KR20190072165A - Fault diagnosis system of motor - Google Patents

Abstract

The present invention relates to a motor fault diagnosis system, which can monitor the operation of a motor in real time, catch fault indications, identify fault causes, and provide preventive maintenance. The motor fault diagnosis system comprises: a first monitoring unit for measuring and analyzing the rotation angle of the motor and performing modeling; a second monitoring unit for measuring and analyzing a vibration state of the motor and performing modeling; a third monitoring unit for measuring and analyzing the frequency of the motor and performing modeling; a display unit for receiving data modeled by the first monitoring unit, the second monitoring unit, and the third monitoring unit and outputting the data on a designated screen; and a fault discrimination unit for storing learning data obtained from the first, the second, and the third monitoring unit and outputting a warning phrase through the display when the stored learning data exceeds a predetermined reference value.

Description

{Fault diagnosis system of motor}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a motor fault diagnosis system, and more particularly, to a motor fault diagnosis system capable of detecting a motor fault through 3D modeling of a motor, 3D vibration state, frequency analysis and real time data communication.

In general, motors, especially three-phase induction motors, are used as power sources for machines in industry, and in case of failure, they cause considerable opportunity loss due to discontinued production of machines. These motor failures occur more than 80% due to stator failures such as bearing failure, insulation breakdown, rotor bar rotor failure, and mechanical failure. There have been various methods of monitoring and diagnosing faults and diagnosing faults in real time.

That is, various sensors have been used to detect the failure of a motor, and these sensors can be used to detect stator voltages and currents, air gaps (voids), external magnetic flux density, rotor position and speed, output torque, Were measured.

Therefore, there are thermal diagnosis method, vibration diagnosis method, and electrical diagnosis method using sensors. However, these methods are excellent in detecting mechanical defects of motors but insufficient to detect electric elements.

Accordingly, in recent years, there has been used an electric signal analysis method for early diagnosis of a failure of a motor through analysis of voltage, current, and magnetic flux characteristics without a sensor.

The electric signal analysis method calculates the motor speed in operation from the voltage transformer (MCC: Motor Control Center) and the current transformer (CT: Current Transformer), calculates the slip of the induction motor, Bearing frequency and so on.

That is, in the motor current signature analysis (MCSA) method, a component generated by a defect of a stator, a rotor, and a bearing is modulated to a current frequency by a Fast Fourier Transform (FFT) Finding the fault frequency has the advantage of acquiring only the current signal and diagnosing the motor status remotely without the sensor, but it has a disadvantage in that it is less accurate in diagnosing mechanical faults.

As an example of a conventional technique, a motor current signal analyzing system having an audio amplified speaker output of US Pat. No. 5,689,194 issued to Pramatom Technolol Co., Ltd. monitors and processes the motor current to obtain motor characteristics for various loads The voltage and current are simultaneously sampled using a spectrum analyzer having a plurality of digitally controlled switch capacitor filters under the control of a personal computer, and the motor performance is monitored by causing instantaneous motor power and power factor to be calculated. To this end, a sensor for judging the performance of the motor is provided. The sensor consists of a current transducer, leads and load resistors, and the current transducer provides a variety of current signals proportional to the time of the bearing current flowing through the leads. Various current signals on this lead are applied to both ends of the load resistor to generate various voltages across the resistor proportional to the motor current. This voltage signal is applied to the signal processing system to amplitude and phase demodulate and the signal is applied to the signal conditioner, which can be applied to a variety of filters with frequency cutoffs set to eliminate undesirable spectra associated with frequencies and harmonics that can not be applied to motor current noise analysis Which allow the system to select and suppress a number of different frequency components of the motor current signal, such as motor slip, rotational speed, drive belt speed, and the like. Therefore, it is possible to analyze the fault state of the motor using an audio warning or analyzer.

Pramatom Technologies has released the model "EMPATH system" as a diagnostic device for motors, which is characterized by degradation and deterioration of the rotor bar, electrical and mechanical abnormality of the stator, motor speed and slip, average operating current and voltage , Motor torque, torsional vibration, dynamic load and bearing defects. By monitoring the various faults by analyzing the spectrum of the motor current voltage, it is possible to use the algorithm of the fast Fourier transform for the current conversion in the time domain of the above- have. Using this algorithm, the motor diagnosis is required to condition the motor current signal, which is done by Root Means Squares (RMS) modulation. That is, the diagnosis is represented by various modulation spectra of the RMS-processed current and voltage signals, and fault diagnosis is performed with this spectrum.

