JP5349980B2 - Signal processing apparatus, signal processing apparatus control method, and signal processing apparatus control program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a tachometer capable of improving following properties of a rotary primary peak and accurately detecting a varying rotational speed with an improved response, to provide a method of controlling the tachometer, and to provide a control program for the tachometer. <P>SOLUTION: A signal processor includes: an FFT (Fast Fourier Transformation) operation means (12a) for performing Fast Fourier Transformation to input signal; a determination means (12b) for determining whether a storage means (16) stores peak spectrum data; a searching means (12d) for searching spectrum data; and a searching means (12e) for searching spectrum data having a frequency nearly n times larger than that of spectrum data and spectrum data having a frequency nearly 1/n times larger than that of the spectrum data (in this case, n is a natural number of not less than 2). <P>COPYRIGHT: (C)2010,JPO&amp;INPIT

Description

  The present invention relates to a signal processing device that measures the rotational speed of a measurement object that rotates, a control method for the signal processing device, and a control program for the signal processing device.

The number of rotations of the measurement object that rotates is measured by attaching a tachometer as a signal processing device such as a rotary encoder to the measurement object. However, since such a tachometer needs to be directly attached to the measurement object, it is necessary to prepare a jig for attaching the tachometer, or to change the characteristics of the rotational movement of the measurement object due to the effect of the attached tachometer. Therefore, there is a problem that the rotational speed of the measurement target cannot be measured accurately.
Therefore, a tachometer capable of measuring without adversely affecting the rotation speed of the measurement object has been proposed and introduced into the general market (for example, see Patent Document 1). The invention of Patent Document 1 detects the leakage magnetic flux of the rotating motor to be measured, obtains the leakage magnetic flux spectrum (analysis data) by fast Fourier transform (FFT), and rotates the frequency indicating the maximum value of the spectrum. The number of rotations is obtained by determining the rotation frequency of the motor.

  In such a tachometer algorithm, for example, a search range is set based on the measurement value obtained in the previous measurement cycle, and the maximum peak in the range is used as the measurement value.

JP 2008-196876 A

  However, in the measurement method using the maximum peak within a certain range as a measurement value, components that are not directly related to the measurement (frequency components other than the primary peak component of the rotation frequency of the rotating body) suddenly appear and are not directly related to this measurement. In some cases, the peak component is larger than the primary peak component of the rotation frequency of the rotating body, and stable rotation speed measurement cannot be realized.

  In particular, stable rotation speed measurement could not be realized during acceleration / deceleration of the rotating body.

  Even in the conventional method of limiting the search range, an algorithm using the maximum peak may cause a measurement error.

  The present invention has been made in consideration of such a conventional technique, and the object of the present invention is to track the rotation primary peak by simultaneously tracking other order peaks in addition to the rotation primary peak. To provide a signal processing device, a signal processing device control method, and a signal processing device control program capable of improving followability and accurately detecting a fluctuating rotational speed with good response.

The present invention solves the above problems by the following means. For ease of understanding, the reference numerals corresponding to the embodiments of the present invention will be described in parentheses, but the present invention is not limited thereto.

(1) FFT (Fast Fourier Transform) calculation means (12a) for calculating spectrum data by performing fast Fourier transform on the input signal, and peak spectrum data as one of the spectrum data is stored in the storage means (16). A judgment means (12b) for judging whether or not the peak spectrum data is stored in the storage means (16) by the judgment means (12b). Search range setting means (12 c ) for setting a nearby frequency as a search range, search means (12d) for searching the spectrum data calculated by the FFT calculation means (12a) in the search range, and the search The spectrum data searched by the means (12d) Search means (12e) for searching for spectrum data having a frequency approximately n times the wave number and spectrum data having a frequency approximately 1 / n times the frequency of the spectrum data, and the frequency of the spectrum data by the search means (12e) When at least one of spectrum data having a frequency of approximately n times the frequency and spectrum data having a frequency of approximately 1 / n is retrieved, the spectrum data is determined as the peak spectrum data, and the peak is stored in the storage means. And a signal processing device (where n is a natural number of 2 or more).

  According to the invention according to (1), by having the above-described configuration, a frequency near the previously measured peak spectrum data is set as a search range based on the previously measured peak spectrum data, and the search is performed. Spectrum data having a frequency of approximately n times the frequency of the spectrum data and spectrum data having a frequency of approximately 1 / n times the frequency of the spectrum data, and a spectrum having a frequency of approximately n times the frequency of the spectrum data. When at least one of the data and spectrum data having a frequency approximately 1 / n times is searched, the spectrum data is determined as peak spectrum data (where n is a natural number of 2 or more).

  As described above, according to the invention according to (1), the spectrum data stored in the storage unit 16 is not the spectrum data in which only the main component (rotation primary component) is searched but added to the peak of the rotation primary component. Therefore, it is possible to improve the followability of the peak of the rotation primary component (no longer follow the peak due to noise).

  That is, not only the peak due to the main component (rotation primary component) but also other order peaks are simultaneously tracked, noise is removed, and more stable rotation speed measurement can be realized even during acceleration / deceleration.

  (2) When there are a plurality of spectrum data to be searched by the search means (12d), the search means (12e) determines the frequency of the spectrum data from the spectrum data having the largest spectrum data size. The signal processing device according to (1), wherein spectrum data having a frequency of n times and spectrum data having a frequency of approximately 1 / n times are searched (where n is a natural number of 2 or more).

  According to the invention according to (2), when there are a plurality of searched spectrum data, the frequency of the spectrum data is determined from the spectrum data having the largest size of the spectrum data (most likely peak spectrum data). Since the spectrum data of the frequency of about n times the frequency and the spectrum data of the frequency of about 1 / n times are searched, it is possible to measure the rotation speed of the tachometer quickly and in a short time at least with the amount of calculation (here N is a natural number of 2 or more).

  (3) There are a plurality of spectrum data in which at least one of spectrum data having a frequency approximately n times the frequency of the spectrum data and spectrum data having a frequency approximately 1 / n times is searched in the search means (12e). When there is the spectrum data in which both the spectrum data having a frequency approximately n times the frequency of the spectrum data and the spectrum data having a frequency approximately 1 / n times are present, the frequency of the spectrum data (1) or (2), wherein the spectrum data in which both the spectrum data of the n-fold frequency and the spectrum data of the approximately 1 / n-fold frequency are searched is determined as the peak spectrum data. Signal processing device (where n is a natural number of 2 or more).

  According to the invention according to (3), both spectrum data having a frequency approximately n times the frequency of spectrum data and spectrum data having a frequency approximately 1 / n times (where n is a natural number of 2 or more) are searched. Therefore, it is possible to continuously measure the accurate rotation speed of the tachometer stably and more reliably.

  (4) A plurality of spectrum data in which both spectrum data having a frequency approximately n times the frequency of the spectrum data and spectrum data having a frequency approximately 1 / n times (where n is a natural number of 2 or more) are searched. If present, the signal processing apparatus according to any one of (1) to (3), wherein the spectrum data having the largest spectrum data size is determined as the peak spectrum data.

  According to the invention according to (4), the spectrum data in which both spectrum data having a frequency approximately n times the frequency of the spectrum data and approximately 1 / n times the frequency (where n is a natural number of 2 or more) is searched. In the case where there are a plurality, since the spectrum data having the largest spectrum data size is determined as the peak spectrum data, it is easy to determine which spectrum data should be determined as the rotation speed of the tachometer, It becomes possible to prevent other different spectrum data from being erroneously determined as the rotation speed of the tachometer.

  In addition, since the noise component is efficiently removed, it is possible to continuously and more reliably measure the accurate rotational speed of the tachometer.

  (5) The signal is a detection result of a time-series physical phenomenon caused by the rotational motion measured from the measurement target, and the frequency of the peak spectrum data determined by the calculation means (12f) is the measurement target The signal processing device according to any one of (1) to (4), wherein

  According to the invention according to (5), since the signal processing device uses the detection result of the time-series physical phenomenon caused by the rotational motion measured from the measurement object as a signal, it easily results from the rotation of the measurement object. The rotational speed of the measurement object can be obtained from the physical phenomenon.

  (6) The search range setting means (12c) determines the upper limit of the search range at a frequency smaller than n times the frequency of the peak spectrum data, and sets the lower limit of the search range as the frequency of the peak spectrum data. The signal processing device according to any one of (1) to (5), wherein n is a natural number equal to or greater than 2.

  According to the invention according to (6), the upper limit of the search range is calculated as a frequency smaller than n times the frequency of the specific spectrum, and the lower limit of the search range is lower than 1 / n times the frequency of the specific spectrum. Since the calculation is performed as a large frequency, the 1 / n-order component and the n-order component of a specific spectrum can be excluded from the search range, and a search for a different spectrum by mistake can be prevented (where n is 2 or more natural number).

  According to the invention according to (7), it is possible not only to exclude a spectrum such as a 1/2 order component or a secondary component of a frequency of a specific spectrum, but also to set a more appropriate frequency range as a search range. Become.

  (8) FFT calculation step for calculating spectrum data by performing fast Fourier transform on the input signal, and a determination step for determining whether peak spectrum data which is one of the spectrum data in the storage step is stored. In the determination step, when it is determined in the determination step that the peak spectrum data in the storage step is stored, a search range setting step for setting a frequency near the peak spectrum data in a search range, and in the search range, A search step for searching for the spectrum data calculated in the FFT calculation step, spectrum data having a frequency approximately n times the frequency of the spectrum data searched in the search step, and approximately 1 / n of the frequency of the spectrum data Search process to search for spectrum data of double frequency and previous In the search step, when at least one of spectrum data having a frequency of approximately n times the frequency of the spectrum data and spectrum data having a frequency of approximately 1 / n is searched, the spectrum data is converted into the peak spectrum data. And a calculation step of storing the peak spectrum data in the storing step (where n is a natural number of 2 or more).

  According to the invention according to (8), by having the above-described configuration, a frequency near the previously measured peak spectrum data is set as a search range based on the previously measured peak spectrum data, and the search is performed. Spectrum data having a frequency of approximately n times the frequency of the spectrum data and spectrum data having a frequency of approximately 1 / n times the frequency of the spectrum data, and a spectrum having a frequency of approximately n times the frequency of the spectrum data. When at least one of the data and spectrum data having a frequency approximately 1 / n times is searched, the spectrum data is determined as peak spectrum data (where n is a natural number of 2 or more).

  Thus, according to the invention according to (8), the spectrum data stored in the storing step is not the spectrum data in which only the main component (rotation primary component) is searched, but in addition to the peak of the rotation primary component. Since the spectrum data is searched for other order peaks, it is possible to improve the followability of the peak of the rotation primary component (no longer follow the peak due to noise).

  That is, not only the peak due to the main component (rotation primary component) but also other order peaks are simultaneously tracked, noise is removed, and more stable rotation speed measurement can be realized even during acceleration / deceleration.

  (9) A program that causes a computer to execute the method according to (8).

  According to such a configuration, the same effect as in (8) can be expected by causing the computer to execute the program.

  According to the present invention, the spectrum data stored in the storage unit 16 is not the spectrum data in which only the main component (rotation primary component) is searched, but in addition to the peak of the rotation primary component, other order peaks are searched. Since it is spectral data, it is possible to improve the followability of the peak of the rotation primary component (no longer follow the peak due to noise).

  In other words, not only the peak due to the main component (rotation primary component) but also other order peaks are simultaneously tracked, so that noise is removed and more stable and accurate rotation speed measurement can be realized even during acceleration and deceleration.

It is a block diagram which shows the structure of the tachometer as one Example of the signal processing apparatus by this invention. It is a figure which shows an example of the spectrum data processed with the tachometer which concerns on this embodiment. It is a flowchart with which it uses for description about the process sequence by the tachometer which concerns on this embodiment. It is a figure which shows another example of the spectrum data processed with the tachometer which concerns on this embodiment.

In order to provide a signal processing apparatus and a control method for a signal processing apparatus that can accurately read a specific spectrum with a fluctuating frequency, the present invention provides a rotation first-order peak that is a frequency of the fluctuating specific spectrum. In addition, by tracking other order peaks at the same time, the search range for searching for a specific spectrum is followed.

  Hereinafter, the present invention will be described in more detail with reference to the drawings and the like.

  [Entire system configuration]

  FIG. 1 is a block diagram showing the configuration of a tachometer as an embodiment of a signal processing apparatus according to the present invention. FIG. 2 is a diagram showing spectral data processed by the tachometer according to the present embodiment. FIG. 3 is a flowchart for explaining the processing procedure by the tachometer according to the present embodiment.

  A tachometer (signal processing device) 10 includes an input unit 11, an operation control unit 12 (an FFT operation unit as an FFT operation unit, a determination unit as a determination unit, a search range setting unit as a search range setting unit, and a search unit) A search unit, a search unit as a search unit, a calculation unit as a calculation unit), a display unit 13, an operation unit (not shown), an input amplifier unit 14, a filter unit 15 and a storage unit 16, and a gasoline engine It is a measuring instrument that measures the number of rotations of a rotation mechanism provided in the gasoline engine 30 based on vibration data (signal) detected by a vibration meter 20 disposed at 30. In the present embodiment, the frequency f1 of the spectrum of the main component (primary component) based on the vibration data of the gasoline engine 30 detected from the vibration meter 20 and the rotation frequency of the rotation mechanism of the gasoline engine 30 have a relationship of 1: 1. It will be explained as being.

  The input unit 11 is an A / D converter that discretizes time-series vibration data detected by the vibration meter 20 and continuously inputs the vibration data to the operation control unit 12 of the tachometer 10.

  The calculation control unit 12 includes an FFT calculation unit (FFT calculation unit) 12a, a determination unit (determination unit) 12b, a search range setting unit (search range setting unit) 12c, a search unit (search unit) 12d, and a search unit (search unit). 12e and a calculation unit (calculation means) 12f, which is a control circuit that performs overall control of each unit of the tachometer 10. In addition, the measurement time and search range are set by the operation unit (the measurement time and search range can be numerically input, and the measurement time range and search range are set by switching the measurement time and search range. Can also).

  The FFT calculation unit 12a is a calculation circuit that performs fast Fourier transform (FFT) on discrete vibration data continuously input from the input unit 11 and calculates vibration spectrum data.

