JP2019045378A - Oscillation type load sensor - Google Patents

Abstract

To provide an oscillation type load sensor capable of making a load measurement rate constant.SOLUTION: A tuning-fork sensor 100 incudes: a load receiving unit 16 for receiving a load F; a tuning-fork oscillator 1 for changing an oscillation frequency in accordance with magnitude of the load F received by the load receiving unit 16; a piezoelectric element 4b for detecting an oscillation waveform of the tuning-fork oscillator 1; an AD converter 62 for converting the oscillation waveform into a digital signal at a constant sampling frequency; a digital processing unit 64 for specifying the oscillation frequency of the tuning-fork oscillator 1 for each sampling period corresponding to a sampling frequency, on the basis of the digital signal; and a load specification unit 65 for specifying the magnitude of the load F received by the load receiving unit 16, on the basis of the specified oscillation frequency.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a vibratory load sensor.

  BACKGROUND Conventionally, there has been known a vibration type load sensor that measures a load using a vibrator that changes a natural frequency (frequency) according to a load (force). For example, Patent Document 1 discloses a load conversion mechanism for measuring a load based on a change in vibration frequency of a tuning fork type vibrator. In such a vibration type load sensor, as a method of measuring the vibration frequency of the vibrator, for example, a reciprocal method as disclosed in Patent Document 2 is known.

Japanese Examined Patent Publication No. 3-49059 JP, 2011-221007, A

  By the way, the above-mentioned reciprocal system is a system which calculates | requires a frequency by calculating | requiring the period of an input signal first, and calculating the reciprocal of the said period. Therefore, in the reciprocal method, the timing at which the vibration frequency is obtained depends on the period of the input signal. That is, in the reciprocal method, the oscillation frequency can not be obtained at a constant measurement rate without depending on the period of the input signal. For this reason, there is a problem that it is difficult to apply the reciprocal method when it is necessary to perform dynamic control (for example, control to move a predetermined mechanism according to a load change) based on a load measurement result.

  Therefore, one aspect of the present invention is to provide a vibration type load sensor capable of making a load measurement rate constant.

  A vibration type load sensor according to one aspect of the present invention includes a load receiving portion that receives a load, a vibrator that changes a vibration frequency according to the size of a load received by the load receiving portion, and a vibration waveform of the vibrator Digital processing that specifies the vibration frequency of the vibrator for each sampling cycle corresponding to the sampling frequency based on the detection unit that detects, the AD conversion unit that converts the vibration waveform to a digital signal at a fixed sampling frequency, and the digital signal And a load specifying unit that specifies the magnitude of the load received by the load receiving unit based on the identified vibration frequency.

  The vibration type load sensor converts the vibration waveform into a digital signal at a constant sampling frequency, and acquires the vibration frequency of the vibrator for each sampling cycle based on the digital signal. Further, the magnitude of the load at each sampling time point is specified from each vibration frequency obtained for each sampling cycle. As described above, according to this vibration type load sensor, the load measurement result can be obtained for each sampling cycle, so that the load measurement rate can be made constant.

  The digital processing unit acquires a first digital signal having the same phase as the digital signal, and Hilbert transforms the digital signal to acquire a second digital signal that is 90 ° out of phase with the digital signal. Even if the phase angle is calculated based on the ratio of one digital signal to the second digital signal and the vibration frequency is calculated based on the phase angle calculated at each of two consecutive sampling points for each sampling period. Good. In this case, it is possible to easily realize a configuration in which the load measurement rate is constant by utilizing the Hilbert transform.

  The digital processing unit may specify the vibration frequency by performing a filtering process using an adaptive notch filter on the digital signal. In this case, by performing the filtering process using the adaptive notch filter, it is possible to easily realize a configuration that makes the load measurement rate constant.

  The vibration type load sensor further includes a band pass filter for inputting the digital signal converted by the AD conversion unit and attenuating frequency components outside the predetermined band of the digital signal, and the digital processing unit is a band pass filter The vibration frequency may be specified based on the digital signal after passing through. In this case, the band pass filter can reduce noise contained in the digital signal. As a result, the measurement accuracy of the vibration frequency (that is, the accuracy of load measurement based on the vibration frequency) can be improved.