Models "EPX3000" and "EPX4000" have been developed by SKF baker as a dynamic motor analyzer as another conventional embodiment and can be commercially obtained commercially. These analyzers can be remotely monitored from the Motor Control Center (MCC) and employ high frequency resolution techniques called Digital Frequency Locked Loop (DFLL) to provide voltage / current spectrum, demodulation And a comprehensive analysis of the current / voltage on the frequency as a domain of harmonics and spectra and harmonics.

As another conventional example, a system using a fault diagnosis technology of an induction motor based on a finite evaluation of a voltage harmonic using a coil sensor of A-Tech Co. is being sold. In this system, the electromagnetic field calculation domain is used to calculate the electromagnetic field torque and the electromagnetic force acting on the rotor and to calculate the right angle component of the magnetic flux density, that is, the time variation of the magnetic flux density Bx and the harmonic spectrum The method consists of generating time-varying spectral of the radial component of density (Bx), spectrum of harmonics, time-variance of output voltage, spectrum of harmonics, and RMS value of output voltage, and comparing the steady state and fault state. This system is known to be somewhat lacking in reliability.

As described above, the related art relates to a technique for calculating the operation speed of the motor using the rotor bar passing frequency or using the eccentricity frequency peak, and the signal processing for extracting the operation speed of the motor is performed from the time domain data Speed Fourier transform for extracting a frequency spectrum or calculating an air gap torque by a voltage and a current applied to the motor, calculating a torque of the motor shaft by inputting electrical and mechanical losses, and then calculating a torque change amount And it is possible to diagnose not only electrical defects but also mechanical defects of the motor by checking the fluctuation of the load.

However, this conventional technique is based on the current signal analysis, which is insufficient in finding the cause of all the failures of the motor during the operation of the motor, resulting in a reduction in the reliability of the previously diagnosed fault diagnosis.

KR 10-1521119 B1 KR 10-1348543 B1

An object of the present invention to solve the problems derived from the prior art is to provide a motor fault diagnosis system which can monitor the operation of a motor in real time to catch a fault, identify the cause of the fault, and perform preventive maintenance .

Meanwhile, the object of the present invention is not limited to the above-mentioned objects, and other objects not mentioned can be clearly understood from the following description.

According to an embodiment of the present invention, the above-mentioned object can be accomplished by a first monitoring unit for measuring, analyzing and modeling a rotation angle of a motor; A second monitoring unit for measuring, analyzing and modeling the vibration state of the motor; A third monitoring unit for measuring and analyzing the frequency of the motor and modeling the frequency; A display unit for receiving data modeled by the first monitoring unit, the second monitoring unit, and the third monitoring unit and outputting the data to a designated screen; And a failure judgment part for storing learning data obtained from the first monitoring part, the second monitoring part and the third monitoring part, and outputting a warning phrase through the display when the stored learning data exceeds a predetermined reference value, The motor fault diagnosis system comprising:

Preferably, the first monitoring unit calculates the motor rotational force by subtracting the frictional force from the driving torque of the motor (wherein the frictional force includes both the frictional force of the rotating shaft, the frictional force inside the motor, and the frictional force of the gearbox connected to the motor The rotational inertia is calculated by dividing the rotational inertia by the motor rotational force to calculate the angular acceleration of the motor, wherein the rotational inertia is a value obtained by rotating the motor Is an inertia mass of the inertia mass measured by an experimental method and is stored in advance in a storage medium).

Preferably, the second monitoring unit includes: a vibration detection module attached to the motor to detect motor vibrations in the axial and radial directions; and a controller for linearly amplifying the signal according to the vibration value detected by the vibration detection unit and outputting And a signal amplification module.

Preferably, the third monitoring unit includes: a current detecting module that detects a current value output from the motor; a current detecting unit that converts a current value sensed by the current detecting unit into a frequency domain signal, A database for storing item-specific parameters of a motor processed by the signal processing module into a frequency band signal; and a parameter setting unit for setting parameters of the motor detected by the signal processing module, And a comparison determination module for calculating a difference with respect to the data and outputting a result for each item of the motor corresponding to the difference.

Advantageously, the current detection module comprises respective R, S and T sensors for sensing current on R, S, T 3.

Preferably, the signal processing module performs a fast Fourier transform on the current sensed by the current detection module to convert the current into a signal of a frequency band, and outputs a parameter for each item of each frequency band.

Preferably, the database stores the normal current, the normal low pressure, the frequency, the connection type, the RPM, the CT ratio, and the PT ratio, which are setting data for each item of each motor.

Preferably, the failure discriminator comprises a neural network module having a plurality of nodes therein and storing learning data obtained as a result of iterative learning with each other, an algorithm pre-stored in the neural network module, And performing a proportional integral derivative control to determine whether the reference value is greater than the reference value.

