Sound source estimation method

Abstract

To increase estimation accuracy of a source of sound.SOLUTION: A sound source estimation method performs: first processing of acquiring a frequency of the largest sound at each time as first sound information from sound data during a predetermined period to estimate a source of the sound on the basis of the first sound information (S13); second processing of generating processed sound data from the sound data and acquiring a frequency of the largest sound at each time as second sound information from the processed sound data to estimate the source of the sound on the basis of the second sound information (S15); and third processing of estimating the source of the sound on the basis of an estimation result in the first processing and an estimation result in the second processing (S17).SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a sound source estimation method. [Background technology]

[0002] Patent Document 1 discloses a sound source estimation system that estimates candidates for the source of an abnormal sound. In this system, a maximum sound frequency, which is a frequency indicating the loudest sound, is extracted from frequency sound data that represents changes in a frequency spectrum during a predetermined period. An order is calculated based on the maximum sound frequency and the rotation speed of the part. Then, the part having the order closest to the maximum sound order is estimated as the source of the abnormal sound. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2022-100139 A Summary of the Invention [Problem to be solved by the invention]

[0004] In the above system that estimates the source of an abnormal sound using the maximum sound frequency, it becomes difficult to estimate the source of the abnormal sound when noise other than the abnormal sound is loud. [Means for solving the problem]

[0005] A sound source estimation method for solving the above problem has a control device execute a first process of acquiring the frequency of the loudest sound at each time from sound data during a predetermined period as first sound information and estimating a source of the sound based on the first sound information, a second process of generating processed sound data from the sound data by excluding sound data in the frequency range of the sound emitted from the sound source estimated in the first process, acquiring the frequency of the loudest sound at each time from the processed sound data as second sound information and estimating the source of the sound based on the second sound information, and a third process of estimating the source of the sound based on the estimation results of the first process and the second process. Effect of the Invention

[0006] According to the above sound source estimation method, it is possible to improve the accuracy of estimating the sound source. [Brief description of the drawings]

[0007] [Figure 1] FIG. 1 is a flow chart illustrating an embodiment of a sound source estimation method. [Diagram 2] In FIG. 2, (A) is a diagram showing an example of the first sound information, (B) is a diagram showing a transition of the noise frequency, and (C) is a diagram showing the estimation result of the source of the abnormal sound in the first processing. [Diagram 3] In FIG. 3, (A) is a diagram showing an example of the second sound information, and (B) is a diagram showing the estimation result of the source of the abnormal sound in the second processing. [Figure 4] FIG. 4 is a diagram showing the results of estimation of the source of abnormal noise in the third process. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0008] Hereinafter, one embodiment of a sound source estimation method will be described with reference to FIGS. <Configuration of sound source estimation method> One example of a sound source estimation method is a method for estimating the source of an abnormal sound generated in a vehicle. When an occupant of the vehicle notices the generation of an abnormal sound, the occupant operates, for example, a button provided in the vehicle. This button is an operation unit for starting a series of processes for identifying the source of the abnormal sound.

[0009] A series of processes for identifying the source of an abnormal noise is shown in Figure 1. The series of processes is executed by the control device. In step S11, the control device acquires sound data for a predetermined period of time by having a sound collection device such as a microphone collect sound. The sound data for the predetermined period of time is time-series data including the loudness and frequency of the collected sound.

[0010] In step S13, the control device executes a first process of estimating the source of the abnormal noise by analyzing sound data during a predetermined period. Specifically, the control device acquires the maximum sound frequency at each time from the sound data during a predetermined period as the first sound information (S131). The maximum sound frequency is the frequency of the loudest sound. Note that FIG. 2(A) illustrates an example of the first sound information.

[0011] Next, the control device estimates the source of the abnormal noise based on the first information (S133). In Fig. 2(B), a first line L1 indicates the frequency change of switching noise that occurs when an on-board switch is operated. A second line L2 indicates the frequency change of 48th order noise, and a third line L3 indicates the frequency change of 24th order noise as the motor is driven.

[0012] The control device acquires the number of plots included in a predetermined frequency range including the first line L1, the number of plots included in a predetermined frequency range including the second line L2, and the number of plots included in a predetermined frequency range including the third line L3 from among the multiple plots shown in Fig. 2. Then, the control device calculates the matching rate of the switching noise, the matching rate of the 48th order of the motor noise, and the matching rate of the 24th order of the motor noise. For example, the control device can calculate the matching rate of the switching noise by dividing the number of plots included in the predetermined frequency range including the first line L1 by the total number of plots in Fig. 2.

[0013] 2C shows an example of an estimation result of the source of the abnormal sound based on the first sound information. In this case, the control device indicates that the switch is most likely to be the source of the abnormal sound.

[0014] Then, when the control device ends the first process, the process proceeds to the next step S15. In step S15, the control device executes a second process of estimating the source of the abnormal noise by analyzing the processed sound data, which will be described later.

[0015] Specifically, the control device generates processed sound data based on the sound data acquired in step S11 and the first sound information (S151). Specifically, the control device generates processed sound data by excluding sound data in the frequency range of the sound emitted from the source of the abnormal sound estimated in the first process from the sound data.

[0016] Next, the control device acquires the maximum sound frequency at each time from the processed sound data as second sound information (S153). Note that FIG. 3(A) illustrates an example of the second sound information. The control device then estimates the source of the abnormal sound based on the second information (S155). The method of estimating the source here is the same as the estimation method used in step S133. Figure 3(B) shows an example of the result of estimating the source of the abnormal sound based on the second information.

[0017] When the control device ends the second process, the process proceeds to the next step S17. In step S17, the control device executes a summing process, which corresponds to a "third process" for estimating the source of a sound based on the estimation result in the first process and the estimation result in the second process.

[0018] FIG. 4 illustrates the result of the summation process. Specifically, the control device multiplies both the first match rate, which is the match rate in the first process, and the second match rate, which is the match rate in the second process, by a weighting coefficient. The control device then calculates the sum of the first match rate and the second match rate after correction with the weighting coefficient as the match rate. Note that in FIG. 4, the weighting coefficient for the first match rate is 1, and the weighting coefficient for the second match rate is 0.8. However, these are merely examples, and different values ​​may be set as the weighting coefficients.

[0019] When the control device finishes the summation process, the process proceeds to step S19. In step S19, the control device notifies the occupant of the result of the summation process. For example, the control device may display the result on a display mounted on the vehicle.

[0020] <Effects of this embodiment> In this embodiment, the source of the abnormal sound is estimated taking into consideration sounds other than the loudest sound. This makes it possible to identify multiple sources of abnormal sound when there are multiple sources of abnormal sound. [Explanation of symbols]

[0021] S13: first process, S15: second process, S17: summation process.

Claims

[Claim 1] a first process of acquiring a frequency of a loudest sound at each time from sound data during a predetermined period as first sound information, and estimating a source of the sound based on the first sound information; a second process of generating processed sound data from the sound data by excluding sound data in a frequency range of the sound emitted from the sound source estimated in the first process, obtaining a frequency of the loudest sound at each time from the processed sound data as second sound information, and estimating the sound source based on the second sound information; and a third process of estimating a source of the sound based on an estimation result in the first process and an estimation result in the second process. Sound source estimation method.

Images