  The predetermined band is a band set based on the natural frequency of the vibration mode predetermined as the vibration mode used for load measurement and the amount of change of the frequency according to the predetermined measurable load. May be In this case, frequency components in bands other than the frequency band necessary for load measurement can be appropriately attenuated, and noise contained in the digital signal can be effectively reduced.

  The vibrator may be a tuning fork type vibrator. In this case, the above-described vibration type load sensor can be specifically realized by using a tuning fork type vibrator.

  According to one aspect of the present invention, it is possible to provide a vibration type load sensor capable of making a load measurement rate constant.

It is a schematic perspective view of a tuning fork sensor which is one embodiment of a vibration type load sensor. It is a block diagram which shows the function structure of a load measurement part. It is a figure which shows the frequency band required for the load measurement by a tuning fork sensor. It is a figure which shows an example of the voltage waveform obtained by A / D conversion processing. It is a figure which shows an example of the voltage waveform obtained by band pass filter processing. It is a figure which shows an example of the frequency waveform obtained by frequency conversion which used Hilbert transform. It is a block diagram which shows schematic structure of a digital filter. It is a figure which shows an example of the frequency waveform obtained by a digital filter. (A) The weight measurement result at the time of using a reciprocal system, and (B) It is a figure which shows the weight measurement result by the measuring method which concerns on this embodiment. It is a figure which shows schematic structure of an adaptive notch filter. It is a figure which shows the circuit structural example of an adaptive notch filter.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings. The same parts or corresponding parts in the respective drawings are denoted by the same reference numerals, and redundant description will be omitted.

  FIG. 1 is a schematic perspective view of a tuning fork sensor that is an embodiment of a vibrating load sensor according to the present invention. The tuning fork sensor 100 includes the tuning fork vibrator 1 configured to change the natural frequency (frequency) according to the magnitude of the load F applied to the load receiving unit 16. The tuning fork sensor 100 also has a load measuring unit 6 that measures the size of the load F based on the vibration frequency of the tuning fork vibrator 1. The load measuring unit 6 may be configured by a computer including a processor such as a CPU, a memory, and the like. Hereinafter, the configuration of tuning fork sensor 100 will be described in detail.

  The tuning fork vibrator 1 includes two rectangular plate-shaped vibrating bars 1a and 1b arranged in parallel with each other and a U-shaped connecting portion 2a connecting both ends of the vibrating bars 1a and 1b in the longitudinal direction D. , 2b, vibrating pieces 1a and 1b, and connecting portions 2a and 2b. The thin plate-like support pieces 3a and 3b support the vibrating portion on both sides in the longitudinal direction D.

  Piezoelectric elements 4a and 4b are attached to both side surfaces of one coupling portion (here, coupling portion 2b) of the tuning fork vibrator 1 by, for example, adhesion and vapor deposition. The piezoelectric elements 4a and 4b are respectively connected to the output part and the input part of the amplifier 5 installed outside. Thus, the piezoelectric element 4a operates as an excitation for vibrating the tuning fork vibrator 1 at a constant frequency, and the piezoelectric element 4b operates as a pickup for detecting the vibration of the tuning fork vibrator 1. That is, the piezoelectric element 4 b functions as a detection unit that detects the vibration waveform of the tuning fork vibrator 1. The vibrating bars 1a and 1b are vibrated at a predetermined frequency in advance based on the gain (gain) set in the amplifier 5, the frequency characteristic, and the like.

  The tuning fork sensor 100 has a lever portion 10 and a base 11 as a transmission mechanism for transmitting the load F of the load receiving portion 16 to the tuning fork vibrator 1. One end of the tuning fork vibrator 1 in the longitudinal direction D is connected to one end of the lever portion 10 via the support piece 3 a. The one end of the lever portion 10 constitutes a power point 12 in the lever portion 10. The other end of the tuning fork vibrator 1 in the longitudinal direction D is attached to the base 11 via the support piece 3 b. The lever portion 10 is supported by a thin supporting point 13 connected to the base 11. The other end of the lever 10 opposite to the power point 12 with respect to the fulcrum 13 constitutes a weight 14. The load receiving portion 16 is suspended to the weight 14 via a thin plate-like tension piece 15. With such a configuration, the load F applied to the load receiving portion 16 pulls the tuning fork vibrator 1 via the lever portion 10. The magnitude of the force that the load F acts on the tuning fork vibrator 1 is adjusted by the ratio of the distance from the fulcrum 13 to the force point 12 and the distance from the fulcrum 13 to the weight 14 or the like.

  The attachment holes 17 and 18 of the base 11 are for fixing the main body (portion excluding the amplifier 5 and the load measurement unit 6) of the tuning fork sensor 100 to a housing or the like (not shown). The hooks 19 of the load receiving portion 16 shown by a broken line in FIG. 1 are auxiliary used as needed depending on the shape of the object to be measured and the like. The size and shape of the main body of the tuning fork sensor 100 (for example, the shapes of the lever 10 and the base 11) are appropriately changed according to the place of use, use, etc. (for example, use for measuring a small load of 5 kgf or less). obtain.

  Next, measurement processing by the load measurement unit 6 will be described. The outline of the measurement process is as follows. When a load F is applied to the load receiving portion 16, the tuning fork vibrator 1 is pulled through the transmission mechanism such as the lever portion 10 and the base 11, and the vibration frequency of the tuning fork vibrator 1 changes. The load measuring unit 6 obtains, via the piezoelectric element 4b, a voltage waveform (waveform in which the horizontal axis represents time and the vertical axis represents voltage) which is an analog signal indicating the magnitude of the voltage according to the vibration of the tuning fork vibrator 1 Do. As a result, the load measuring unit 6 can obtain a voltage waveform in which the change in the vibration frequency of the tuning fork vibrator 1 is reflected. Subsequently, the load measuring unit 6 executes A / D (analog / digital) conversion processing and digital processing to be described later on the voltage waveform to vibrate the tuning fork vibrator 1 according to the size of the load F. Identify the frequency (eg, the natural frequency corresponding to the primary mode of vibration). Subsequently, the load measuring unit 6 specifies the size of the load F based on the specified vibration frequency.

  FIG. 2 is a block diagram showing a functional configuration of the load measuring unit 6. As shown in FIG. 2, the load measuring unit 6 includes an analog LPF (low pass filter) 61, an AD converter 62 (AD converter), a BPF (band pass filter) 63, a digital processing unit 64, and a load. And an identification unit 65. The digital processing unit 64 includes a Hilbert transform unit 641 and a digital filter 642. Hereinafter, each function of the load measuring unit 6 will be described.

  The analog LPF (low pass filter) 61 inputs a voltage waveform which is an analog signal acquired through the piezoelectric element 4 b, and removes (attenuates) a frequency component higher than a frequency band required for load measurement by the tuning fork sensor 100 ) Is an analog filter configured as

  FIG. 3 is a diagram showing a frequency band required for load measurement by the tuning fork sensor. FIG. 3 shows a waveform obtained by converting a voltage waveform, which is a time waveform observed in a state where a predetermined load F is applied to the load receiving portion 16, into a spectrum in the frequency domain. In addition, the said waveform is shown for description, It is not necessary to acquire by the load measurement part 6. FIG.

  In the present embodiment, as an example, load measurement is performed based on the natural frequency corresponding to the primary mode vibration of the tuning fork vibrator 1. That is, the primary mode of the tuning fork vibrator 1 is predetermined as a vibration mode used for load measurement. Here, the natural frequency of the primary mode in the default state (that is, the state in which the load F is not applied to the load receiving portion 16) is represented as f, and the load F included in a predetermined measurable load range Is the maximum variation of the natural frequency when it is applied to the load receiving portion 16 as .DELTA.f. The f and Δf are grasped in advance based on the main body shape of the tuning fork sensor 100, the measurable load, and the like. In this case, a region R between the frequency f and the frequency f + Δf is a frequency band required for load measurement by the tuning fork sensor 100. In this case, the analog LPF 61 may be configured to remove frequency components higher than the frequency f + Δf.

