Abnormal sound determination program, abnormal sound determination device, abnormal sound determination system, and abnormal sound determination method

Abstract

To provide an abnormal sound determination program, an abnormal sound determination device, an abnormal sound determination system, and an abnormal sound determination method for determining occurrence of an abnormal sound.SOLUTION: An abnormal sound determination program causes a control device to execute the steps of: generating a specific loudness differential value by differentiating, by time, a specific loudness signal, which is obtained by calculating specific loudness corresponding to sound continuously generated by an internal combustion engine; and determining an occurrence of an abnormal sound in the internal combustion engine on the basis of the specific loudness differential value.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to an abnormal sound detection program, an abnormal sound detection device, an abnormal sound detection system, and an abnormal sound detection method. [Background technology]

[0002] A known example of an invention related to a conventional abnormal sound detection program is an abnormal sound analysis device described in Patent Document 1. In this abnormal sound analysis device, a processor performs a process of identifying abnormal sounds that occur periodically, a process of removing data on the identified abnormal sounds, and a process of calculating data on the abnormal sounds based on the processed data. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2016-27324 A Summary of the Invention [Problem to be solved by the invention]

[0004] As described above, there is a demand for determining the occurrence of abnormal sounds.

[0005] SUMMARY OF THE PRESENT DISCLOSURE An object of the present invention is to provide an abnormal sound detection program, an abnormal sound detection device, an abnormal sound detection system, and an abnormal sound detection method for determining the occurrence of an abnormal sound. [Means for solving the problem]

[0006] The first aspect is a generating step of generating a specific loudness differential value by differentiating with respect to time a specific loudness signal obtained by calculating a specific loudness corresponding to a sound continuously generated by the internal combustion engine; a determination step of determining whether an abnormal sound occurs in the internal combustion engine based on the specific loudness differential value; causing the control device to execute This is a program for detecting abnormal sounds.

[0007] The second aspect is, In the determination step, it is determined that an abnormal sound has occurred when the loudness differential value becomes larger than a first threshold value, causing the specific loudness signal to become larger than a second threshold value. 1 is a program for detecting an abnormal sound according to a first aspect of the present invention;

[0008] The third aspect is, The determining step includes: a frequency calculation step of calculating an abnormal sound frequency indicating the number of times the abnormal sound occurs per unit time; a level determination step of determining an abnormal sound level indicating the degree of the abnormal sound based on the abnormal sound frequency and the specific loudness signal at the time when the abnormal sound occurs; Contains The abnormal sound detection program according to the second aspect.

[0009] The fourth aspect is the determination step includes a first zero-cross detection step, a second zero-cross detection step, and a determination value calculation step, which are repeated during a measurement period; In the first zero cross detection step, a first zero cross point where the specific loudness differential value changes from negative to positive is detected, In the second zero cross detection step, a second zero cross point at which the specific loudness differential value changes from positive to negative for the first time after the first zero cross point is detected, In the determination value calculation step, a determination value is calculated using a maximum value of the specific loudness differential value between the first zero cross point and the second zero cross point and a maximum value of the specific loudness signal between the first zero cross point and the second zero cross point. 1 is a program for detecting an abnormal sound according to a first aspect of the present invention;

[0010] The fifth aspect is The determination step further includes a determination value calculation step, a time rate level calculation step, and a threshold value calculation step, In the judgment value calculation step, a plurality of judgment values ​​are calculated based on the specific loudness differential value, In the time rate level calculation step, a predetermined % time rate level is calculated based on the plurality of judgment values ​​obtained in the judgment value calculation step, In the threshold calculation step, a third threshold is calculated using the predetermined percentage time level and a predetermined coefficient; In the determination step, it is determined that an abnormal sound has occurred for a determination value that is greater than the third threshold value among the plurality of determination values. 1 is a program for detecting an abnormal sound according to a first aspect of the present invention;

[0011] The sixth aspect is: An abnormal sound detection device, a generating step of generating a specific loudness differential value by differentiating with respect to time a specific loudness signal obtained by calculating a specific loudness corresponding to a sound continuously generated by the internal combustion engine; a determination step of determining whether an abnormal sound occurs in the internal combustion engine based on the specific loudness differential value; Execute This is an abnormal sound detection device.

[0012] The seventh aspect is: The abnormal sound detection device according to the sixth aspect, A plurality of microphones; Equipped with Each of the plurality of microphones is arranged at a different spatial position and captures sounds continuously generated by the internal combustion engine. This is an abnormal sound detection system.

[0013] The eighth aspect is: The plurality of microphones acquire sounds successively generated by different parts of the internal combustion engine. The abnormal sound detection system according to the seventh aspect.

