JP5569940B2 - Characteristics estimation method and replacement method of eddy current displacement sensor - Google Patents

Description

The present invention, with the eddy current type displacement sensor for detecting vibration such as a rotary machine to be monitored, its impedance - when installing or replacing the voltage converter (hereinafter, referred to as Z / V converter) It is related to the technology used. That is, the present invention estimates the characteristics without removing the sensor head of the eddy current displacement sensor from a rotating machine or the like, and estimates the characteristics of the eddy current displacement sensor that can be replaced with an appropriate Z / V converter. The present invention relates to a method and an exchange method.

  Conventionally, an eddy current type displacement sensor is used to monitor vibrations of a rotating machine or the like. In order to measure the vibration of a rotating machine or the like with this eddy current type displacement sensor, a high-frequency magnetic field is generated by flowing a high-frequency current through a sensor coil housed inside the sensor head. When a measurement object (conductor such as metal) exists in the high-frequency magnetic field, an eddy current flows around the magnetic flux passing through the surface of the measurement object by electromagnetic induction, and the impedance of the sensor coil accommodated inside the sensor head Changes. When the impedance of the sensor coil changes, the amplitude of the alternating voltage output from the sensor head changes. Therefore, the distance between the sensor coil and the measurement object can be measured by detecting a change in the amplitude of the AC voltage using an eddy current displacement sensor. As an eddy current type displacement sensor, for example, Japanese Patent Publication No. 2-19884 “Eddy Current Displacement Meter” of Patent Document 1 has been proposed.

Japanese Patent Publication No. 2-19884

  By the way, the eddy current type displacement sensor is generally fixed to a rotating machine or the like to be monitored. If any failure occurs in the eddy current displacement sensor, the failed eddy current displacement sensor must be replaced. However, considering the installation status and cost of the sensor head, the sensor head is monitored. In many cases, it cannot be removed. In addition, since the Z / V converter is not the sensor head in which the sensor coil is housed, and the Z / V converter is almost always broken, it can be repaired by replacing only the Z / V converter.

  However, since there are many manufacturers of eddy current type displacement sensors and the specifications differ from manufacturer to manufacturer, the Z / V converter must be replaced with one that matches the characteristics of the sensor head. If the sensor head cannot be removed from the object to be monitored, the manufacturer of the eddy current type displacement sensor cannot be specified, and there is a problem that an appropriate Z / V converter cannot be selected. It was.

The present invention has been proposed in view of the above-described circumstances, and when an eddy current type displacement sensor fails, it is possible to select an appropriate Z / V converter by accurately estimating the sensor characteristics, and to quickly It is an object of the present invention to provide a characteristic estimation method and replacement method for an eddy current type displacement sensor capable of handling a failure.

The characteristic estimation method and replacement method of the eddy current type displacement sensor of the present invention have the following characteristic points in order to achieve the above-described object. That is, the characteristic estimation method and replacement method of the eddy current type displacement sensor according to the present invention have, as a common configuration, sweeping the excitation frequency of the eddy current type displacement sensor to be a characteristic estimation target (continuously within a predetermined range). A step (S1, S2, S3) of measuring a Q value characteristic including a maximum value of the Q value and an excitation frequency at the maximum value, and a Q value characteristic of the eddy current type displacement sensor to be a characteristic estimation target; A step (S4) of comparing a previously measured Q value characteristic of a known eddy current type displacement sensor and a step of estimating the type and characteristic of an eddy current type displacement sensor to be a characteristic estimation target based on the comparison result (S4) S5).

  In addition, the comparison between the Q value characteristic measured for the eddy current type displacement sensor to be the characteristic estimation target and the previously measured Q value characteristic of the eddy current type displacement sensor should be performed by statistical processing using a neural network. Is possible.