According to the present invention, it is possible to diagnose a mechanical defect or an electric defect of a motor by monitoring the operation of the motor in real time and capturing a fault symptom. As a result, it is possible to prevent malfunction and burnout of the motor in advance, and also, it is possible to extend the service life of the motor through efficient management of the motor and to reduce enormous management costs due to the reduction of parts.

1 is a block diagram illustrating a motor fault diagnosis system according to an embodiment of the present invention,
2 is a flowchart for explaining the first monitoring unit of the present invention,
3 is a block diagram showing a third monitoring unit of the present invention,
FIG. 4 is a block diagram illustrating a failure detection unit according to the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention and the manner of achieving them will become apparent with reference to the embodiments described in detail below with reference to the accompanying drawings. However, the present invention is not limited to the embodiments described below, but may be embodied in various different forms, and these embodiments are not intended to be exhaustive or to limit the scope of the present invention to the precise form disclosed, It is provided to inform the person completely of the scope of the invention. And the terminology used herein is for the purpose of illustrating embodiments and is not intended to be limiting of the present invention. The singular forms herein include plural forms unless the context clearly dictates otherwise.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. Brief Description of Drawings FIG. 1 is a block diagram of a computer system according to an embodiment of the present invention; FIG. 2 is a block diagram of a computer system according to an embodiment of the present invention; FIG.

1, a motor fault diagnosis system according to an embodiment of the present invention includes a first monitoring unit 100, a second monitoring unit 200, a third monitoring unit 300, and a display unit 500 And a failure determination unit 700. [

First, the first monitoring unit 100 measures, analyzes, and models the rotation angle of the motor M.

Specifically, as shown in FIG. 2, the first monitoring unit 100 calculates a motor average voltage applied to the motor by multiplying the battery voltage by a PWM duty (DUTY) value for modeling. Then, the motor average voltage is divided by the pre-stored motor reactance to calculate the motor drive current. Thereafter, the motor drive current is multiplied by the previously stored torque conversion coefficient to calculate the drive torque of the motor.

When the driving torque of the motor is calculated as described above, the first monitoring unit 100 can calculate the motor rotational force by subtracting the frictional force from the driving torque of the motor M. [ Here, the frictional force includes all the frictional forces acting as resistance to the rotation, including the frictional force of the rotating shaft, the frictional force inside the motor, and the frictional force of the gearbox connected to the motor, which are measured by an experimental method and stored in advance in the storage medium.

Accordingly, the motor angular acceleration can be calculated by dividing the rotational inertia by the motor rotational force, and the position (rotational angle) of the motor can be calculated from the angular velocity of the motor. At this time, the angular velocity of the motor may be integrated with respect to time to calculate the angular velocity of the motor, and the calculated angular velocity of the motor may be integrated with respect to time to calculate the rotational angle of the motor. Here, the rotational inertia is an inertia mass when the motor M is rotated, is measured by an empirical method and stored in advance in the storage medium.

Next, the second monitoring unit 200 measures, analyzes, and models the vibration state of the motor M.

The second monitoring unit 200 includes a vibration detecting module attached to the motor M and detecting vibration of the motor M in the axial and radial directions, And then outputting the amplified signal.

The vibration detecting module detects a motor vibration in the axial and radial directions by attaching a piezoelectric film sensor or an acceleration sensor to the motor.

The signal amplification module linearly amplifies a signal according to the vibration value detected by the vibration detection module by an operational amplifier.

Next, the third monitoring unit 300 measures and analyzes the frequency of the motor M and models it.

3, the third monitoring unit 300 includes a current detecting module 320 for detecting a current value output from the motor M, and a current detecting unit 320 for detecting a current value output from the current detecting unit 320. [ A signal processing module 340 for converting the sensed current value into a frequency domain signal and detecting a parameter for each item of the motor for each frequency band signal, A difference between the parameter of the motor M detected by the signal processing module 340 and the motor data preset in the database 360 and a difference corresponding to the difference And a comparison determination module 380 for outputting the results of the respective items of the motor M to be output.

The current detection module 320 constitutes each of the R, S and T sensors for detecting the currents on the R, S, and T 3, and senses a current output according to the phase difference.

More specifically, when voltage and current are applied to the motor M, the motor M is operated without load or as a load. In this case, the current varies depending on the load, and if there is a mechanical or electrical failure of the motor M, the current changes finely. Therefore, it is possible to detect a minute current change.

The signal processing module 340 converts the current value detected by the current detection module 320 into a frequency domain signal and detects a parameter for each item of the motor M for each frequency band signal.

Specifically, the signal processing module 340 performs Fast Fourier Transform (FFT) on the current sensed by the current detection module 320 to convert the current into a signal of a frequency band, Outputs parameters for each item. At this time, it is preferable to perform the fast Fourier transform by detecting the result by applying a general equation.