[0014] The ninth aspect is: a generating step of generating a specific loudness differential value by differentiating with respect to time a specific loudness signal obtained by calculating a specific loudness corresponding to a sound continuously generated by the internal combustion engine; a determination step of determining whether an abnormal sound occurs in the internal combustion engine based on the specific loudness differential value; causing the control device to execute This is a method for detecting abnormal sounds. Effect of the Invention

[0015] According to the present invention, it is possible to determine whether or not an abnormal sound has occurred. [Brief description of the drawings]

[0016] [Figure 1] FIG. 1 is a block diagram of the abnormal sound detection system 1. [Diagram 2] FIG. 2 is a diagram showing the waveform of a sound pressure signal Sig1a, a waveform of a specific loudness signal Sig11a, and a waveform of a specific loudness differential value Sig12a. [Diagram 3] FIG. 3 is a diagram showing waveforms of the specific loudness differential values ​​Sig12a to Sig12i. [Figure 4] FIG. 4 is a flowchart showing the operations executed by the processor 10. As shown in FIG. [Diagram 5] FIG. 5 is a flowchart showing a subroutine of step S4 in FIG. [Figure 6] FIG. 6 is a graph used to determine the abnormal sound level. [Figure 7] FIG. 7 is a flowchart showing the subroutine of step S4. [Figure 8] FIG. 8 is a diagram showing the waveform of the specific loudness signal Sig11a and the waveform of the specific loudness differential value Sig12a. [Figure 9] FIG. 9 is a diagram showing a procedure for calculating the predetermined % time level Y. In FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0017] (Embodiment) [Configuration of abnormal sound detection system 1] The configuration of an abnormal sound detection system 1 according to an embodiment of the present invention will be described below with reference to the drawings. FIG.

[0018] The abnormal sound detection system 1 performs a determination regarding the occurrence of an abnormal sound in an internal combustion engine 100. In this embodiment, the abnormal sound detection system 1 determines an abnormal sound level indicating the degree of an abnormal sound generated in the internal combustion engine 100. In this embodiment, the abnormal sound is a knocking sound generated in the internal combustion engine 100. The internal combustion engine 100 may be a multi-cylinder engine or a single-cylinder engine, for example, an in-line three-cylinder engine having a displacement of 1000 cc. As shown in FIG. 1, the abnormal sound detection system 1 includes a processor 10, microphones 12a to 12i, an amplifier 14, an A / D (analog-to-digital converter) 16, a storage device 18, and a display device 20.

[0019] An abnormal sound detection program is stored in the storage device 18. The storage device 18 is, for example, a solid state drive (SSD) or a hard disk drive (HDD).

[0020] The processor 10 is a control device (abnormal sound detection device) that executes an abnormal sound detection program stored in the storage device 18. The processor 10 is, for example, a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit).

[0021] The display device 20 displays an image based on a video signal output from the processor 10. The display device 20 is, for example, a liquid crystal display or an organic EL display. The above-described processor 10, storage device 18, and display device 20 are realized, for example, by a personal computer.

[0022] The microphones 12a to 12i are arranged at different spatial positions and capture sounds continuously generated by the internal combustion engine 100. More specifically, the microphones 12a to 12i capture sounds continuously generated by different parts of the internal combustion engine 100. The microphones 12a to 12i generate sound pressure signals Sig1a to Sig1i indicating sounds generated by the internal combustion engine 100. The sound pressure signals Sig1a to Sig1i output from the microphone array 12 are analog signals. The microphones 12a to 12i are arranged in a matrix of 3 rows and 3 columns. The arrangement of the microphones 12a to 12i is determined in consideration of the frequency band of the signal to be observed and acoustic reflection in the observation environment. The interval between the microphones 12a to 12i is half the wavelength of the signal wavelength to be observed from the observation object to several times or more. In addition, it is preferable that the microphones 12a to 12i are arranged so as not to be affected by interference from reflected sound. Specifically, the microphones 12a to 12i are arranged so as not to be located in a valley of interference fringes caused by reverberation. The microphones 12a to 12i do not need to be spaced apart uniformly. For example, the microphones 12a to 12i can be arranged at positions 200 mm to 500 mm from the engine. In this embodiment, the microphones 12a to 12i are arranged at positions 200 mm from the engine. In this embodiment, the distance between the centers of two adjacent microphones is 100 mm.

[0023] The amplifier 14 amplifies the amplitude of the sound pressure signals Sig1a to Sig1i generated by the microphones 12a to 12i, respectively.

[0024] The A / D 16 converts the sound pressure signals Sig1a to Sig1i into digital signals, whereby the processor 10 acquires the sound pressure signals Sig1a to Sig1i, which are digital signals.