The determination unit (determination unit) 12b stores peak spectrum data that is one of the spectrum data in the storage unit 16 (spectrum data in a group of spectrum data that has been subjected to FFT calculation at a certain timing in the FFT calculation unit 12a) and has a predetermined condition. It is determined whether or not spectral data satisfying the above (details will be described later ) are stored.

  If the peak spectrum data is stored, the rotation speed of the tachometer is calculated at least once in the calculation unit 12f. In the present embodiment, a case where the rotation number of the tachometer is calculated subsequently when the rotation number of the tachometer has been calculated at least once will be described in detail.

When the determination unit 12b determines that the peak spectrum data is stored in the storage unit 16, the search range setting unit 12c sets a frequency near the peak spectrum data as a search range (in the search range setting unit 12c). , The search range of the spectrum of the principal component to be searched is calculated).

  The spectrum data calculated by the FFT calculation unit 12a includes the spectrum of the principal component (primary component) spectrum, the half component of the principal component, the secondary component of the principal component, and the tertiary component of the principal component. In addition, the spectrum of each order component of noise is also included.

  In the present embodiment, as an example, the search range is automatically set so as to exclude the spectrum of the 1/2 order component and the 2nd order component. Specifically, when the frequency f1 of the spectrum of the main component (primary component) during the initial operation of the gasoline engine 30 is 2 kHz, the spectra of the 1 / 2-order component and the secondary component appear as 1 kHz and 4 kHz, respectively. . Therefore, it is necessary to set the search range that does not include the 1/2 order and secondary components. For example, the lower limit frequency fL is set to 40% reduction of the frequency f1 (0.6 times f1). By setting the upper limit frequency fH to 80% of the frequency f1 (1.8 times f1), the spectrum of the 1/2 order component and the secondary component is excluded from the search range.

  Therefore, the search range setting unit 12c calculates the search range based on the frequency f1 of the principal component spectrum. For example, when the frequency f1 of the principal component spectrum is 2 kHz, the lower limit frequency fL of the search range is 1.2 kHz, and the upper limit frequency fH is 3.6 kHz.

  Furthermore, the search range setting unit 12c may set an arbitrary fixed frequency as the frequency near the peak spectrum data as the search range, and the arbitrary fixed frequency may be set from an operation unit (not shown). It may be entered.

  The search unit 12d is a circuit that sequentially searches spectrum data in the spectrum data calculated by the FFT calculation unit 12a during measurement of the rotation speed of the gasoline engine 30. Here, the spectrum data may be searched from the entire spectrum data, but is preferably searched from the search range calculated by the search range setting unit 12c described above.

  The search unit 12e searches for spectrum data having a frequency that is approximately n times the frequency of the spectrum data searched by the search unit 12d and spectrum data having a frequency that is approximately ½ times the frequency of the spectrum data (where n is 2 or more natural number).

  Here, when the frequency approximately n times the frequency of the spectrum data searched by the search unit 12d and the frequency approximately 1 / n times the frequency of the spectrum data are out of the search range calculated by the search range setting unit 12c. Is the spectrum data (n is a natural number of 2 or more) having a frequency approximately n times and a frequency approximately 1 / n times from a frequency range outside the search range calculated by the search range setting unit 12c. Searched by.

  That is, the range of the frequency that is approximately n times the frequency of the spectrum data searched by the search unit 12d and the frequency that is approximately 1 / n times the frequency of the spectrum data (where n is a natural number of 2 or more) is a search range setting. The search range calculated by the part 12c may be exceeded.

  In addition, when there are a plurality of spectrum data searched by the search unit 12d, the search unit 12e is substantially the same as the spectrum data having a frequency that is approximately n times the frequency of the spectrum data from the spectrum data having the largest spectrum data size. Search for spectrum data having a frequency of 1 / n times (where n is a natural number of 2 or more).

  Further, spectrum data having a frequency approximately n times the frequency of the spectrum data searched by the search unit 12d and / or spectrum data having a frequency approximately 1 / n times (where n is a natural number of 2 or more). May not exist. In such a case, the search unit 12e performs subsequent processing.

  The calculation unit 12f uses the search unit 12e to obtain at least one of spectrum data having a frequency approximately n times the frequency of spectrum data and spectrum data having a frequency approximately 1 / n times (where n is a natural number of 2 or more). When the search is performed, the spectrum data searched for at least one of them is determined as the peak spectrum data, and the peak spectrum data is stored in the storage unit 16.

  The calculation unit 12f includes a plurality of spectrum data in which at least one of spectrum data having a frequency approximately n times the frequency of spectrum data and spectrum data having a frequency approximately 1 / n times is searched in the search unit 12e. When there is the spectrum data in which both the spectrum data having a frequency of approximately n times the frequency of the spectrum data and the spectrum data having a frequency of approximately 1 / n times are present, it is approximately n times the frequency of the spectrum data. The spectrum data in which both the spectrum data of the frequency and the spectrum data of the frequency approximately 1 / n times (where n is a natural number of 2 or more) are searched is determined as the peak spectrum data.

