sound wave constituting a reference is emitted from each lattice point as virtual abnormal sound source, calculating by simulation the waveforms of the sound waves collected by the microphone when the sound wave constituting the reference is emitted from each lattice point from the shape of the inspection object, the physical property value of each of materials constituting the inspection object and the position on inspection object of the microphone. The calculation unit squares a difference between the waveform of collected sound wave and the waveform of the calculated sound wave, and calculates a lattice point, of which the evaluation function obtained by taking time integration is minimum.SELECTED DRAWING: Figure 1

Description

The present invention relates to an in-device abnormal sound coordinate detection system and an in-device abnormal sound coordinate detection method.

When abnormal noise is generated inside an inspection target such as a machine, various methods have been attempted to identify the cause of the abnormal noise. For example, after determining whether the operating sound inside the machine is normal or abnormal, the position where the operating sound is generated is specified by the sound signals received by a plurality of microphones attached to the machine (Patent Documents). See 1 and 2). The positions where these operating sounds are generated are all specified as being arranged in an open space.

Japanese Unexamined Patent Publication No. 2017-133885 Japanese Unexamined Patent Publication No. 2006-208075

However, when the position of a different sound source is non-destructively detected when an abnormal noise is generated in a sealed mechanism such as a mission, an engine, or a differential that rotates, engages, or operates a gear or a clutch of an automobile, for example, these Since the device is a mixture of air, oil, metal, etc., the position of the different sound source cannot be detected only by the phase and intensity of the abnormal sound wave. Furthermore, there are cases where the abnormal noise disappears when decomposed, for example, the abnormal noise is caused by the presence of oil, and it is necessary to specify the position of the different sound source without disassembling it.

In view of the above problems, an object of the present invention is to provide an in-device abnormal sound coordinate detection system and an in-device abnormal sound coordinate detection method for detecting the coordinates of a different sound source from the inside of the device without disassembling the device. To do.

The first aspect of the present invention is an in-device abnormal sound coordinate detection system, which includes a microphone mounted on an inspection target to collect sound waves, a storage unit for storing the collected sound wave data, and an inspection. The shape of the object, the physical properties of the materials that make up the object to be inspected, and the position of the microphone on the object to be inspected are input and stored in the storage unit. Then, assuming that a reference sound wave is emitted using each lattice point as a virtual different sound wave, the shape of the inspection target stored in the storage unit, the physical property values of the materials constituting the inspection target, and the position of the microphone on the inspection target. Therefore, it is equipped with a calculation unit that calculates the waveform of the sound wave collected by the microphone when the reference sound wave is emitted from each grid point by simulation, and the calculation unit is further calculated as the waveform of the sound wave collected. The gist is to calculate the grid point at which the evaluation function obtained by squared the difference from the waveform of the sound wave and further time-integrated is the minimum value.

In the first aspect of the present invention, a second microphone attached at a position different from the microphone and collecting sound waves is further provided, and the calculation unit constitutes the shape of the inspection target and the inspection target stored in the storage unit. By simulating the waveform of the second sound wave collected by the second microphone when the reference sound wave is emitted from each lattice point from each physical property value of the material and the position on the inspection target of the second microphone. It is equipped with a calculation unit for calculating, and the calculation unit further squares the difference between the waveform of the collected sound wave and the calculated waveform of the second sound wave, and further obtains a second evaluation obtained by time integration. The function may be compared with the notation evaluation function, and the lattice point that is the minimum value of the second evaluation function and the evaluation function may be calculated.

In the first aspect of the present invention, the waveform of the reference sound wave may be set by inputting the waveform of the reference sound wave in the input unit.

A second aspect of the present invention is an in-device abnormal noise coordinate detection method, in which a step of collecting sound waves at a specific position on an inspection target and a CAE model of the inspection target are used to divide the inspection target into a grid pattern. Then, assuming that each grid point is used as a virtual different sound wave and a reference sound wave is emitted, from each grid point from the shape of the inspection target, each physical property value of the material constituting the inspection target, and a specific position on the inspection target. The difference between the step of calculating the waveform of the sound wave collected at a specific position on the inspection target when the reference sound wave is emitted by simulation and the waveform of the collected sound wave and the calculated sound wave waveform. The gist is to include a step of calculating a grid point at which the evaluation function obtained by squaring is squared and further time-integrated is the minimum value.

In the second aspect of the present invention, the step of collecting abnormal sound waves at a position different from a specific position on the inspection target, the shape of the inspection target, the physical property values of the materials constituting the inspection target, and the inspection. Simulates the waveform of the second sound wave collected at a position different from the specific position on the inspection target when the reference sound wave is emitted from each grid point from a position different from the specific position on the target. The difference between the step calculated by and the waveform of the collected sound wave and the calculated waveform of the second sound wave is squared, and the second evaluation function and the valence function obtained by time integration are compared. Then, a step of calculating the lattice point which is the minimum value of the second evaluation function and the evaluation function may be further provided.

According to the present invention, it is possible to provide an in-device abnormal sound coordinate detection system and an in-device abnormal sound coordinate detection method for detecting the coordinates of a different sound source from the inside of the device without disassembling the device.

It is a block diagram which shows an example of the structure of the abnormal sound coordinate detection system in an apparatus which concerns on embodiment of this invention. It is a figure which shows the physical property value with respect to the position inside the inspection target. It is a waveform diagram obtained by measuring the sound wave of the abnormal noise generated from the inside of the inspection object shown in FIG. 2 with a microphone. It is a flowchart explaining the method of detecting the abnormal sound coordinates in a device by the system for detecting the abnormal sound coordinates in a device which concerns on embodiment of this invention.

Next, an embodiment of the present invention will be described with reference to the drawings. In the description of the drawings according to the embodiment, the same or similar parts are designated by the same or similar reference numerals. However, it should be noted that the drawings are schematic, and the relationship between the thickness and the plane dimensions, the ratio of the thickness of each member, etc. are different from the actual ones. Therefore, the specific thickness and dimensions should be determined in consideration of the following explanation. In addition, it goes without saying that the drawings include parts having different dimensional relationships and ratios from each other.

Further, the embodiment illustrates an apparatus or method for embodying the technical idea of the present invention, and the technical idea of the present invention describes the configuration, arrangement, layout, etc. of each component as follows. It is not something that is specific to something. The technical idea of the present invention may be modified in various ways within the technical scope specified by the claims stated in the claims.

(Embodiment)
The method of detecting the abnormal sound coordinates in the device by the in-device abnormal sound coordinate detection system according to the embodiment of the present invention will be described below. FIG. 1 shows an example of the overall configuration of the in-device abnormal sound coordinate detection system 10 according to the embodiment. The in-device abnormal sound coordinate detection system 10 shown in FIG. 1 includes a first microphone 101 ₁ , a second microphone 101 ₂ , a first ADC 102 ₁ , a second ADC 102 ₂ , a storage unit 103, an input unit 104, and a calculation unit. It is composed of 105 and an output unit 106.

The first microphone 101 ₁ and the second microphone 101 ₂ are attached on the surface to be inspected. Here, the first microphone 101 ₁ and the second microphone 101 ₂ are attached at different positions from each other. The first microphone 101 ₁ and the second microphone 101 ₂ collect sound waves due to abnormal noise generated from the inside of the inspection target. The sound waves collected by the first microphone 101 ₁ and the second microphone 101 ₂ are converted from the analog data format to the digital data format by _{the first ADC 102 1} and the second ADC 102 _{2, respectively, and then stored.} It is input to and stored in unit 103. The input unit 104 is connected to the storage unit 103. From the input unit 104, the 3D model to be inspected, the physical property values of the materials of the parts constituting _{the inspection target, the positions of the first microphone 101 1} and the second microphone 101 _{2 on} the inspection target, and the like are input and stored. It is stored in the unit 103. The calculation unit 105 performs a calculation based on the data stored in the storage unit 103, and outputs the result to the output unit 106. The output to the output unit 106 may be, for example, display on a display or the like, printing on a paper medium or the like, output by voice, or the like.

Using the in-device abnormal noise coordinate detection system according to the embodiment shown in FIG. 1, an inspection target is a device in which oil is sealed inside a space sealed by an outer wall made of iron, for example, a car mission. , The case of detecting the coordinates of the abnormal noise generated from the inside of the inspection target will be described. Since the in-device abnormal sound coordinate detection system according to the present embodiment can specify the coordinates of different sound sources without disassembling the device, the device to be inspected is a device that is difficult or impossible to disassemble. For example, a device whose structure changes when disassembled, such as a device in which a fluid is sealed, a device in which a fluid or the like is sealed and an abnormal noise may be generated by the fluid, or the like is suitable.

FIG. 2 shows the physical property values for each position inside the inspection target. The physical property values shown in FIG. 2 are the speed of sound in the material. In FIG. 2, the speed of sound of iron arranged between _{position a and position a'and between position b and position b'is v 1} , and the speed of sound of oil sealed between position a'and position b is v ₂ . ..

_{The first microphone 101 1} and the second microphone 101 ₂ are attached to the positions a and b'of the inspection target in FIG. 2, respectively. In the inspection target shown in FIG. 2, the oil is sealed by an iron outer wall, and the first microphone 101 ₁ and the second microphone 101 ₂ attached to the positions a and b'are the surfaces of the outer wall to be inspected. It is in a state of being mounted on the top. An abnormal noise is generated from the inside of the inspection target, and the radiated abnormal noise sound wave is collected by _{the first microphone 101 1} and the second microphone 101 _2. The waveform of the collected sound wave is shown in FIG. FIG. 3A shows the waveform of the sound wave collected by _{the first microphone 101 1} attached to the position a, and FIG. 3B shows the _{sound wave collected by the second microphone 101 2 attached to the position b'.} The sound waves collected by the first microphone 101 ₁ and the second microphone 101 ₂ are stored in the storage unit 103 via _{the first ADC 102 1} and the second ADC 102 _2.

From the input unit 104, the 3D model to be inspected, the respective physical property values, and the positions of the first microphone 101 ₁ and the second microphone 101 _{2 on} the inspection target are input. These input data are stored in the storage unit 103. The calculation unit 105 models the inspection target by CAE (Computer Aided Engineering) model based on these data stored in the storage unit 103, and considers that the inspection target is divided into grid points by the finite difference method. Assuming that a reference sound wave is emitted from each lattice point as a virtual different sound source _{, the output waveforms to the first microphone 101 1} and the second microphone 101 ₂ are calculated and obtained by simulation. The sound wave calculated by the simulation is stored in the storage unit 103.

Examples of the reference sound wave include an impulse wave, a continuous sine wave, a rectangular wave, and a triangular wave. Since which waveform of the reference sound wave is effective depends on the type of abnormal sound, an arbitrary waveform may be set according to the type of abnormal sound to be analyzed. The waveform to be set may be any of the above waveforms such as an impulse wave, or a combination of the above waveforms.

Here, the calculation unit 105 uses the waveforms stored in the storage unit 103 from each lattice point obtained by simulation and the waveforms collected by _{the first microphone 101 1} and the second microphone 101 _{2 shown in FIG.} Of the waveforms from each grid point obtained by the simulation, the grid point that is the closest to the collected waveform shown in FIG. 3 can be regarded as the estimated coordinates of the different sound source. To compare the waveform from each grid point by simulation with the waveform collected from each grid point shown in FIG. 3, take the difference between the waveform from each grid point by simulation and the waveform collected from each grid point shown in FIG. The closest waveform is determined by taking the sum of squares.

In the following, the difference between the waveform from each grid point by simulation _{and the waveform collected by the first microphone 101 1} and the second microphone 101 ₂ shown in FIG. 3 is taken, and the sum of squares of this difference is taken and compared. .. Figure 3 a waveform obtained by collecting the first microphone 101 ₁ shown in _{(a) g r1 (t)} , a function of the square of the difference between the resulting waveform _{g sim1} (t) of the simulation _{f a1} ( If t), _fa1 (t) is represented by a function that takes time t as an argument.
_{f a1 (t) = (g} r1 (t) -g sim1 (t)) 2
Is.

As the evaluation _function, attention is paid to _F a1 (t) is the time integral of _f a1 (t). here,
F _a1 (t) = ∫f _a1 (t) dt
Is. The waveform collected by the second microphone 101 ₂ is also subjected to the same processing as the processing for the waveform collected by the _{first microphone 101 1 described above to obtain the evaluation function Fa2} (t). Evaluation function The evaluation functions F _a1 (t) and F _a2 (t) at each lattice point are compared, and the lattice point _{having the smallest value among the evaluation functions F a} (t) and F _a2 (t) at each lattice point is determined. It can be regarded as the closest to the coordinate values of different sound sources. It is possible to estimate the coordinate values of different sound sources only from the waveform collected by the first microphone 101 _{1 , and by using the waveform collected by the second microphone 101 2} to estimate the coordinate values of different sound sources. , The accuracy of position measurement is improved.

Here has been exploring the search start point of the minimum value of F _{a (t)} and F _{a2 (t)} from the coordinates x, because in general the function F _a there are a plurality of minimum values, F a _(t ) And _Fa2 (t), evolutionary computation in which multiple calculation start points are provided before and after the initial setting of coordinates, more specifically, genetic algorithm (GA). And taboo search method (TS) can be used. Further, when adjusting a plurality of physical property values, that is, the speed of sound at the same time, the steepest descent method or the conjugate gradient method (CG method) in which the convergence calculation is performed using each partial differential can be used. Further, if there is a problem with the convergence speed, the conjugate gradient method with incomplete Cholesky decomposition (ICCG method) can be used.

The method for detecting abnormal noise coordinates in the device according to the embodiment will be described with reference to the flowchart shown in FIG. In the following flowchart, a method of calculating the coordinates of different sound sources using only _{the first microphone 101 1 will be described.} In step S401, collects sound waves of noise in the first microphone 101 _1. The detected sound wave is converted from the analog data format to the digital data format by _{the first ADC 102 1, and then input to the storage unit 103 and stored.}

In step S402, the inspection target is CAE modeled, and the inspection target is divided into grid points by the finite difference method.

In step S403, the calculation unit 105 determines the 3D model of the inspection target stored in advance in the storage unit 103, the physical property values of the materials of the parts constituting the inspection target, and the first microphone 101 ₁ attached to the inspection target. It is assumed that a reference sound wave is emitted from each grid point set in step S402 as a virtual different sound source from the data of the position on the inspection target, _{and the output waveform to the first} microphone 1011 is calculated and obtained by simulation. The calculated output waveform is input to the storage unit 103 and stored.

In step S404, an evaluation function is obtained from the difference between the waveform obtained by collecting the sound wave irradiated in step S401 with the first microphone 101 _{1 and the waveform calculated by simulation in step S403.}

In step S405, the evaluation functions for all the grid points are compared. The grid point with the smallest evaluation function can be regarded as the closest to the coordinate values of different sound sources.

As described above, it goes without saying that the present invention includes various embodiments not described here. Therefore, the technical scope of the present invention is defined only by the matters specifying the invention relating to the reasonable claims from the above description.

10 In-device abnormal noise coordinate detection system 101 ₁ First microphone 101 ₂ Second microphone 102 ₁ First ADC
102 ₂ Second ADC
103 Storage unit 104 Input unit 105 Calculation unit 106 Output unit

Claims (5)

A microphone that is mounted on the inspection target and collects sound waves,
A storage unit that stores the sound wave data collected and
An input unit for inputting the shape of the inspection target, each physical property value of the material constituting the inspection target, and the position of the microphone on the inspection target, and storing the microphone in the storage unit.
It is assumed that the inspection target is CAE modeled, the inspection target is divided into a plurality of lattice points, and a reference sound wave is emitted from each lattice point as a virtual different sound wave, and the shape of the inspection target stored in the storage unit. , The waveform of the sound wave collected by the microphone when the reference sound wave is emitted from each of the lattice points from the respective physical property values of the material constituting the inspection target and the position of the microphone on the inspection target. Is provided with a calculation unit that calculates the above by simulation, and the calculation unit further squares the difference between the waveform of the collected sound wave and the waveform of the calculated sound wave, and further obtains an evaluation obtained by time integration. An in-device abnormal sound coordinate detection system characterized in that the function calculates the grid point at which the minimum value is obtained. A second microphone, which is attached at a position different from the microphone and collects the sound wave, is further provided, and the calculation unit has a shape of the inspection target stored in the storage unit and a material constituting the inspection target, respectively. The waveform of the second sound wave collected by the second microphone when the reference sound wave is emitted from each lattice point from the physical property value of the second microphone and the position of the second microphone on the inspection target. It is provided with a calculation unit calculated by simulation, and the calculation unit further squares the difference between the waveform of the collected sound wave and the waveform of the calculated second sound wave, and further obtains it by time integration. The device according to claim 1, wherein the second evaluation function is compared with the evaluation function, and the lattice point that is the minimum value of the second evaluation function and the evaluation function is calculated. Abnormal sound coordinate detection system. The in-device abnormal sound coordinate detection system according to claim 1 or 2, wherein the reference sound wave waveform is set by inputting the reference sound wave waveform in the input unit. The step of collecting sound waves at a specific position on the inspection target,
It is assumed that the inspection target is CAE modeled, the inspection target is divided into a plurality of grid points, and a reference sound wave is emitted from each grid point as a virtual different sound source, and the shape of the inspection target and the inspection target are configured. The waveform of the sound wave collected at the specific position on the inspection target when the reference sound wave is emitted from each of the lattice points from each physical property value of the material and the specific position on the inspection target. Steps calculated by simulation and
It is provided with a step of squaring the difference between the waveform of the collected sound wave and the waveform of the calculated sound wave, and further calculating the grid point at which the evaluation function obtained by taking the time integration is the minimum value. A characteristic method for detecting abnormal sound coordinates in a device. A step of collecting abnormal sound waves at a position different from the specific position on the inspection target, and
The case where the reference sound wave is emitted from each of the lattice points from a position different from the shape of the inspection target, each physical property value of the material constituting the inspection target, and a specific position on the inspection target. The step of calculating the waveform of the second sound wave collected at a position different from the specific position on the inspection target by simulation, and
The difference between the waveform of the collected sound wave and the waveform of the calculated second sound wave is squared, and the second evaluation function obtained by time integration is compared with the evaluation function. The in-device abnormal sound coordinate detection method according to claim 4, further comprising a second evaluation function and a step of calculating a grid point that is the minimum value of the evaluation functions.

Images