Here, the parameters for each item in accordance with the frequency band include defects due to fixed unloading defects of the motor M, unbalance / gearbox / transmission defects, rotor defects, stator defects, bearing defects, and other problems. That is, it may include parameters for defects in various facilities such as a pump, a fan, a compressor, a conveyor, a blower, an air conditioner, and a press, which are driving facilities connected to the motor M.

In the database 360, the normal current, the normal low voltage, the frequency, the connection type, the RPM, the CT ratio, and the PT ratio, which are setting data for each item of each motor M, are stored. In the signal processing module 340, data on defects due to fixing failure of the motor M according to the frequency band, unbalance / gearbox / transmission defects, rotor defects, stator defects, bearing defects, Base and store it. At this time, each parameter of the motor M that stores the data base in the signal processing module 340 can be continuously updated through modeling learning of the motor.

The comparison determination module 380 compares the data of the ideal theoretical value of the motor M previously stored in the database 360 with the parameter data obtained by measuring the actual motor M. [ In other words, the measured data of the motor M is compared with the data of the theoretical motor M, and the difference is calculated. Then, it is judged whether or not there is a defect corresponding to the value according to the difference. Here, when the difference between the theoretical motor data and the actual measured motor data is equal to or larger than the predetermined value, it is determined that the corresponding item is defective. On the other hand, when the difference between the theoretical motor data and the actual measured motor data is less than the predetermined value, the motor M is considered to be in the normal range, and the modeling learning of the motor is performed again to monitor the motor M continuously.

Next, the display unit 500 receives the data modeled by the first monitoring unit 100, the second monitoring unit 200, and the third monitoring unit 300, and outputs the data to a designated screen.

Next, the failure discrimination unit 700 stores the learning data obtained from the first monitoring unit 100, the second monitoring unit 200, and the third monitoring unit 300, And outputs a warning message through the display unit 500 when the predetermined reference value is exceeded.

4, the failure determination unit 700 includes a neural network module 720 having a plurality of nodes 724 therein and storing learning data obtained as a result of iterative learning with each other, And a determination module 740 that processes the algorithm stored in the neural network module 720 and performs proportional integral derivative control to determine whether the learning data exceeds the reference value.

The neural network module 720 learns in both directions between the nodes 724.

The determination module 740 processes the neural network module 720 in both directions. That is, the neural network module 720 learns the feed forward direction and / or the feedback word direction, that is, both directions, between the internal nodes 724 to determine the reference value of the multivariable system, and stores the learning data obtained by the learning.

According to the embodiments of the present invention described so far, the following effects can be expected.

First, even if a slight error occurs in the performance deterioration of the motor or the contact resistance, the fault can be diagnosed.

Second, by using the voltage and current of the motor, it is possible to diagnose mechanical faults or electrical faults of motors as well as to diagnose mechanical faults such as rotating equipment connected to the motors, thereby diagnosing the whole facilities and managing the life span.

Third, by monitoring the operation of the motor in real time, it can improve the productivity by reducing the abnormal fault, and it can reduce the huge facility management cost due to the extension of the life span and parts reduction through efficient management of facility assets.

The foregoing has outlined rather broadly the features and technical advantages of the present invention in order that the claims of the invention to be described below may be better understood. It will be understood by those skilled in the art that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. It is therefore to be understood that the above-described embodiments are illustrative in all aspects and not restrictive. The scope of the present invention is defined by the appended claims rather than the foregoing detailed description, and all changes or modifications derived from the appended claims and their equivalents should be construed as being included within the scope of the present invention.

M: Motor
100: first monitoring section
200: second monitoring unit
300: Third monitoring unit
320: current detection module
340: Signal processing module
360: Database
380: comparison judgment module
500:
700: Fault discrimination unit
720: Neural network module
724: Node
740:

Claims (8)

A first monitoring unit for measuring and analyzing the rotation angle of the motor and modeling the rotation angle;
A second monitoring unit for measuring, analyzing and modeling the vibration state of the motor;
A third monitoring unit for measuring and analyzing the frequency of the motor and modeling the frequency;
A display unit for receiving data modeled by the first monitoring unit, the second monitoring unit, and the third monitoring unit and outputting the data to a designated screen; And
A failure detection unit for storing learning data obtained from the first monitoring unit, the second monitoring unit, and the third monitoring unit, and outputting a warning message through the display when the stored learning data exceeds a predetermined reference value;
The motor fault diagnosis system comprising: The method according to claim 1,
Wherein the first monitoring unit
Wherein the frictional force is calculated by subtracting the frictional force from the driving torque of the motor, wherein the frictional force includes all of the frictional force of the rotating shaft, the frictional force inside the motor, and the frictional force of the gearbox connected to the motor, