JP3915684B2 - Vibration wave determination device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a vibration wave determination device that determines a state of an object, for example, a fitting state, based on a vibration wave generated during the operation of the object such as an assembly operation.
[0002]
[Prior art]
Conventionally, for example, as described in Patent Document 1, a target is tapped with a hammer or the like, and a sound or vibration generated at that time is detected. Based on the determination of the vibration wave, the content of the target is determined. A technique for inspecting a crack of an object is known. In this specification, sound and vibration are collectively referred to as vibration waves.
[0003]
[Patent Document 1]
Japanese Patent Laid-Open No. 10-300730
[Problems to be solved by the invention]
By the way, if the object has the same shape and the same properties, the generated vibration waves have substantially the same characteristics, but for example, in the case of a resin molded product, the shape may be slightly different depending on the mold, and in that case It has been found that, for example, the level of vibration waves generated during the operation of an object such as an assembling operation changes and affects the determination accuracy of the content of the object.
[0005]
The present invention has been made in view of the above points, and even when a plurality of different objects are determined, it is possible to accurately determine the state of the object based on vibration waves generated during operation of the object. An object of the present invention is to provide a simple vibration wave determination device.
[0006]
[Means for Solving the Problems]
In order to achieve the above object, the technical means described in claims 1 to 6 are employed.
[0007]
According to the first aspect of the present invention, the vibration wave input means for detecting and inputting the vibration wave generated simultaneously or continuously at a plurality of locations at the time of operation of the object, and the frequency of the input vibration wave are separated, and the object A signal processing means for extracting a first time-series signal for each predetermined frequency band in which a target operation sound indicating a predetermined operation state of the object is generated; and a first time-series set for each of the plurality of different objects A plurality of individual correction tables in which correction coefficients for correcting the signal level are set, a correction table means capable of selecting the individual correction table in accordance with a selection command signal corresponding to the object, and the selected individual correction table Based on the correction coefficient, correction means for correcting the level of the first time series signal for each predetermined frequency band to generate a second time series signal, and second time for each frequency band generated by the correction means Series From by comparing the sound pressure signal level and obtains the area equivalent value by product of the generation period, the total value of the area equivalent value in a plurality of frequency bands with a predetermined threshold, determination determines the state of the object Means.
[0008]
Accordingly, even when a plurality of different objects are determined, the correction state corresponding to the different objects is selected and the level of the first time series signal is corrected, so that the operating state of each object is accurately determined. It becomes possible.
[0009]
According to the second aspect of the present invention, the signal processing means includes the wavelet transform calculating means for frequency-separating the input vibration wave to form a first time-series signal for each predetermined frequency band. It becomes possible to effectively frequency-separate the first time-series signal in a predetermined frequency band.
[0010]
According to the third aspect of the present invention, the signal processing means includes a wavelet transform operation means for frequency-separating the input vibration wave to form a time-series signal for each predetermined frequency band, and the time-series signal. And a filter means for taking out a first time-series signal for each predetermined frequency band in which a target operation sound indicating a predetermined operation state of the object is generated, so that a first time-series signal in a predetermined frequency band is obtained. It is possible to effectively separate the frequencies and extract a signal in a required frequency band.
[0011]
According to the fourth aspect of the present invention, an input device that can be selected and operated from the outside is provided, and the correction table means is determined in advance from an individual correction table selected according to a selection command signal instructed by the input device. By reading out the correction coefficient set for each frequency band and giving it to the correction means, it is possible to select the correction coefficient continuously according to the selection command from the outside, and it is possible to continuously determine different objects It becomes.
[0012]
According to the fifth aspect of the present invention, if the determination means determines that the object is in the predetermined operating state, is the second time-series signal level for each predetermined frequency band within the reference range? Based on whether or not the correction coefficient correction means for correcting the correction coefficient in the individual correction table is provided, the state of the object changes little by little over time, and the initially assumed distribution of variation changes However, the correction coefficient can be corrected by detecting such a change from the level change of the second time-series signal, and the determination accuracy can be stably maintained.
[0013]
According to the sixth aspect of the present invention, the correction coefficient correcting means reduces the correction coefficient when the level of the second time series signal is greater than the reference range, while the level of the second time series signal is within the reference range. When the value is smaller, the correction coefficient can be set to a correction coefficient corresponding to the change of the object by increasing the correction coefficient.