[0025] [Operation of abnormal sound detection system 1] Next, an overview of the operation of the abnormal sound detection system 1 will be described with reference to Fig. 2 and Fig. 3. Fig. 2 is a diagram showing the waveform of the sound pressure signal Sig1a, the waveform of the specific loudness signal Sig11a, and the waveform of the specific loudness derivative value Sig12a. The horizontal axis of Fig. 2 is time. The vertical axis of Fig. 2 is the sound pressure signal Sig1a, the value of the specific loudness signal Sig11a, and the specific loudness derivative value Sig12a. Fig. 3 is a diagram showing the waveforms of the specific loudness derivative values ​​Sig12a to Sig12i. The horizontal axis of Fig. 3 is time. The vertical axis of Fig. 3 is the value of the specific loudness derivative value Sig12a.

[0026] The loudness of a sound perceived by humans varies depending on the frequency of the sound even if the sound pressure is the same. For this reason, it is difficult for the abnormal sound detection system 1 to determine whether an abnormal sound has occurred based only on the sound pressure. Therefore, the abnormal sound detection system 1 uses the specific loudness signal Sig11a. This enables the abnormal sound detection system 1 to perform a determination regarding the occurrence of an abnormal sound that is close to that made by an experienced person. Therefore, the abnormal sound detection system 1 performs the following operations.

[0027] The processor 10 performs loudness analysis on the sound continuously generated by the internal combustion engine, and calculates the specific loudness to generate a specific loudness signal Sig11a (see FIG. 2). The specific loudness is the loudness of a specific frequency among the loudness of the sound continuously generated by the internal combustion engine. The specific frequency is a 10 kHz band. The 10 kHz band means, for example, the 10 kHz critical band defined by ISO532. In the loudness analysis, processing is performed in accordance with the standard "DIN 45631 / A1 2010-3".

[0028] Next, the processor 10 differentiates the specific loudness signal Sig11a obtained by calculating the specific loudness for the sound continuously generated by the internal combustion engine with respect to time to generate a specific loudness differential value Sig12a. The specific loudness differential value Sig12a indicates the rate of change of the specific loudness signal Sig11a.

[0029] Next, the processor 10 judges whether an abnormal sound has occurred in the internal combustion engine 100 based on the specific loudness signal Sig11a and the specific loudness differential value Sig12a. In this specification, the judgment regarding the occurrence of an abnormal sound includes the judgment of the level of the abnormal sound as well as the presence or absence of the occurrence of the abnormal sound. Specifically, the processor 10 judges that an abnormal sound has occurred when the specific loudness signal Sig11a becomes larger than the second threshold value TH2 as a result of the specific loudness differential value Sig12a becoming larger than the first threshold value TH1.

[0030] Here, the microphones 12a-12i output sound pressure signals Sig1a-Sig1i, respectively. Therefore, the processor 10 acquires the sound pressure signals Sig1a-Sig1i. The processor 10 then executes the same process on the sound pressure signals Sig1b-Sig1i as on the sound pressure signal Sig1a. That is, as shown in FIG. 3, the processor 10 generates specific loudness differential values ​​Sig12b-Sig12i. Then, the processor 10 makes a determination regarding the occurrence of an abnormal sound in the internal combustion engine 100 based on the specific loudness signals Sig11b-Sig11i and the specific loudness differential values ​​Sig12b-Sig12i in addition to the specific loudness signal Sig11a and the specific loudness differential value Sig12a.

[0031] Next, specific operations of the abnormal sound detection system 1 will be described with reference to the drawings. FIG. 4 is a flowchart showing operations executed by the processor 10. FIG. 5 is a flowchart showing a subroutine of step S4 in FIG. 4. FIG. 6 is a graph used to determine the abnormal sound level. The vertical axis represents the abnormal sound frequency, and the horizontal axis represents the specific loudness signal. The following operations are realized by an abnormal sound detection program stored in the storage device 18 causing the processor 10 (control device / abnormal sound detection device) to execute the flowchart.

[0032] First, the processor 10 acquires the sound pressure signals Sig1a to Sig1i (acquisition step: step S1). The processor 10 generates the specific loudness signals Sig11a to Sig11i by calculating the specific loudness corresponding to the sounds continuously generated by the internal combustion engine 100 (first generation step: step S2). Note that the details of the loudness analysis have already been explained, so the explanation will be omitted.

[0033] Next, the processor 10 differentiates the specific loudness signals Sig11a to Sig11i with respect to time to generate specific loudness differential values ​​Sig12a to Sig12i (second generation step: step S3). In this way, the processor 10 obtains the specific loudness differential values ​​Sig12a to Sig12i having the waveforms of FIG.