  In addition, the calculation unit 12f in the search unit 12e is at least one of spectrum data having a frequency approximately n times the frequency of spectrum data and spectrum data having a frequency approximately 1 / n times (where n is a natural number of 2 or more). In the case where there are a plurality of spectrum data searched for only the spectrum data, the spectrum data having the largest spectrum data size may be determined as the peak spectrum data.

  Further, when there are a plurality of spectrum data in which both spectrum data having a frequency approximately n times the frequency of spectrum data and spectrum data having a frequency approximately 1 / n times (where n is a natural number of 2 or more) are searched. Alternatively, the spectrum data having the largest spectrum data size may be determined as the peak spectrum data.

  The display unit 13 displays the peak spectrum data calculated by the calculation unit 12f. The peak spectrum data displayed on the display unit 13 indicates the rotation frequency of the rotation mechanism of the gasoline engine 30.

  In addition, the display unit 13 may display a symbol such as a number, a character, or the like and a search range input by the operation unit.

  The input amplifier unit 14 is an amplifier that amplifies vibration data detected from the vibrometer 20.

  The filter unit 15 is a filter including an AAF (anti-aliasing filter) that removes aliasing noise included in the vibration data amplified by the input amplifier unit 14.

  The storage unit 16 stores at least one of spectrum data having a frequency approximately n times the frequency of spectrum data and spectrum data having a frequency approximately 1 / n times (where n is a natural number of 2 or more) by the search unit 12e. The spectrum data searched for one is stored.

  The vibration meter 20 is an accelerometer that is fixed to the casing of the gasoline engine 30 and detects vibrations of a rotation mechanism of the gasoline engine 30 as vibration data of acceleration.

  The gasoline engine 30 has a rotation mechanism inside, and sucks a mixture of fuel (gasoline) and air into a cylinder, compresses the mixture with a piston, and then ignites, burns and expands (premixed combustion) piston. Is an internal combustion engine that reciprocates.

  As described above, according to the present invention, the spectrum data stored in the storage unit 16 is not the spectrum data in which only the main component (rotation primary component) is searched, but other order peaks in addition to the peak of the rotation primary component. Therefore, it is possible to improve the followability of the peak of the rotation primary component (no longer follow the peak due to noise).

  That is, not only the peak due to the main component (rotation primary component) but also other order peaks are simultaneously tracked, noise is removed, and more stable rotation speed measurement can be realized even during acceleration / deceleration.

  Next, an example of the embodiment will be described with reference to FIG.

In FIG. 2A, the measurement peak (spectrum data) f1 of the rotation speed of the rotation mechanism provided in the gasoline engine 30, the spectrum data 2f1 having a frequency approximately twice the frequency of the measurement peak f1, and the frequency of the spectrum data. This shows a state where spectrum data (f1 / 2) having a frequency approximately ½ times the frequency is retrieved.

  2B, in the measurement cycle subsequent to the measurement cycle of FIG. 2A, the FFT operation unit 12a converts the discretized vibration data continuously input from the input unit 11 into a fast Fourier transform (FFT). ), And shows the result of calculating the spectrum data of vibration.

  Rotation components such as a noise component larger than the measurement peak f1 in the frequency region around the measurement peak f1 of the rotation speed of the rotation mechanism provided in the gasoline engine 30 measured in the measurement cycle of FIG. An unrelated frequency component f2 appears.

  In the conventional measurement method, the frequency component f2 irrelevant to the rotation component is erroneously determined to be the rotation speed of the rotation mechanism provided in the gasoline engine 30.

  However, in FIG. 2B, spectrum data having a frequency approximately n times that of the frequency component f2, which is not related to the rotation component, and spectrum data having a frequency approximately 1 / n times the frequency of the spectrum data are included in the FFT operation unit 12a. The spectrum data having a frequency approximately n times the frequency component f2 that is not related to the rotation component and the spectrum data having a frequency approximately 1 / n times the frequency of the spectrum data does not exist (here, N is a natural number of 2 or more).

  Therefore, the frequency component f2 is determined by the calculation unit 12f to be a frequency component unrelated to the rotation component of the rotation speed of the rotation mechanism provided in the gasoline engine 30.

The frequency region around the measurement peak f1 is smaller than the spectrum data of the frequency component f2, but is approximately twice the frequency spectrum data (2 (f1 + Δ)) and approximately ½ times the frequency of the spectrum data. There exists spectral data (f1 + Δ) having spectral data ( (f1 + Δ) / 2 ) of the frequency of.

  Accordingly, the calculation unit 12f determines the spectrum data (f1 + Δ) as a measurement peak (spectrum data) of the rotation speed of the rotation mechanism provided in the gasoline engine 30.

  The spectral data (f1 + Δ) is spectral data in which the spectral data f1 has changed due to factors such as fluctuations in the rotational speed of the rotation mechanism between measurement cycles, but is larger than spectral data (spectral data (f1 + Δ) such as other noise components). Even if it has a value), it is accurately determined (determined) as the rotation speed of the rotation mechanism, and it is possible to follow fluctuations in the rotation speed.

  As described above, according to the present invention, in addition to the peak of the rotation primary component, other order peaks are searched, so that the followability of the peak of the rotation primary component can be improved (the peak due to noise is reduced). No longer follow).

  That is, not only the peak due to the main component (rotation primary component) but also other order peaks are simultaneously tracked, noise is removed, and more stable rotation speed measurement can be realized even during acceleration / deceleration.

  [Processing procedure]

Here, specific processing procedures that may be implemented in the case of applying the present invention will be described with reference to a flowchart shown in FIG. The processing procedure shown below is an example, and there are innumerable processing procedures that can be realized in addition to this.

  In step S <b> 1, the tachometer 10 discretizes the time-series vibration data detected by the vibrometer 20 and measures the rotation speed of the rotation mechanism provided in the gasoline engine 30.

  In step S2, the search unit 12d of the tachometer 10 searches for a peak around the measurement peak (peak spectrum data) of the previous (measurement) cycle.

  In this case, if the peak spectrum data of the previous (measurement) cycle exists in the storage unit 16, the search range setting unit 12c calculates and sets the search range that is the frequency range near the peak spectrum data. The search unit 12d searches for certain spectrum data.

  When the peak spectrum data of the previous (measurement) cycle does not exist in the storage unit 16, the search unit 12d may search all frequency ranges, but the search range set by the operation unit (search The range can be input numerically, or the search range can be set by switching the search range of the tachometer 10).

  In step S3, when the search unit 12d of the tachometer 10 searches for a peak around the measurement peak (peak spectrum data) of the previous (measurement) cycle, the search unit 12e of the tachometer 10 The spectrum data having a frequency approximately n times the frequency of the peak (spectrum data) searched by the unit 12d and the spectrum data having a frequency approximately 1 / n times the frequency of the spectrum data are searched (where n is 2 or more). Natural number).

  In step S4, the calculation unit 12f of the tachometer 10 determines whether spectrum data having a frequency of approximately n times and spectrum data having a frequency of approximately 1 / n times exist in the peak (spectrum data) searched by the search unit 12d. It is determined whether or not (where n is a natural number of 2 or more).

  When at least one of spectrum data having a frequency of approximately n times and spectrum data having a frequency of approximately 1 / n times exists in the peak (spectrum data) (step S4: YES), the process proceeds to step S5, and the peak ( In the case where neither spectral data having a frequency of about n times or spectral data having a frequency of about 1 / n times (where n is a natural number of 2 or more) are not present in (spectrum data) (step S4: NO) Proceed to step S6.

  In step S5, the calculation unit 12f of the tachometer 10 determines a measurement peak (peak spectrum data) based on the determination result in step S4.

  The calculation unit 12f of the tachometer 10 has retrieved both spectrum data having a frequency approximately n times the frequency of the spectrum data and spectrum data having a frequency approximately 1 / n times (where n is a natural number of 2 or more). Spectral data is determined as peak spectral data.

  When there is no spectrum data in which both spectrum data having a frequency approximately n times the frequency of the spectrum data and spectrum data having a frequency approximately 1 / n times (where n is a natural number of 2 or more) are not found, The calculation unit 12f of the tachometer 10 determines the spectrum data searched for at least one of them as the peak spectrum data.

  In step S6, the calculation unit 12f of the tachometer 10 has a spectrum of the spectrum data (frequency) determined as the peak spectrum data in step S5 and a frequency approximately n times as high as the spectrum data determined as the peak spectrum data in step S5. At least one of the data (frequency thereof) and spectrum data (frequency thereof) having a frequency approximately 1 / n times (where n is a natural number of 2 or more) is stored in the storage unit 16.

  In step S7, the calculation control unit 12 of the tachometer 10 ends the measurement when the measurement is to be ended (step S7: YES), and returns to step S1 when the measurement is to be continued (step S7: NO). Continue frequency measurement.

  The end of the frequency measurement is determined by inputting a signal from the operation unit or the like.

  In step S8, frequency measurement is performed using a conventional technique (see, for example, Patent Document 1).

  As described above, according to the present invention, the spectrum data stored in the storage unit 16 is not the spectrum data in which only the main component (rotation primary component) is searched, but other order peaks in addition to the peak of the rotation primary component. Therefore, it is possible to improve the followability of the peak of the rotation primary component (no longer follow the peak due to noise).

  That is, not only the peak due to the main component (rotation primary component) but also other order peaks are simultaneously tracked, noise is removed, and more stable rotation speed measurement can be realized even during acceleration / deceleration.

  Next, another example of the embodiment will be described in detail with reference to FIG.

In FIG. 4 (a), the measurement peak (spectrum data) f1 of the rotational speed of the rotation mechanism provided in the gasoline engine 30, spectrum data 2f1 having a frequency approximately twice the frequency of the measurement peak f1, and the frequency of the spectrum data. Shows a state where spectrum data (f1 / 2) having a frequency approximately ½ times the frequency and other noise components f3 and f4 are retrieved.

In FIG. 4 (a), the operation unit 12f of the tachometer 10 uses the spectrum data 2f1 having a frequency approximately doubled and the spectrum data f1 obtained by searching the spectrum data (f1 / 2) having a frequency approximately 1/2 times as gasoline. This is determined as a measurement peak (peak spectrum data) of the rotation speed of the rotation mechanism provided in the engine 30.

  FIG. 4B shows the discrete Fourier data continuously input from the input unit 11 by the FFT calculation unit 12a in the measurement cycle next to the measurement cycle of FIG. ), And shows the result of calculating the spectrum data of vibration.

  In FIG. 4B, the search unit 12d sequentially searches for spectrum data in a range of Δf1 (frequency range to be searched) as a frequency search range near the peak spectrum data.

  In the search range Δf1, the spectrum data f1 determined as the peak spectrum data in FIG. 4A does not exist, but the spectrum data f5, f6, and f7 exist, and these are searched by the search unit 12d.

Next, the search unit 12e sequentially uses the spectrum data f5, f6, and f7 searched by the search unit 12d in order to obtain spectrum data having a frequency approximately n times that of each spectrum data or spectrum data having a frequency approximately 1 / n times ( Here, it is searched whether n is a natural number of 2 or more.

  Here, the order in which the spectrum data f5, f6, and f7 are searched can be arbitrarily set (in the order in which the peak spectrum data is searched earlier).

  Here, for the spectrum data f5, the search unit 12d starts with the spectrum data of approximately n times the frequency or the spectrum data of the approximately 1 / n times the frequency (where n is a natural number of 2 or more). )

  However, since there is no spectrum data having a frequency approximately n times that of the spectrum data f5 and spectrum data having a frequency approximately 1 / n times (here, n is a natural number of 2 or more), it cannot be searched.

  In this case, the search unit 12e does not store the spectrum data f5 in the storage unit 16 or the like.

  Next, the search unit 12e searches the spectrum data f6 for spectrum data having a frequency of approximately n times or spectrum data having a frequency of approximately 1 / n times (where n is a natural number of 2 or more).

Since the spectrum data 2f6 having a frequency approximately twice that of the spectrum data f6 and the spectrum data (f6 / 2) having a frequency approximately ½ times exist as shown in FIG. 4C, the search unit 12e Can be searched.

Accordingly, the search unit 12e stores the spectrum data f6, the spectrum data 2f6 having a frequency approximately twice that of the spectrum data f6, and the spectrum data (f6 / 2) having a frequency approximately 1/2 times that of the spectrum data f6 in the storage unit 16 or the like. Remember.

In this case, the value of the spectrum data f1 retrieved in the previous measurement cycle and stored as spectrum data, the spectrum data 2f1 having a frequency approximately twice that of the spectrum data f1, and approximately 1/2 times the spectrum data f1. The search unit 12e deletes the frequency spectrum data (f1 / 2) from the storage unit 16 or the like.

  Next, the search unit 12e searches the spectrum data f7 for spectrum data having a frequency of approximately n times or spectrum data having a frequency of approximately 1 / n times (where n is a natural number of 2 or more).

  In this case, since there is no spectrum data having a frequency approximately n times that of the spectrum data f7 and spectrum data having a frequency approximately 1 / n times (where n is a natural number of 2 or more), the search cannot be performed.

  Accordingly, the search unit 12e does not store the spectrum data f7 in the storage unit 16 or the like.

  Next, since the search by the search unit 12e for the spectrum data f5, f6, and f7 in the search range Δf1 is completed, the calculation unit 12f performs the next calculation.

  The computing unit 12f checks the spectrum data stored in the storage unit 16.

  As shown in FIG. 4C, spectrum data having a frequency approximately n times that of the spectrum data stored in the storage unit 16 or spectrum data having a frequency approximately 1 / n times that of the spectrum data (where n is 2 or more). Since the spectrum data having at least one of (natural number) is only the spectrum data f6, the calculation unit 12f determines the spectrum data f6 as the peak spectrum data.

  Next, the calculation unit 12 f displays the calculated peak spectrum data on the display unit 13. The peak spectrum data displayed on the display unit 13 indicates the rotation frequency of the rotation mechanism of the gasoline engine 30.

  As described above, according to the present invention, the spectrum data stored in the storage unit 16 is not the spectrum data in which only the main component (rotation primary component) is searched, but other order peaks in addition to the peak of the rotation primary component. Therefore, it is possible to improve the followability of the peak of the rotation primary component (no longer follow the peak due to noise).