  The characteristic estimation method for an eddy current type displacement sensor according to the present invention is characterized in that the type and characteristic of an eddy current type displacement sensor to be a characteristic estimation target are estimated by the procedure described above. According to the eddy current displacement sensor replacement method of the present invention, when the eddy current displacement sensor fails, the type and characteristics of the eddy current displacement sensor to be replaced are estimated by the above-described procedure, and the sensor coil (20). The sensor head having the above is left, and only the Z / V converter (30) adjusted according to the estimated Q value characteristic of the eddy current displacement sensor is replaced (S6, S7). is there. The type and characteristics of the eddy current type displacement sensor of the present invention can be estimated by the above-described procedure. In the event of a failure, the sensor head having the sensor coil (20) is left, and the eddy current type displacement sensor estimated. Only the Z / V converter (30) adjusted according to the type and characteristics of the sensor can be replaced.

  The sensor head in the present invention includes the sensor coil (20) built in the measurement object and the coaxial cable (40) connected to the sensor coil. When the eddy current displacement sensor fails, Only the Z / V converter (30) is replaced, leaving only the sensor head.

  In the characteristic estimation method, replacement method, and eddy current type displacement sensor of the eddy current type displacement sensor having such a configuration, the impedance changes when a high frequency current is passed through the circuit including the sensor coil (20). Based on this change in impedance, it is possible to specify a Q value characteristic consisting of the maximum Q value and the excitation frequency at the maximum value.

  Then, the Q value characteristic of a known eddy current type displacement sensor that is commercially available is measured in advance, and the Q value characteristic of an unknown eddy current type displacement sensor that is the target of characteristic estimation, and the known previously measured value. By comparing the Q value characteristics of the eddy current type displacement sensor, it is estimated which of the known eddy current type displacement sensors corresponds to the unknown eddy current type displacement sensor to be the characteristic estimation target.

  At this time, the characteristics of the eddy current type displacement sensor are estimated quickly and accurately by comparing the Q value characteristics by statistical processing using a neural network.

According to the characteristic estimation method, replacement method, and eddy current type displacement sensor of the present invention, it is possible to easily and appropriately estimate the type and characteristics of an eddy current type displacement sensor that has been difficult to estimate. Can do. For this reason, when an eddy current type displacement sensor fails, it becomes possible to select an appropriate maintenance part and perform quick failure response. In particular, the type and characteristics of the eddy current type displacement sensor can be estimated when the sensor head cannot be removed from the monitored object due to the installation status and cost of the eddy current type displacement sensor. The measurement monitoring system including the eddy current type displacement sensor can be restored on the spot without waiting for ordering / delivery of equivalent parts. For this reason, not only the work efficiency in the repair of the eddy current type displacement sensor is improved, but also the operation stoppage of the equipment using the eddy current type displacement sensor can be minimized.

  Further, when the Q value characteristics are compared by statistical processing using a neural network, unlike the manual comparison of the Q value characteristics, the types and types of eddy current type displacement sensors are more quickly and accurately compared. It is possible to estimate characteristics.

It is a schematic diagram which shows the displacement detection principle in the eddy current type displacement sensor which concerns on embodiment of this invention. It is a schematic diagram which shows the measuring apparatus of the Q value characteristic of the eddy current type displacement sensor which concerns on embodiment of this invention. It is an equivalent circuit diagram of the eddy current type displacement sensor according to the embodiment of the present invention. It is a schematic diagram which shows the structure of the measuring apparatus for measuring the Q value characteristic of a known eddy current type displacement sensor. It is explanatory drawing which shows the measurement result of the Q value characteristic of a known eddy current type displacement sensor. It is a flowchart which shows the replacement | exchange method of the eddy current type displacement sensor which concerns on embodiment of this invention. It is explanatory drawing which shows the relationship between the lift-off of an eddy current type displacement sensor, and an output characteristic.

  DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments of an eddy current displacement sensor characteristic estimation method, replacement method, and eddy current displacement sensor according to the present invention will be described with reference to the drawings.

<Eddy current type displacement sensor>
FIG. 1 is a schematic diagram showing the principle of displacement detection in an eddy current displacement sensor according to an embodiment of the present invention.
As shown in FIG. 1, the eddy current type displacement sensor according to the embodiment of the present invention is arranged close to the measurement object (10), and the sensor coil (20), the Z / V converter (30), The sensor coil (20) and the voltage converter are connected by a coaxial cable (40). The Z / V converter (30) includes an oscillator (31), a resonance circuit (32), a detection circuit (33), an amplification circuit (34), and a linearizer (35).