[0014]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described with reference to the drawings.
[0015]
In the following description, an example will be described in which an assembly (fitting) failure of a component is determined by detecting an assembly sound (fitting sound) generated during assembly of the component (or product) as an object. However, the vibration wave determination device 100 of the present invention is not limited to this application, and can be applied to other applications as long as the state of the object reflects the vibration wave. It can also be applied to detection of vibration instead of sound.
[0016]
(First embodiment)
FIG. 1 shows a system configuration in the first embodiment of the present invention.
[0017]
Here, as an object 1, in this example, automatic assembly of resin molded parts, particularly as an example in which a loud noise is generated during assembly, a snap fit mechanism as shown in FIGS. An example of sound generation from the work process for assembling both parts 110 and 120 is given. FIG. 2 shows the projections 104 provided in the plurality of fitting portions 103 in the bottomed cylindrical lower lid component 120 in the holes 102 opened in the plurality of fitting portions 101 in the bottomed cylindrical upper lid component 110. The state to make it fit is shown. In this case, the lower lid part 120 is inserted into the upper lid part 110, and the protrusions 104 are fitted into the hole parts 102 almost simultaneously and are fixed by the snap-fit mechanism.
[0018]
FIGS. 3A, 3B, and 3C are examples of vibration waves generated when the two parts 110 and 120 are assembled. FIG. 3A shows the assembly operation state, and FIG. When it is good, (c) shows the vibration wave at the time of poor assembly. As can be seen from the assembled sound level in FIGS. 3B and 3C, the sound pressure level of the fitting sound changes according to the quality of the assembly.
[0019]
As a feature of sound generation at the time of assembly, a mechanical operation sound accompanying the movement of the assembly equipment and a fitting sound that conveys the fitting status (that is, a target operation sound to be judged) are generated. It has a sound pressure level corresponding to the magnitude of the assembly speed and the like, and the sound pressure levels of both sounds change in correlation with each other.
[0020]
The microphone 2 detects the vibration generated in the object 1 for vibration wave determination as a sound wave and converts it into an electric signal. The sound pressure electrical signal input from the microphone 2 is input to the amplifier 3 of the vibration wave determination device 100 and output to the A / D converter 4. The A / D converter 4 converts the sound pressure signal into a digital signal, which is output to the storage device 5 at the subsequent stage for storage processing.
[0021]
The wavelet transform computing unit 6 constitutes a signal processing means, takes in the digital sound pressure signal S0 stored in the storage device 5 at a predetermined timing, and this digital sound pressure signal S0 is set in advance. Are separated for each frequency band and converted to a time-series signal S1. In general, the wavelet transform computing unit 6 is a computing unit that separates the digital sound pressure signal S0 into time-series signals for each frequency band by enlarging or reducing the basis function (wavelet function). In this example, a frequency band is set as an assembly sound in accordance with a fitting sound that is a target operation sound generated from one or more snap-fit mechanisms. This frequency band is composed of a set band of one or a plurality of frequency bands, depending on the characteristics of the target fitting sound.
[0022]
The filter processor 7 constitutes signal processing means, and reduces or cuts a signal in a frequency band other than the fitting sound from a predetermined time-series signal for each frequency band, and outputs it as a time-series signal S2. Note that the filter processor 7 is omitted, and the time series signal S1 of the wavelet transform arithmetic unit 6 is supplied to the corrector 10 described later, and a correction coefficient for each predetermined frequency band applied to the corrector 10 is devised. By doing so, it is also possible to reduce or cut signals in a frequency band other than the fitting sound.
[0023]
The correction table device 8 constitutes correction table means, and selects a correction coefficient (or correction amount or gain) to be applied to the corrector 10 described later according to the shape and properties of the resin molded part to be assembled and fitted. It comprises a table selection means 81 and a plurality of individual correction tables 82 shown in FIGS. 4 (a) and 4 (b). The input device 9 connected to the table selection means 81 is a device that inputs, for example, the types of resin molding dies of the parts 110 and 120 shown in FIG. 2 as selection commands. It is divided into multiple types according to the degree, size, shape, and material. This is because it has been found that the variation of both the parts 110 and 120 has a distribution that differs mainly for each resin mold.
[0024]
In the correction table device 8, as shown in FIG. 4A, individual correction tables (Da1, Da2,...) 82 are set for only combinations of resin molds used for vibration wave determination. One (for example, Da1) constitutes the individual correction table shown in FIG. In FIG. 4A, A, B, C, and D indicate the type of the component 110, and 1, 2, 3, and 4 indicate the type of the component 120. Therefore, the table selection unit 81 reads out the correction coefficients a11, b11,... Set for each predetermined frequency band from the individual correction table 82 selected according to the selection command signal of the input device 9, and performs a filter process. Every time the time-series signal S2 is input from the adjuster 7, the corrector 10a, 10b,.
[0025]
The corrector 10 is a collection of one or more correctors 10a, 10b,... That normally constitute correction means and receive the fitting sound signal S1. Each of the correctors 10a and 10b receives correction coefficients a11, b11,... For each frequency band determined in advance from the individual correction table 82, weights the time series signal S2 for each frequency band, and performs time series signal S3. Form. This is because the time series signal S2 in the frequency band other than the assumed fitting sound is regarded as noise and the level is lowered or cut, while the frequency band with less noise in the fitting sound is amplified and the S / N ratio is increased. Has improved.
[0026]
The determination device 11 can be fitted from all fitting locations even when each fitting sound (target operation sound) generated from a plurality of fitting locations in the object 1 overlaps or continues on the time axis. It is necessary to determine whether or not a sound is generated. Therefore, the reference value of the sound pressure signal determined by the sound pressure signal level of the fitting operation sound when all the fitting points overlap, or the reference value of the sound pressure signal determined by the sound pressure signal level and the generation period thereof (this Is a value obtained by adding an area equivalent value, which is a product of the sound pressure signal level and the generation period, for each predetermined frequency (so-called sound volume equivalent value), and is set as a threshold value L in the threshold value table 12. In order to determine the fitting state at a plurality of locations, it is possible to improve the determination accuracy by using an area equivalent value considering the occurrence period, more preferably a volume equivalent value considering the fitting sound volume in a plurality of frequency bands.
[0027]
As an example of the determiner 11, when the fitting sound exists in the time series signal S3 portion for each frequency band generated by the corrector 10 or in a plurality of frequency bands, each time series signal S3 portion By determining the total value of the area equivalent values (volume equivalent value) and comparing the value with the threshold value L, it is determined whether the parts are assembled (fitted). Based on the determination result, the determination unit 11 displays the result on the display unit 13, and outputs the alarm to the alarm unit 14 if it fails.
[0028]
Next, the determination flow of the vibration wave determination apparatus 100 configured as described above is summarized as shown in FIG. FIG. 6 is a signal waveform diagram of a vibration wave.
[0029]
When the apparatus 100 is instructed to start the determination, the vibration wave (FIG. 6A) generated from the object 1 is recorded as the digital sound pressure signal S0 by the microphone 2 to the storage device 5 (step 201). The wavelet transform computing unit 6 separates and extracts the digital sound pressure signal S0 into a time-series signal S1 having a frequency having a frequency band (FIG. 6B) that matches the fitting sound that is the target operation sound (step S1). 202). In the example of FIG. 6, the sampling frequency of the digital sound pressure signal S0 is 44 KHz, and among the first frequency bands, the frequency band 11 is 11 to 22 KHz, the frequency band 10 is 5.5 to 11 KHz, and the frequency band 9 is 2.8. The fitting sound is noticeable at ~ 5.5KHz. In addition, the mechanical operating sound appears in frequency bands 8 to 6 below 2.8 KHz, and in this case, it can be seen that separation is possible by distinguishing data for each frequency band. Next, the filter processor 7 reduces or cuts the signal in the frequency band other than the fitting sound from these time-series signals S1, and extracts the fitting sound (step 203). FIG. 6C shows a time series signal (fitting sound signal) S2 obtained by filtering the time series signal S1 in accordance with the fitting sound.
[0030]
Next, in the correction table device 8, the individual correction table 82 is selected according to the selection command signal from the input device 9, and the correction coefficients a11, b11,... For each frequency band are given to the corrector 10 (step 204). Therefore, the corrector 10 corrects the fitting sound level of the time-series signal S2 based on the correction coefficients a11, b11,... (Step 205).
[0031]
Next, the determiner 11 obtains a sound volume equivalent value based on the time-series signal S3 for each frequency band generated by the corrector 10, and compares the value with a threshold value L to determine whether the product is assembled (fitted). Is determined (step 206). If it is greater than or equal to the threshold value L, a pass display (good fitting) is obtained, and if it is less than the threshold value L, a fail display (with poor fitting) and alarm output are performed (steps 207 and 208).
[0032]
Here, how to create each individual correction table 82 of the above-described correction table device 8 will be described with reference to FIGS.
[0033]
7 is a test system for creating the correction coefficients a11, b11,... Of each individual correction table 82. An apparatus 100A similar to the fitting sound determination apparatus 100 described above is prepared, and each individual correction is performed. The table 82 is composed of rewritable nonvolatile storage means such as flash memory or EEPROM. In addition, in order for each combination of the resin molds A to D and 1 to 4 as shown in FIG. The assembly equipment 1 </ b> A is set so that the parts can be automatically replaced so that the parts 110 and 120 are combined, and the generation order of selection command signals generated from the input device 9 is set.
[0034]
FIG. 8 shows the processing flow of this test system. First, the parts 110 and 120 are set in the assembly facility 1A, and a selection command signal is input from the input device 9 to select the individual correction table 82 (steps 301 and 302). A vibration wave generated from the object 1 in the assembly facility 1A is input from the microphone 2 and recorded as a digital sound pressure signal (step 303). This sound pressure signal is wavelet transformed (further filtered) to be separated into time-series signals for each predetermined frequency band (step 304), and time-series signals (fitting sound signals) for each predetermined frequency band. ) Is extracted (step 305).
[0035]
Here, correction work of the correction coefficient is performed only when the sound pressure signal level of each time series signal is within the allowable range, thereby preventing abnormal correction work. Therefore, the sound pressure signal level of each time series signal is determined (step 306). If it is within the allowable range, each time series signal is corrected by a corrector based on the initial value of the individual correction cable 82, A corrected time-series signal for each predetermined frequency band is formed (step 307).
[0036]
In order to obtain the optimum sound pressure signal magnitude for determination by the determination device in the subsequent stage based on each corrected time series signal, each correction for the time series signal in each frequency band f1, f2,. The coefficient is incremented or decremented by a predetermined value from the initial value, corrected to an optimum correction coefficient, and stored (steps 308 to 310). The reference range is the sound pressure for each corrected time series signal so that the optimum sound pressure signal size can be obtained based on each corrected time series signal. It means the range of signal level.
[0037]
(Second Embodiment)
In this example, during the vibration wave determination process, the correction coefficients a11, b11,... For each frequency band preset in the individual correction table 82 are set in accordance with the levels of the respective correction time series signals from the corrector 10. This is an example in which a so-called learning function is added to enable correction.
[0038]
The correction coefficients a11, b11,... Of each individual correction table 82 are set in advance by obtaining a process as shown in FIGS. 7 and 8, but slight wear and deformation of the resin mold during long-term flow. Due to this, the shape, surface state, etc. of the resin-molded parts 110, 120 may change, and the distribution of variations originally assumed may change. As a result, the sound pressure signal level of the fitting sound generated when fitting the parts 110 and 120 also changes, which may affect the determination accuracy of the determiner 11. Therefore, in this example, such a change is detected to correct the correction coefficients a11, b11,... And will be described with reference to FIGS.
[0039]
FIG. 9 shows a system configuration in the second embodiment. FIG. 9 is different from FIG. 1 in that a correction coefficient correction unit 15 is provided and a rewritable nonvolatile storage unit such as a flash memory or an EEPROM is used as the individual correction table 82. The point to do is different. FIG. 10 shows a determination flow of the vibration wave determination apparatus 100, which is different from FIG. 5 in that a correction coefficient correction flow 250 is added.
[0040]
In this example, when the pass determination is made at step 207, the sound pressure signal level of each corrected time series signal S3 for each predetermined frequency band from the corrector 10 is checked, and each sound pressure signal level is determined in advance. In order to be within the reference range provided for each frequency band, the set correction coefficient value is increased or decreased by a predetermined value, corrected to an optimal correction coefficient, and updated and stored in the corresponding individual correction table 82 ( Steps 209-211).