[0034] Next, the processor 10 performs a determination regarding the occurrence of an abnormal sound in the internal combustion engine 100 based on the specific loudness signals Sig11a to Sig11i and the specific loudness differential values ​​Sig12a to Sig12i (determination step: step S4). Step S4 will be described below with reference to FIG.

[0035] 2, the processor 10 determines whether the specific loudness differential value Sig12a is greater than the first threshold value TH1 (step S11). If the specific loudness differential value Sig12a is greater than the first threshold value TH1, the process proceeds to step S12. If the specific loudness differential value Sig12a is not greater than the first threshold value TH1, the process proceeds to step S14.

[0036] If the specific loudness differential value Sig12a is greater than the first threshold TH1, the processor 10 determines whether the specific loudness signal Sig11a is greater than the second threshold TH2 (step S12). If the specific loudness signal Sig11a is greater than the second threshold TH2, the process proceeds to step S13. If the specific loudness signal Sig11a is not greater than the second threshold TH2, the process proceeds to step S14.

[0037] When the specific loudness signal Sig11a is greater than the second threshold TH2, the processor 10 determines that an abnormal sound has occurred (step S13). An abnormal sound has a property that the value of the specific loudness signal increases as the abnormal sound level increases, and the sound has a property of suddenly increasing in volume. Therefore, in the determination step, the processor 10 determines that an abnormal sound has occurred when the specific loudness differential value Sig12a becomes greater than the first threshold TH1, causing the specific loudness signal Sig11a to become greater than the second threshold TH2. After this, the process proceeds to step S15.

[0038] In step S14, the processor 10 determines that no abnormal sound has occurred (step S14). After that, the process proceeds to step S15.

[0039] In step S15, the processor 10 determines whether or not the processes of steps S11 to S14 have been performed on the specific loudness signal Sig11a and the specific loudness differential value Sig12a over the determination period (step S15). If the processes of steps S11 to S14 over the determination period have been completed, the process proceeds to step S16. If the processes of steps S11 to S14 over the determination period have not been completed, the process returns to step S11. The determination period is a time period long enough to determine whether an abnormal sound has occurred, and is, for example, several seconds or more and several tens of seconds or less. In this embodiment, the determination period is 3.2 seconds.

[0040] When the processing of steps S11 to S14 over the determination period is completed, the processor 10 determines whether or not a determination has been made regarding the occurrence of an abnormal sound for all the specific loudness signals Sig11a to Sig11i and all the specific loudness differential values ​​Sig12a to Sig12i (step S16). When a determination regarding the occurrence of an abnormal sound has been made regarding all the specific loudness signals Sig11a to Sig11i and all the specific loudness differential values ​​Sig12a to Sig12i, the process proceeds to step S17. When a determination regarding the occurrence of an abnormal sound has not been made regarding all the specific loudness signals Sig11a to Sig11i and all the specific loudness differential values ​​Sig12a to Sig12i, the process returns to step S11. In this case, the processor 10 makes a determination regarding the occurrence of an abnormal sound for the next specific loudness signal and specific loudness differential value. Then, the processor 10 repeats steps S11 to S15 until the determination regarding the occurrence of an abnormal sound is performed for all the specific loudness signals Sig11a to Sig11i and all the specific loudness differential values ​​Sig12a to Sig12i.

[0041] When the determination regarding the occurrence of abnormal sounds has been performed for all the specific loudness signals Sig11a-Sig11i and all the specific loudness differential values ​​Sig12a-Sig12i, the processor 10 calculates an abnormal sound frequency indicating the number of times abnormal sounds have occurred per unit time (frequency calculation step: step S17). Specifically, the processor 10 calculates the abnormal sound frequency by dividing the number of times it has been determined in steps S11-S16 that abnormal sounds have occurred by the determination period. The abnormal sound frequency is calculated for each of the specific loudness signals Sig11a-Sig11i. Thus, nine abnormal sound frequencies are calculated.

[0042] Next, the processor 10 judges the abnormal sound level indicating the degree of the abnormal sound based on the abnormal sound frequency and the specific loudness signal when the abnormal sound occurs (level judgment step, step S18). Specifically, the storage device 18 stores a table corresponding to the graph of FIG. 6. The graph of FIG. 6 includes curves indicating abnormal sound levels KL0 to KL2. The abnormal sound levels KL0 to KL2 indicate the degree of the abnormal sound. The degree of the abnormal sound increases in the order of the abnormal sound level KL0, the abnormal sound level KL1, and the abnormal sound level KL2. The abnormal sound level KL0 is a state in which the skilled person cannot hear the abnormal sound at all. The abnormal sound level KL1 is a state in which the skilled person can slightly hear the abnormal sound. The abnormal sound level KL2 is a state in which the skilled person can clearly hear the abnormal sound.