  The AD converter 62 converts a vibration waveform (a voltage waveform after passing through the analog LPF 61 in this embodiment) detected by the piezoelectric element 4 b into a digital signal at a predetermined sampling frequency. For example, the AD converter 62 can be configured as an electronic circuit that performs known A / D conversion processing on the input analog signal and outputs the converted digital signal. FIG. 4 is a diagram showing an example of a voltage waveform obtained by the A / D conversion process. The waveform examples shown in FIG. 4 and FIGS. 5, 6, and 8 described below are shown as extreme waveform examples in order to facilitate understanding of the contents of various conversion processes in the present embodiment. This may differ from the waveform data obtained by the actual measurement.

  The BPF (band pass filter) 63 is a digital circuit that receives the digital signal converted by the AD converter 62 and attenuates frequency components outside the predetermined band of the digital signal. Here, it is set based on the natural frequency of the vibration mode (in this case, the primary mode) predetermined as the vibration mode used for load measurement, and the variation of the frequency according to the predetermined measurable load. Can be set as the above-mentioned "predetermined band". That is, in the present embodiment, the frequency band (region R between the frequency f and the frequency f + Δf) necessary for the above-described load measurement may be set as the above-described “predetermined band”.

  The BPF 63 is used to attenuate the noise contained in the voltage waveform (see FIG. 4). The above-mentioned "noise" is a portion that can be a factor that lowers the measurement accuracy of the vibration frequency of the tuning fork vibrator 1 (here, the natural frequency of the first mode) in the voltage waveform. For example, the noise may include a vibration component related to a vibration mode (in the present embodiment, a second or higher order mode) that is not used for load measurement of the tuning fork vibrator 1. Further, the noise may include disturbance noise (for example, a vibration component due to a vibration of a support (not shown) supporting the base 11 of the tuning fork sensor 100) transmitted to the main body of the tuning fork sensor 100.

  FIG. 5 is a diagram showing an example of a voltage waveform obtained by the band pass filter processing of the BPF 63. As shown in FIG. In this example, frequency components outside the region R (see FIG. 3) are attenuated by the band pass filter processing, and a voltage waveform having a relatively neat shape is obtained.

  The digital processing unit 64 specifies the vibration frequency of the tuning fork vibrator 1 for each sampling period T corresponding to the sampling frequency in the A / D conversion processing based on the digital signal (see FIG. 5) that has passed through the BPF 63. In the present embodiment, the digital processing unit 64 includes a Hilbert transform unit 641 and a digital filter 642.

  The Hilbert transform unit 641 executes the following first to third processing.

(First processing)
The Hilbert transform unit 641 obtains the first digital signal D1 having the same phase as the digital signal (see FIG. 5) that has passed through the BPF 63, and Hilbert transforms the digital signal to obtain a 90 ° phase with respect to the digital signal. A different second digital signal D2 is obtained. The circuit for executing the first process is configured by a known digital circuit including, for example, an FIR filter, a delay circuit, and the like. The first digital signal D1 and the second digital signal D2 are represented, for example, by the following equations (1) and (2). Here, A represents an amplitude, ω represents an angular frequency, and t represents an arbitrary sampling point.
D1 = A · sin (ωt) (1)
D2 = A · cos (ωt) (2)

(Second processing)
Subsequently, the Hilbert transform unit 641 calculates the phase angle θ (= ωt) based on the ratio (D1 / D2) of the first digital signal D1 to the second digital signal D2 obtained by the first process. Specifically, the Hilbert transform unit 641 finds the inverse tangent (tan ⁻¹ ) by calculation from the relational expression (D1 / D2 = sin (ωt) / cos (ωt) = tan (ωt)) of the ratio. Thus, the phase angle θ is calculated.

(Third process)
Subsequently, the Hilbert transform unit 641 determines, based on the phase angles θ and θ ₋₁ calculated by the second process at each of two consecutive sampling time points t and t ₋₁ (t = t ₋₁ + T). The vibration frequency at the sampling time point t is calculated. The sampling time point t ₋₁ is a sampling time point one before the sampling time point t. Specifically, the Hilbert transform unit 641 calculates the vibration frequency f _t at the sampling time point t by the following equation (3).
f _t = (θ−θ ₋₁ ) / 2πT (3)

  FIG. 6 is a diagram showing an example of a frequency waveform obtained by frequency conversion using Hilbert transform (that is, the first to third processes of the Hilbert transform unit 641).

  The digital filter 642 is a digital filter to further attenuate noise according to the application. Here, “noise” means noise of a level not removed by the BPF 63, noise superimposed by Hilbert transform, and the like.

  FIG. 7 is a block diagram showing a schematic configuration of the digital filter 642. As shown in FIG. 7, the digital filter 642 is configured by a low pass filter (LPF) that passes only the component close to the direct current of the input signal (the digital signal obtained by the above-described Hilbert transform unit 641). In the present embodiment, as an example, the digital filter 642 is a multistage filter by a first stage filter circuit including the FIR filter 21 and the decimator 22 and a second stage filter circuit including the FIR filter 23 and the decimator 24. It is configured. However, the digital filter 642 may be configured as a multistage filter of three or more stages. Further, in the present embodiment, a multi-stage configuration is adopted in order to suppress the number of taps that can be covered by the CPU power of the load measuring unit 6, but a single-stage configuration may be used if the CPU power is sufficient.

  The FIR filters 21 and 23 are configured as low pass filters (LPF) that allow only components close to direct current to pass. The low pass filters (FIR filters 21 and 23) may be implemented as a finite impulse response filter (FIR filter) as shown in FIG. 7 or as an infinite impulse response filter (IIR filter).

  The decimators 22 and 24 are circuits that execute decimation (down sampling) which is thinning processing of data. In the present embodiment, the decimator 22 is configured to decimate data every M times, and the decimator 24 is configured to decimate data every N times. Here, M and N are predetermined integers.

  FIG. 8 is a diagram showing an example of a frequency waveform obtained by the digital filter 642. As shown in FIG. 8, by the processing of the digital filter 642, a frequency waveform in which noise is further reduced as compared with the frequency waveform (see FIG. 6) obtained by the Hilbert transform unit 641 is obtained.

  The load specifying unit 65 specifies the magnitude of the load F received by the load receiving unit 16 based on the vibration frequency specified by the digital processing unit 64 as described above. Specifically, the load identification unit 65 converts the identified vibration frequency into the magnitude of the load F. For example, a function or the like that defines the relationship between the natural frequency of the primary mode of the tuning fork sensor 100 and the magnitude of the load F is prepared in advance. In this case, the load identification unit 65 can convert the identified vibration frequency into a weight value representing the magnitude of the load F by using the function.

  The tuning fork sensor 100 described above converts the vibration waveform (voltage waveform which is an analog signal) acquired from the piezoelectric element 4 b into a digital signal at a constant sampling frequency, and based on the digital signal, tuning fork vibration at every sampling period T The vibration frequency of the child 1 is acquired. Further, the magnitude of the load F at each sampling point is specified from each vibration frequency obtained for each sampling period T. As described above, according to the tuning fork sensor 100, since the load measurement result can be obtained for each sampling cycle T, the load measurement rate can be made constant.

In addition, the Hilbert converter 641 (digital processor 64) acquires the first digital signal D1 having the same phase as the digital signal (here, the digital signal after passing through the BPF 63 (see FIG. 5)), and the digital signal The second digital signal D2 having a phase difference of 90 ° with respect to the digital signal is obtained by Hilbert transform, and the phase angle θ is calculated based on the ratio between the first digital signal D1 and the second digital signal D2, The vibration frequency f _t is calculated based on the phase angles θ and θ ₋₁ calculated at each of two successive sampling time points t and t ₋₁ for each sampling period T. Thus, by utilizing the Hilbert transform, a configuration can be easily realized in which the load measurement rate is made constant.