[0041]
Note that the reference range in step 209 is the same as in FIG. 8, and the magnitude of the sound pressure signal optimal for determination by the determination unit 11 at the subsequent stage is based on each correction time series signal from the correction unit 10. It means the range of the sound pressure signal level for each corrected time series signal to be obtained.
[Brief description of the drawings]
FIG. 1 is a configuration diagram showing a system configuration of a first embodiment of the present invention.
FIG. 2 is an explanatory view showing a part of the part assembling step of FIG. 1;
3A is an explanatory diagram showing a main part of the part assembly process of FIG. 2, and FIGS. 3B and 3C are diagrams showing sound pressure waveforms detected in the part assembly process. FIG.
4A is a diagram illustrating the entire correction table of FIG. 1, and FIG. 4B is a diagram illustrating an individual correction table 82 of FIG.
FIG. 5 is a flowchart showing a processing flow of the vibration wave determination apparatus 100 shown in FIG. 1;
6 is a signal waveform diagram of the vibration wave determination apparatus 100 shown in FIG. 1; FIG.
FIG. 7 is a block diagram showing a test system for setting a correction coefficient of a correction table.
8 is a flowchart showing a processing flow of the test system shown in FIG.
FIG. 9 is a configuration diagram showing a system configuration of a second exemplary embodiment of the present invention.
10 is a flowchart showing a processing flow of the vibration wave determination apparatus 100 shown in FIG. 9;
[Explanation of symbols]
1 Object 2 Microphone (vibration wave input means)
5 Storage device 6 Wavelet transform computing unit (signal processing means)
7 Filter processor (Signal processing means)
8 Correction table device (correction table means)
9 Input device 10 Corrector (correction means)
11 Judgment device (judgment means)
15 Correction coefficient correction means 81 Table selection means 82 Individual correction table

Claims (6)

A vibration wave input means for detecting and inputting vibration waves generated simultaneously or continuously at a plurality of locations when the object is operated;
Signal processing means for frequency-separating the input vibration wave to extract a first time-series signal for each predetermined frequency band in which a target operation sound indicating a predetermined operation state of the object is generated;
A plurality of individual correction tables that are set for a plurality of different objects and set a correction coefficient for correcting the level of the first time-series signal, and according to a selection command signal corresponding to the object Correction table means capable of selecting the individual correction table;
Correction means for correcting a level of the first time series signal for each predetermined frequency band based on the selected correction coefficient of the individual correction table to generate a second time series signal;
From the second time-series signal for each frequency band generated by the correction means, an area equivalent value is obtained by the product of the sound pressure signal level and the generation period, and a total value of the area equivalent values in a plurality of frequency bands is predetermined. A vibration wave determination apparatus comprising: a determination unit that determines a state of the object by comparing with a threshold value .   The said signal processing means has a wavelet transformation calculation means which forms the said 1st time series signal for every predetermined frequency band by frequency-separating the said vibration wave input. The vibration wave determination apparatus described.   The signal processing means includes a wavelet transform computing means for frequency-separating the inputted vibration wave to form a time-series signal for each predetermined frequency band, and a predetermined operating state of the object from the time-series signal. The vibration wave determination device according to claim 1, further comprising: a filter unit that extracts the first time-series signal for each predetermined frequency band in which a target operation sound indicating the frequency is generated.   The correction table means is set for each predetermined frequency band from the individual correction table selected according to the selection command signal instructed by the input device. The vibration wave determination apparatus according to claim 1, wherein a correction coefficient is read and provided to the correction unit.   If the determination means determines that the object is in a predetermined operating state, based on whether the level of the second time series signal for each predetermined frequency band is within a reference range, 5. The vibration wave determination apparatus according to claim 1, further comprising correction coefficient correction means for correcting a correction coefficient in the individual correction table.   The correction coefficient correction means reduces the correction coefficient when the level of the second time series signal is greater than the reference range, and corrects when the level of the second time series signal is less than the reference range. 6. The vibration wave determination apparatus according to claim 5, wherein the coefficient is increased.

Images