<Displacement detection principle>
In the Z / V converter (30), an excitation voltage V (excitation frequency _fe : frequency at which parallel resonance occurs) is applied from the oscillator (31) to the resonance circuit (32), and an excitation current is applied to the sensor coil (20). I _c flows. As a result, a magnetic flux Φ _c is generated from the sensor coil (20). When the magnetic flux [Phi _c acts on the measurement target, the magnetic flux [Phi _e occurs flows an eddy current I _e in the measurement object based on the law of electromagnetic induction (10). When the magnetic flux Φ _e acts on the sensor coil (20), the flux linkage number Ψ of the sensor coil (20) changes. The flux linkage number ψ is the sum of the magnetic flux Φ _c from the sensor coil (20) itself and the magnetic flux Φ _e due to the eddy current, and can be expressed by the following equation (1).

Ψ = N (Φ _c −Φ _e ) [Wb] (1)
N: Number of turns of sensor coil (times)

Since the flux linkage number Ψ of the sensor coil (20) varies depending on the displacement x between the sensor coil (20) and the measurement object (10), the resistance R and the inductances L and Q of the sensor coil (20) depend on the displacement x. The value changes. When the resistance R, the inductance L, and the Q value change, the amplitude and phase of the current flowing through the sensor coil (20) change. By detecting this current, the voltage V _c is extracted from the Z / V converter (30). Therefore, the displacement x can be detected by measuring the voltage V _c . Specifically, the displacement x can be detected by passing V _c through the detection circuit (33) and making it an output voltage linearized through the amplification circuit (34) and the linearizer 35.

  The Q value is a quantity representing the sharpness of resonance. When the vibration of energy W at resonance angular frequency ω loses energy of only S per unit time due to Joule heat or radiation loss, Q = ωW / S = (2π × vibration energy of resonance) / (loss per cycle).

If the logarithmic attenuation of the damped vibration is δ, then Q = π / δ. In a resonance circuit in which an inductance L, an electric capacitance C, and an electric resistance R are connected in series, Q = ωL / R = 1 / R (C / L) ^1/2 , and L, C, and R are connected in parallel. Q = R / ωL = R (C / L
) ^1/2 .
Further, in the resonance circuit when the electrical resistance R and the inductance L are connected in series and connected in parallel to the capacitance C, Q = ωL / R = 1 / R (C / L) ^1/2. .

<Principle of Q value detection of eddy current displacement sensor>
FIG. 2 is a schematic diagram showing a Q-value characteristic measuring apparatus for an eddy current displacement sensor according to an embodiment of the present invention. FIG. 3 is an equivalent circuit diagram of the eddy current displacement sensor according to the embodiment of the present invention.
Next, the Q value detection principle of the eddy current type displacement sensor which is the object of characteristic estimation in the present invention will be described.
As shown in FIG. 2, the eddy current displacement sensor includes a sensor coil (20), a coaxial cable (40), and a Z / V converter (30). The sensor coil (20) It is arranged close to the object (10). In the eddy current type displacement sensor having such a configuration, the principle of measuring the displacement x between the measurement object (10) and the sensor coil (20) will be described.

Measuring the displacement x between the measurement object (10) and the sensor coil (20) means obtaining V _c using the impedance Z _s as shown in FIG. That is, the sensor coil (20) can be regarded as a series circuit of an inductance L _c [H] and a resistance R _c [Ω]. In addition, the coaxial cable (40) can be regarded as a system including an inductance L ₂ [H], a resistance R ₂ [Ω], and a capacitance C ₂ [F]. A high frequency excitation voltage V ₁ [V] having an excitation angular frequency ω [rad / s] is supplied from the oscillator (31) to the sensor coil (20) via the additional impedance Za, and _a current I _c [A] flows. .

Then, as shown in FIG. 2, when the obtained voltage V _c is passed through the detection circuit 33 and amplified by the amplifier circuit 34, the relationship between the voltage V _c and the displacement x can be obtained. Further, when this value is linearized through the linearizer 35, the relationship between the output and the displacement x can be obtained.