[0043] The processor 10 plots points determined by each of the abnormal sound frequencies calculated in step S17 and the values ​​of the specific loudness signals Sig11a to Sig11i when the abnormal sound occurs on the graph of FIG. 6. In this embodiment, nine pairs of abnormal sound frequencies and specific loudness signals Sig11a to Sig11i exist, so nine points are plotted. Then, the processor 10 determines the abnormal sound level based on the curve of the abnormal sound level KL0, the curve of the abnormal sound level KL1, the curve of the abnormal sound level KL2, and the nine points. If the point is located below the curve of the abnormal sound level KL0, the processor 10 determines that the abnormal sound level is 0. If the point is located below the curve of the abnormal sound level KL1 and above the abnormal sound level KL0, the processor 10 determines that the abnormal sound level is 1. If the point is located below the curve of the abnormal sound level KL2 and above the abnormal sound level KL1, the processor 10 determines that the abnormal sound level is 2. The determination is made for each of the nine points, after which the process ends.

[0044] [effect] The abnormal sound detection system 1 can determine whether an abnormal sound has occurred. More specifically, an abnormal sound has a property of suddenly increasing in volume. That is, the specific loudness derivative value Sig12a has a peak waveform. Therefore, the processor 10 determines whether an abnormal sound has occurred in the internal combustion engine 100 based on the specific loudness derivative value Sig12a. In this way, the abnormal sound detection system 1 can determine whether an abnormal sound has occurred.

[0045] The abnormal sound detection system 1 can accurately determine whether an abnormal sound has occurred. More specifically, an abnormal sound has a property that the value of the specific loudness signal increases as the abnormal sound level increases. Furthermore, an abnormal sound has a property that the specific loudness increases rapidly in a short time, and in particular, the specific loudness increases rapidly in a short time. That is, the specific loudness signal Sig11a has a large value, and the specific loudness differential value Sig12a has a peak waveform. Therefore, in the determination step, the processor 10 determines that an abnormal sound has occurred when the specific loudness signal Sig11a becomes larger than the second threshold value TH2 as a result of the specific loudness differential value Sig12a becoming larger than the first threshold value TH1. This allows the abnormal sound detection system 1 to accurately determine whether an abnormal sound has occurred.

[0046] According to the abnormal sound determination system 1, the abnormal sound level can be determined. More specifically, the abnormal sound level depends on the abnormal sound frequency and the specific loudness signal when the abnormal sound occurs. For example, even if the specific loudness signal is large, if the abnormal sound frequency is low, the abnormal sound level is low. Similarly, even if the abnormal sound frequency is high, if the specific loudness signal is small, the abnormal sound level is low. On the other hand, if the specific loudness signal is large and the abnormal sound frequency is high, the abnormal sound level is high. Also, if the specific loudness signal is large and the abnormal sound frequency is low, the abnormal sound level is medium. Also, if the specific loudness signal is small and the abnormal sound frequency is high, the abnormal sound level is medium. Therefore, the processor 10 determines the abnormal sound level indicating the degree of the abnormal sound based on the abnormal sound frequency and the specific loudness signal when the abnormal sound occurs. As a result, according to the abnormal sound determination system 1, the abnormal sound level can be determined.

[0047] The abnormal sound detection system 1 can accurately detect the occurrence of an abnormal sound. More specifically, when one microphone generates a sound pressure signal, if the position of the microphone changes, there are cases where the processor can detect the abnormal sound and cases where the processor cannot detect the abnormal sound. This is because the sound generated by the internal combustion engine 100 interferes.

[0048] Therefore, the abnormal sound detection system 1 is equipped with microphones 12a to 12i. As a result, as shown in Fig. 3, the microphones 12a to 12i are arranged at different spatial positions and capture sounds continuously generated by the internal combustion engine 100. As a result, the abnormal sound detection system 1 can accurately determine the occurrence of an abnormal sound.

[0049] Furthermore, if the internal combustion engine 100 is a multiple cylinder engine, the position at which the abnormal sound is generated varies depending on the cylinder in which the abnormal sound occurs. Therefore, the abnormal sound detection system 1 is equipped with microphones 12a to 12i. This allows the abnormal sound detection systems 1, 1a to identify the cylinder in which the abnormal sound is generated. At this time, a method such as sound source separation may be used to estimate the position at which the abnormal sound is generated.

[0050] (Modification) Next, an abnormal sound detection system 1a according to a modified example will be described with reference to the drawings. Fig. 7 is a flowchart showing a subroutine of step S4. Fig. 8 is a diagram showing the waveform of the specific loudness signal Sig11a and the waveform of the specific loudness differential value Sig12a. Fig. 9 is a diagram showing the calculation procedure of the predetermined % time rate level Y. A block diagram of the abnormal sound detection system 1a is shown in Fig. 1, similar to the abnormal sound detection system 1.