  The tuning fork sensor 100 also includes a BPF 63 that receives the digital signal (see FIG. 4) converted by the AD converter 62 and attenuates frequency components outside the predetermined band of the digital signal. Then, the digital processing unit 64 specifies the vibration frequency of the tuning fork vibrator 1 based on the digital signal (see FIG. 5) after passing through the BPF 63. In this case, the band pass filter can reduce the noise contained in the digital signal (see FIG. 4). As a result, it is possible to improve the measurement accuracy of the vibration frequency (that is, the accuracy of the load measurement based on the vibration frequency) based on the noise-reduced digital signal (see FIG. 5).

  The above-mentioned "predetermined band" is the natural frequency of the vibration mode (first mode in this case) predetermined as the vibration mode used for load measurement and the frequency corresponding to the predetermined measurable load. It is a zone (here, the region R shown in FIG. 3) set based on the amount of change. In this case, frequency components in bands other than the frequency band (that is, region R) necessary for load measurement can be appropriately attenuated, and noise included in the digital signal (see FIG. 4) can be effectively reduced.

  The vibration type load sensor according to the present embodiment is the tuning fork sensor 100 including the tuning fork type vibrator (the tuning fork vibrator 1) as a vibrator. Thus, the above-described vibration type load sensor can be specifically realized by the configuration using the tuning fork type vibrator.

  FIG. 9 shows the result of weight measurement when vibration is applied to the main body of the tuning fork sensor 100 and no load F is applied to the load receiving portion 16 (ie, when F = 0). ing. The graph (A) shows the result of weight measurement in the case where the processing by the Hilbert transformer 641 described above in the measurement method (load measurement unit 6) according to the present embodiment is replaced with the conventional reciprocal processing. A graph (B) has shown a weight measurement result at the time of using a measurement method (load measurement part 6) concerning this embodiment.

  In the reciprocal method, the frequency is calculated by first measuring the period and calculating the reciprocal of the measured period. That is, in order to obtain the vibration frequency, a vibration waveform for at least one cycle of the input signal is required. That is, in the reciprocal method, at least one cycle of the input signal is required to perform load measurement by specifying the vibration frequency from the vibration waveform (input signal) detected by the piezoelectric element 4b. For this reason, for example, if noise that affects the accuracy of the load measurement is included within the time of one cycle of the input signal retroactively from the current time, the noise is affected. That is, in the reciprocal method, the influence of the noise remains on the obtained load measurement result until the time corresponding to one cycle of the input signal elapses on the basis of the point when the noise disappears. On the other hand, in the present embodiment, since it is not necessary to specify the cycle first to specify the frequency as in the reciprocal system, a time shorter than one cycle of the input signal (specifically, after the noise disappears) It becomes possible to obtain a load measurement result without the influence of noise in the above-mentioned sampling period T).

  Therefore, as shown in the graph (A), when the reciprocal method is used, it is easily affected by the noise caused by the vibration applied to the main body of the tuning fork sensor 100, and the weight measurement result is relatively relatively large. Large errors (difference from weight value 0) occur. On the other hand, as shown in the graph (B), the measurement method according to this embodiment (that is, the method of executing the processing by the Hilbert transform unit 641 described above instead of the reciprocal method) is more than the case using the reciprocal method. The influence of noise is reduced and the measurement accuracy of the weight value (0 g) is improved. In addition, since the measurement method according to the present embodiment is excellent in noise resistance as compared with the conventional reciprocal method, when the load dynamically changes (for example, the powdery substance is lowered). This is particularly advantageous as compared to the conventional reciprocal method in situations where the noise is likely to occur, such as when measuring the weight of a feeder that is vibrating to cause it to drop.

  The preferred embodiments of the present invention have been described above in detail, but the present invention is not limited to the above embodiments. For example, the analog LPF 61, the BPF 63, and the digital filter 642 in the load measurement unit 6 may be appropriately omitted according to the measurement accuracy obtained in the load measurement. In the present embodiment, the tuning fork sensor 100 has been described as one form of the vibrating load sensor, but the configuration of the present embodiment (in particular, the load measuring unit 6) is a vibrating type other than the tuning fork type (for example, a string vibration type, It is applicable also to a load sensor of crystal type etc.).

  Further, as shown in FIG. 10, the Hilbert transformer 641 may be replaced by an adaptive notch filter (adaptive notch filter) 66. That is, the digital processing unit 64 may specify the vibration frequency of the tuning fork vibrator 1 by performing the filtering process by the adaptive notch filter 66 on the digital signal (see FIG. 5) that has passed through the BPF 63. The adaptive notch filter 66 is a digital circuit that adapts the notch frequency to the peak frequency of the input signal. As described above, also by performing the filtering process by the adaptive notch filter 66 instead of the Hilbert transform unit 641, a configuration for making the load measurement rate constant can be easily realized.

  As a circuit configuration of the adaptive notch filter 66, a circuit configuration of a known adaptive notch filter can be used. For example, as shown in FIG. 10, the adaptive notch filter 66 includes an adaptive algorithm execution unit A that executes an adaptive algorithm. The adaptation algorithm is an algorithm for adapting the notch frequency to the peak frequency of the input signal x (n) (that is, the digital signal output by the BPF 63). The adaptive algorithm execution unit A is a digital circuit that executes such an adaptive algorithm. Here, "n" indicates a sampling point. The adaptive algorithm execution unit A outputs an error e (n) between the input signal x (n) and the result of executing the adaptive algorithm on the input signal x (n). Then, the adaptive algorithm execution unit A adapts the notch frequency to match the peak frequency of the input signal by updating the adaptive algorithm based on the error e (n).

FIG. 11 is a diagram showing an example of the circuit configuration of the adaptive notch filter 66. As shown in FIG. Moreover, following formula (1)-(3) is an example of the adaptive algorithm in the said circuit structure. The following equation (3) is a calculation equation used when calculating the initial value of a (n). In the following formulas (1) to (3), a (n) is a gain coefficient updated at each sampling time point. u (n) is the sum of the input signal x (n) and the signal output from the adaptive algorithm. r is the notch Q value (notch sharpness and steepness). μ is the step size. μ ₀ is the initial step size. λ is the forgetting factor. f (n) is a notch frequency. f _s is a sampling frequency.

  DESCRIPTION OF SYMBOLS 1 ... Tuning fork vibrator, 4b ... Piezoelectric element (detection part), 6 ... Load measurement part, 16 ... Load receiving part, 62 ... AD converter (AD conversion part), 63 ... BPF (band pass filter), 64 ... Digital Processing unit, 65: load specifying unit, 66: adaptive notch filter, 100: tuning fork sensor (vibration type load sensor), 641: Hilbert converter, 642: digital filter.

Claims (6)

A load receiving portion that receives a load,
A vibrator for changing the vibration frequency according to the size of the load received by the load receiving portion;
A detection unit that detects a vibration waveform of the vibrator;
An AD converter for converting the vibration waveform into a digital signal at a constant sampling frequency;
A digital processing unit that specifies the vibration frequency of the vibrator at each sampling cycle corresponding to the sampling frequency based on the digital signal;
A load identification unit that identifies the magnitude of the load received by the load receiving unit based on the identified vibration frequency;
, Vibrational load sensor. The digital processing unit
Acquiring a first digital signal in phase with the digital signal, and Hilbert-transforming the digital signal to acquire a second digital signal that is 90 ° out of phase with the digital signal;
Calculating a phase angle based on a ratio between the first digital signal and the second digital signal;
The vibration frequency is calculated based on the phase angle calculated at each of two consecutive sampling points in each sampling period.
The vibration type load sensor according to claim 1. The digital processing unit specifies the vibration frequency by performing filtering processing using an adaptive notch filter on the digital signal.
The vibration type load sensor according to claim 1. The digital signal converted by the A / D conversion unit is input, and it further comprises a band pass filter for attenuating frequency components outside the predetermined band of the digital signal
The digital processing unit specifies the vibration frequency based on the digital signal after passing through the band pass filter.
The vibration type load sensor as described in any one of Claims 1-3. The predetermined band is a band set based on the natural frequency of the vibration mode predetermined as the vibration mode used for load measurement and the change amount of the frequency according to the predetermined measurable load. is there,
The vibration type load sensor according to claim 4. The vibrator is a tuning fork vibrator.
The vibration type load sensor as described in any one of Claims 1-5.

Images