<Measurement of Q value characteristic of known eddy current type displacement sensor>
FIG. 4 is an explanatory view showing a Q-value characteristic measuring apparatus of a known eddy current type displacement sensor. FIG. 5 is an explanatory view showing the measurement result of the Q value characteristic of a known eddy current type displacement sensor.
Next, measurement of Q value characteristics of a known eddy current type displacement sensor will be described.
In order to measure the Q value characteristic, a measuring apparatus as shown in FIG. 4 is used. As shown in FIG. 4, the measuring device includes a measuring object (10) connected to a spindle (51) of a micrometer (50), a sensor coil 20 installed close to the measuring object (10), An impedance analyzer 70 is provided, and the sensor coil (20) (sensor head) and the impedance analyzer (70) are connected by a coaxial cable (40). The measurement object (10) and the sensor coil (20) (sensor head) are accommodated in an environmental tester (60) for maintaining a constant temperature. In this embodiment, the temperature in the environmental tester (60) is set to 20 ° C.

  In this measuring apparatus, the micrometer (50) adjusts the displacement x (lift-off) between the measurement object (10) and the sensor coil (20), adjusts the displacement between the measurement object 10 and the sensor coil 20, and The Q value characteristic is measured by changing the excitation frequency of the excitation voltage applied to the sensor coil (20).

  Specifically, the excitation frequency of the excitation voltage applied to the sensor coil (20) is swept in the range of 10 kHz to 1.5 MHz. Here, the object to be measured (10) is assumed to be JIS equivalent material: SCM440 defined in “American Petroleum Institute Standard AP1670 for vibration measurement monitoring of rotating machinery”. Then, the Q value characteristic is measured by the impedance analyzer (70) at a position where the displacement x (lift-off) is 0.8 mm, 1.3 mm, and 1.8 mm.

  In this embodiment, several types of sensors (sensor A, sensor B, sensor C) are prepared as known eddy current type displacement sensors, and when the Q value characteristics are measured, the Q value characteristics shown in FIG. 5 are shown. It was. The Q value characteristic of the known eddy current type displacement sensor measured in this way is used when estimating the type and characteristic of an unknown eddy current type displacement sensor to be replaced due to a failure or the like.

<Neural network>
The neural network used for statistical processing in the present invention is an information processing system that simplifies and imitates the structure of a biological cranial nervous system and is mathematically and engineeringly modeled. That is, a simple arithmetic element that models a brain neuron is connected to each other by a weighted link corresponding to a synapse that is a contact point in signal transmission between the brain neurons, thereby constructing a circuit network. This network has adaptability by learning, acceptability of ambiguous information, parallel processing capability, and the like, and can be used for recognition, prediction, control, and the like.

  A general neural network has a structure in which the structure of the cerebral nervous system (neurons) of a living body is simplified and connected in a multilayered network. A synaptic coefficient sequential calculation method that obtains a result, a differential neuron calculation method that reduces the amount of calculation by calculating only the differential component of an input signal, and the like have been proposed.

  At present, various general-purpose processing software to which such a neural network is applied is commercially available. In the embodiment of the present invention, statistical processing is performed using such general-purpose processing software. That is, the Q value characteristic of the eddy current type displacement sensor for estimating the type and characteristics is measured, and the Q value characteristic of the known eddy current type displacement sensor measured in advance is used to determine the maximum Q value and the maximum value thereof. The type of the eddy current type displacement sensor is estimated by performing statistical processing using a neural network with the characteristics of the excitation frequency as explanatory variables and the type of the eddy current type displacement sensor as a target variable.

<Replacement method of eddy current displacement sensor>
FIG. 6 is a flowchart showing a method of replacing the eddy current type displacement sensor. FIG. 7 is an explanatory diagram showing the relationship between lift-off and output characteristics of an eddy current type displacement sensor.
Next, a method for replacing the eddy current type displacement sensor according to the embodiment of the present invention will be described.
Prior to replacing the failed eddy current type displacement sensor, although not shown, a Q value characteristic comprising a maximum value of a known eddy current type displacement sensor and an excitation frequency at the maximum value is measured in advance. In order to replace the faulty eddy current type displacement sensor, as shown in FIG. 6, the faulty eddy current type displacement sensor is connected to the Z / V converter 30 or the impedance analyzer 70 (S1), and the eddy current is supplied from the oscillator 31. An alternating voltage having an excitation frequency of 10 kHz to 1.5 MHz is applied to the mold displacement sensor (S2), and a Q value characteristic consisting of a maximum Q value and an excitation frequency at the maximum value is measured (S3).

  Subsequently, the Q value characteristic of the known eddy current type displacement sensor measured in advance is compared with the Q value characteristic of the faulty eddy current type displacement sensor (S4), and the fault is obtained by statistical processing using a neural network or the like. The type and characteristics of the eddy current type displacement sensor are estimated (S5). Then, the circuit of the Z / V converter (30) is adjusted so as to be suitable for the estimated eddy current type displacement sensor (S6), and the failed Z / V converter (30) is adjusted to the adjusted Z / V conversion. Replace the vessel (30) (S7).

  By using the Q value characteristic of a known eddy current type displacement sensor measured in advance by such a procedure, the maximum value of the Q value and the characteristic of the excitation frequency at the maximum value are explained as explanatory variables. The objective gain is statistically processed by a neural network to estimate the type and characteristics of the faulty eddy current displacement sensor and calibrate the output gain and offset voltage of the new Z / V converter (30). Thus, it is possible to obtain a Z / V converter (30) adapted to the failed eddy current type displacement sensor.

  When the estimated output characteristics of the eddy current displacement sensor, the actual output characteristics of the eddy current displacement sensor, and the output characteristics after calibrating the output gain and the offset voltage are compared, as shown in FIG. As a result of applying the present invention to estimate the type and characteristics of the sensor and applying necessary calibration to the eddy current type displacement sensor whose characteristics were unknown, the output characteristics with respect to lift-off became ideal.

  The present invention is not limited to the embodiment of the invention described above, and when an eddy current type displacement sensor fails, it is possible to select an appropriate maintenance part by accurately estimating the sensor characteristics, and quick failure Of course, as long as the configuration can cope with the above, the configuration is not limited to the illustrated configuration, and various modifications can be made without departing from the gist of the present invention.

DESCRIPTION OF SYMBOLS 10 Measurement object 20 Sensor coil 30 Z / V converter 31 Oscillator 32 Resonance circuit 33 Detection circuit 34 Amplification circuit 35 Linearizer 40 Coaxial cable 50 Micrometer 51 Spindle 60 Environmental tester 70 Impedance analyzer

Claims (4)

Step (S1, S2, S3) of sweeping the excitation frequency of the eddy current displacement sensor to be the characteristic estimation target and measuring the Q value characteristic composed of the maximum Q value and the excitation frequency at the maximum value;
A step (S4) of comparing the Q value characteristic of the eddy current type displacement sensor to be the characteristic estimation object with a previously measured Q value characteristic of the eddy current type displacement sensor;
And a step (S5) of estimating the type and characteristic of the eddy current type displacement sensor to be the characteristic estimation target based on the comparison result, and a method for estimating the characteristic of the eddy current type displacement sensor.   The comparison between the Q value characteristic of the eddy current type displacement sensor to be the characteristic estimation target and the previously measured Q value characteristic of the eddy current type displacement sensor is performed by statistical processing using a neural network. The method for estimating characteristics of an eddy current displacement sensor according to claim 1. When an eddy current displacement sensor fails, the excitation frequency of the failed eddy current displacement sensor is swept to measure the Q value characteristic composed of the maximum Q value and the excitation frequency at the maximum value ( S1, S2, S3)
Comparing the Q value characteristic of the failed eddy current type displacement sensor with the Q value characteristic of a known eddy current type displacement sensor measured in advance (S4);
Estimating the type and characteristics of the failed eddy current displacement sensor based on the comparison result (S5);
In the failed eddy current type displacement sensor, the sensor head having the sensor coil (20) is left, and the impedance-voltage converter (30) adjusted according to the estimated type and characteristics of the eddy current type displacement sensor. And exchanging only (S6, S7), an eddy current type displacement sensor replacement method.   The comparison between the Q value characteristic of the failed eddy current type displacement sensor and the Q value characteristic of a known eddy current type displacement sensor measured in advance is performed by statistical processing using a neural network. 4. A method for replacing the eddy current type displacement sensor according to claim 3.

Images