[0051] The abnormal sound detection system 1a differs from the abnormal sound detection system 1 in step S4. Step S4 will now be described in detail with reference to FIG.

[0052] 8, the processor 10 detects a first zero-crossing point ZC1 where the specific loudness derivative value Sig12a changes from negative to positive (first zero-crossing detection step, step S21). Furthermore, the processor 10 detects a second zero-crossing point ZC2 where the specific loudness derivative value Sig12a changes from positive to negative for the first time after the first zero-crossing point ZC1 (second zero-crossing detection step, step S22).

[0053] Next, the processor 10 multiplies the maximum value M1 of the specific loudness differential value Sig12a between the first zero cross point ZC1 and the second zero cross point ZC2 by the maximum value M2 of the specific loudness signal Sig11a between the first zero cross point ZC1 and the second zero cross point ZC2 to calculate a judgment value XN (judgment value calculation step, step S23). N is a natural number. The specific loudness differential value Sig12a includes multiple peak waveforms. Therefore, the first judgment value is called a judgment value X1. The Nth judgment value is called a judgment value XN.

[0054] Next, the processor 10 judges whether or not the processing of steps S21 to S23 has been performed on the specific loudness signal Sig11a and the specific loudness differential value Sig12a over the judgment period (step S24). If the processing of steps S21 to S23 over the judgment period has not been completed, the process returns to step S21. If the processing of steps S21 to S23 over the judgment period has been completed, the process proceeds to step S25. In this case, since the specific loudness differential value Sig12a has multiple (M) peak waveforms in the judgment period, the processor 10 obtains judgment values ​​X1 to XM, where M is a natural number. Thus, in the judgment step, the judgment values ​​X1 to XM are calculated based on the specific loudness differential value Sig12a. As described above, the judgment step includes a first zero-cross detection step, a second zero-cross detection step, and a judgment value calculation step, which are repeated in the measurement period.

[0055] When the processing of steps S21 to S23 over the judgment period is completed, the processor 10 calculates a predetermined percentage level Y based on the judgment values ​​X1 to XM obtained in the judgment value calculation step (percentage level calculation step: step S25). The percentage level calculation step will be described below with reference to FIG. 9.

[0056] There are judgment values ​​X1 to XM. First, the processor 10 sorts the judgment values ​​X1 to XM from smallest to largest.

[0057] The processor 10 calculates a predetermined % time rate level Y. The predetermined % time rate level Y is a value that is located at a predetermined % of the total data, counting from the larger numerical value. The predetermined % is a value that is determined in advance. In this embodiment, the processor 10 specifies the judgment value X3 as the predetermined % time rate level Y.

[0058] Next, the processor 10 calculates the third threshold TH3 by using the predetermined percentage level Y and the predetermined coefficient K. In the present embodiment, the processor 10 multiplies the predetermined percentage level Y by the predetermined coefficient K to calculate the third threshold TH3 (threshold calculation step: step S26).

[0059] Next, the processor 10 sets N to 1 (step S27). The processor 10 determines whether the judgment value XN is greater than the third threshold TH3 (step S28). If the judgment value XN is greater than the third threshold TH3, the process proceeds to step S29. If the judgment value XN is not greater than the third threshold TH3, the process proceeds to step S30.

[0060] When the determination value XN is greater than the third threshold value TH3, the processor 10 determines that an abnormal sound has occurred (step S29). After that, the process proceeds to step S31.

[0061] When the determination value XN is not greater than the third threshold value TH3, the processor 10 determines that no abnormal sound has occurred (step S30). After that, the process proceeds to step S31.

[0062] In step S31, the processor 10 determines whether or not N=M holds (step S31). In step S31, the processor 10 determines whether or not a determination as to the occurrence of abnormal sound has been made for all of the judgment values ​​X1 to XM. If N=M does not hold, the process proceeds to step S32. If N=M holds, the process proceeds to step S33. In this manner, the processor 10 determines that an abnormal sound has occurred for a judgment value that is greater than the third threshold value TH3 among the multiple judgment values ​​X1 to XM.

[0063] If N=M is ​​not satisfied, the processor 10 increments N by 1 (step S32). After that, the process returns to step S28.

[0064] If N=M, the processor 10 judges whether or not the processing of steps S21 to S31 has been performed for all the specific loudness signals Sig11a to Sig11i and all the specific loudness differential values ​​Sig12a to Sig12i (step S33). If the processing of steps S21 to S31 has been performed for all the specific loudness signals Sig11a to Sig11i and all the specific loudness differential values ​​Sig12a to Sig12i, the process proceeds to step S34. If the processing of steps S21 to S31 has not been performed for all the specific loudness signals Sig11a to Sig11i and all the specific loudness differential values ​​Sig12a to Sig12i, the process returns to step S21. In this case, the processor 10 performs the processing of steps S21 to S31 for the next specific loudness signal and specific loudness differential value.

[0065] When the processes of steps S21 to S31 have been performed for all the specific loudness signals Sig11a to Sig11i and all the specific loudness differential values ​​Sig12a to Sig12i, the processor 10 calculates an abnormal sound frequency indicating the number of times an abnormal sound has occurred per unit time (frequency calculation step: step S34). Specifically, the processor 10 calculates the abnormal sound frequency by dividing the number of times that it has been determined in steps S21 to S33 that an abnormal sound has occurred by the determination period.

[0066] Next, the processor 10 determines the abnormal sound level indicating the degree of the abnormal sound based on the abnormal sound frequency and the specific loudness signal when the abnormal sound occurs (level determination step: step S35). Step S35 is the same as step S18, so the description will be omitted. After this, the process ends.

[0067] [effect] According to the abnormal sound detection system 1a, it is possible to determine whether or not an abnormal sound has occurred for the same reasons as the abnormal sound detection system 1. According to the abnormal sound detection system 1a, it is possible to accurately determine whether or not an abnormal sound has occurred for the same reasons as the abnormal sound detection system 1.

[0068] According to the abnormal sound detection system 1a, the occurrence of an abnormal sound can be detected accurately for the following reasons. More specifically, the processor 10 calculates the determination value XN using the maximum value M1 of the specific loudness differential value Sig12a between the first zero cross point ZC1 and the second zero cross point ZC2 and the maximum value M2 of the specific loudness signal Sig11a between the first zero cross point ZC1 and the second zero cross point ZC2. The processor 10 calculates a predetermined percentage level Y based on the determination values ​​X1 to XM obtained in the determination value calculation step. Then, the processor 10 detects that an abnormal sound has occurred for a determination value that is greater than the third threshold value TH3 among the multiple determination values ​​X1 to XM. In this way, the processor 10 uses the first zero cross point ZC1 and the second zero cross point ZC2 of the specific loudness differential value Sig12a to perform the determination regarding the occurrence of an abnormal sound. Therefore, the processor 10 can accurately count the occurrence frequency of the abnormal sound.

[0069] Furthermore, in the abnormal sound detection system 1a, a large amount of experimental data is not required to determine the predetermined % time rate level Y. More specifically, in the abnormal sound detection system 1, the first threshold TH1 and the second threshold TH2 are used to determine the occurrence of an abnormal sound. To determine the first threshold TH1 and the second threshold TH2, a large amount of experimental data is required. On the other hand, in the abnormal sound detection system 1a, the predetermined % time rate level Y is used to determine the occurrence of an abnormal sound. The predetermined % time rate level Y is obtained from the specific loudness signals Sig11a-Sig11i and the specific loudness differential values ​​Sig12a-Sig12i. That is, the predetermined % time rate level Y can be determined based on the specific loudness signals Sig11a-Sig11i and the specific loudness differential values ​​Sig12a-Sig12i acquired during the operation of the abnormal sound detection system 1a. Therefore, a large amount of experimental data is not required to determine the predetermined % time rate level Y. As described above, in the abnormal sound detection system 1a, the determination of the predetermined % time rate level Y does not require a large amount of experimental data.

[0070] (Other embodiments) The abnormal sound detection program, abnormal sound detection device, abnormal sound detection system, and abnormal sound detection method according to the present invention are not limited to the abnormal sound detection program, abnormal sound detection device, abnormal sound detection system 1, 1a, and abnormal sound detection method according to the embodiments and modifications, and may be modified within the scope of the gist thereof. Note that the configuration of abnormal sound detection system 1 and the configuration of abnormal sound detection system 1a may be combined in any manner.

[0071] The specific frequency band is not limited to 10 kHz. The specific frequency band depends on the structure of the internal combustion engine 100. For example, the specific frequency band becomes lower as the displacement of the internal combustion engine 100 increases.

[0072] In the abnormal sound determination system 1, 1a, the processor 10 determines whether an abnormal sound has occurred in the internal combustion engine 100 based on the specific loudness signals Sig11a-Sig11i and the specific loudness differential values ​​Sig12a-Sig12i. However, the processor 10 may also determine whether an abnormal sound has occurred in the internal combustion engine 100 based on the specific loudness differential values ​​Sig12a-Sig12i.

[0073] The abnormal sound detection system 1, 1a includes the microphones 12a to 12i. However, the abnormal sound detection system 1, 1a may include one microphone instead of the microphones 12a to 12i.