  That is, not only the peak due to the main component (rotation primary component) but also other order peaks are simultaneously tracked, noise is removed, and more stable rotation speed measurement can be realized even during acceleration / deceleration.

  As mentioned above, although embodiment of this invention was described, this invention is not restricted to embodiment mentioned above. The effects described in the embodiments of the present invention are only the most preferable effects resulting from the present invention, and the effects of the present invention are limited to those described in the embodiments of the present invention. is not.

  3 is recorded in advance on a recording medium such as a hard disk or recorded in advance via a network such as the Internet, and is read and executed by a general-purpose microcomputer or the like. It is possible to cause the general-purpose microcomputer or the like to function as a CPU according to the embodiment.

DESCRIPTION OF SYMBOLS 10 Tachometer 11 Input part 12 Operation control part 12a FFT operation part 12b Judgment part 12c Search range setting part 12d Search part 12e Search part 12f Calculation part 13 Display part 14 Input amplifier part 15 Filter part 16 Storage part 20 Vibrometer 30 Gasoline engine

Claims (8)

FFT (Fast Fourier Transformation) computing means for computing spectral data by performing fast Fourier transform on the input signal;
Determination means for determining whether peak spectrum data, which is one of the spectrum data, is stored in the storage means;
A search range setting means for setting a frequency near the peak spectrum data in a search range when the determination means determines that the peak spectrum data is stored in the storage means;
Search means for searching the spectrum data calculated by the FFT calculation means in the search range;
Search means for searching for spectrum data having a frequency approximately n times the frequency of the spectrum data searched by the search means and spectrum data having a frequency approximately 1 / n times the frequency of the spectrum data;
When at least one of spectrum data having a frequency approximately n times the frequency of the spectrum data and spectrum data having a frequency approximately 1 / n times is searched by the search means, the spectrum data is converted into the peak spectrum data. And calculating means for storing the peak spectrum data in the storage means,
A signal processing device (where n is a natural number of 2 or more). The signal processing device according to claim 1,
When there are a plurality of spectrum data to be searched by the search means, the search means is substantially the same as the spectrum data having the frequency of approximately n times the frequency of the spectrum data from the spectrum data having the largest spectrum data size. A signal processing apparatus that searches for spectrum data having a frequency of 1 / n times (where n is a natural number of 2 or more). The signal processing device according to claim 1 or 2,
There are a plurality of the spectrum data in which at least one of spectrum data having a frequency of approximately n times the frequency of the spectrum data and spectrum data having a frequency of approximately 1 / n times is searched in the search means, When there is the spectrum data in which both spectrum data having a frequency approximately n times the frequency and spectrum data having a frequency approximately 1 / n times are found, a spectrum having a frequency approximately n times the frequency of the spectrum data is present. A signal processing apparatus (where n is a natural number equal to or greater than 2), wherein the spectrum data from which both data and spectrum data having a frequency of about 1 / n times are searched is determined as the peak spectrum data. In the signal processing device according to any one of claims 1 to 3,
When there are a plurality of spectrum data in which both spectrum data having a frequency approximately n times the frequency of the spectrum data and spectrum data having a frequency approximately 1 / n times (where n is a natural number of 2 or more) are searched. In the signal processing apparatus, the spectrum data having the largest size of the spectrum data is determined as the peak spectrum data. In the signal processing device according to any one of claims 1 to 4,
The signal is a detection result of a time-series physical phenomenon caused by a rotational motion measured from a measurement target,
The frequency of the peak spectrum data determined by the computing means is the rotational speed of the measurement object;
A signal processing device. In the signal processing device according to any one of claims 1 to 5,
The search range setting means determines the upper limit of the search range at a frequency smaller than n times the frequency of the peak spectrum data, and sets the lower limit of the search range from 1 / n times the frequency of the peak spectrum data. Set at a higher frequency,
(Where n is a natural number of 2 or more). FFT calculation process for calculating spectrum data by performing fast Fourier transform on the input signal;
A determination step of determining whether or not peak spectrum data that is one of the spectral data in the storage step is stored;
A search range setting step of setting a frequency near the peak spectrum data in a search range when it is determined that the peak spectrum data in the storage step is stored in the determination step;
In the search range, a search step for searching the spectrum data calculated in the FFT calculation step;
A search step of searching for spectrum data having a frequency of approximately n times the frequency of the spectrum data searched in the search step and spectrum data having a frequency of approximately 1 / n times the frequency of the spectrum data;
In the search step, when at least one of spectrum data having a frequency approximately n times the frequency of the spectrum data and spectrum data having a frequency approximately 1 / n times is searched, the spectrum data is converted to the peak spectrum data. And calculating step for storing the peak spectrum data in the storing step;
A signal processing device control method (where n is a natural number of 2 or more). A program causing a computer to execute the method according to claim 7 .

Images