[0074] The processor 10 may determine that an abnormal sound has occurred for a determination value among the plurality of determination values ​​X1 to XM that is equal to or greater than a third threshold value TH3.

[0075] In the determination step, it is sufficient to calculate the plurality of determination values ​​X1 to XM based on the specific loudness differential value Sig12a, and the calculation method is not limited to the one shown in steps S21 to S23.

[0076] In addition, in the judgment step, a threshold value for the specific loudness differential value Sig12a, a threshold value for the specific loudness signal Sig11a, and a threshold value for the abnormal sound frequency may be set, and judgment may be made based on the specific loudness differential value Sig12a, the specific loudness signal Sig11a, the abnormal sound frequency, and these threshold values. [Explanation of symbols]

[0077] 1,1a: Abnormal sound detection system 10: Processor 12a~12i: Microphones 14: Amplifier 16:A / D 18: Storage device 20:Display device 100: Internal combustion engine K: Predetermined coefficient KL0~KL2: Abnormal sound level M1, M2: Maximum value Sig10a~Sig10i: Loudness signals Sig11a~Sig11i: Specific loudness signals Sig12a~Sig12i: Specific loudness differential value Sig1a~Sig1i: Sound pressure signal TH1: First threshold TH2: Second threshold TH3: Third threshold X1~XM,XN: Judgment value Y: Time rate level ZC1: First zero crossing point ZC2: Second zero crossing point

Claims

1. a generating step of generating a specific loudness differential value by differentiating with respect to time a specific loudness signal obtained by calculating a specific loudness corresponding to a sound continuously generated by the internal combustion engine; a determination step of determining whether an abnormal sound occurs in the internal combustion engine based on the specific loudness differential value; causing the control device to execute Abnormal sound detection program.

2. In the determination step, it is determined that an abnormal sound has occurred when the loudness differential value becomes larger than a first threshold value, causing the specific loudness signal to become larger than a second threshold value. The abnormal sound detection program according to claim 1 .

3. The determining step includes: a frequency calculation step of calculating an abnormal sound frequency indicating the number of times the abnormal sound occurs per unit time; a level determination step of determining an abnormal sound level indicating the degree of the abnormal sound based on the abnormal sound frequency and the specific loudness signal at the time when the abnormal sound occurs; Contains The abnormal sound detection program according to claim 2 .

4. the determining step includes a first zero-crossing detection step, a second zero-crossing detection step, and a determination value calculation step, which are repeated during a measurement period; In the first zero-cross detection step, a first zero-cross point where the specific loudness differential value changes from negative to positive is detected, In the second zero crossing detection step, a second zero crossing point at which the specific loudness differential value changes from positive to negative for the first time after the first zero crossing point is detected, In the determination value calculation step, a determination value is calculated using a maximum value of the specific loudness differential value between the first zero cross point and the second zero cross point and a maximum value of the specific loudness signal between the first zero cross point and the second zero cross point. The abnormal sound detection program according to claim 1 .

5. The determination step further includes a determination value calculation step, a time rate level calculation step, and a threshold value calculation step, In the judgment value calculation step, a plurality of judgment values ​​are calculated based on the specific loudness differential value, In the time rate level calculation step, a predetermined % time rate level is calculated based on the plurality of judgment values ​​obtained in the judgment value calculation step, In the threshold calculation step, a third threshold is calculated using the predetermined percentage time level and a predetermined coefficient; In the determination step, it is determined that an abnormal sound has occurred for a determination value that is greater than the third threshold value among the plurality of determination values. The abnormal sound detection program according to claim 1 .

6. An abnormal sound detection device, a generating step of generating a specific loudness differential value by differentiating with respect to time a specific loudness signal obtained by calculating a specific loudness corresponding to a sound continuously generated by the internal combustion engine; a determination step of determining whether an abnormal sound occurs in the internal combustion engine based on the specific loudness differential value; Execute Abnormal sound determination device.

7. The abnormal sound detection device according to claim 6 , A plurality of microphones; It is equipped with Each of the plurality of microphones is arranged at a different spatial position and captures sounds continuously generated by the internal combustion engine. Abnormal sound detection system.

8. The plurality of microphones acquire sounds successively generated by different parts of the internal combustion engine. The abnormal sound detection system according to claim 7 .

9. a generating step of generating a specific loudness differential value by differentiating with respect to time a specific loudness signal obtained by calculating a specific loudness corresponding to a sound continuously generated by the internal combustion engine; a determination step of determining whether an abnormal sound occurs in the internal combustion engine based on the specific loudness differential value; causing the control device to execute Abnormal sound determination method.

Images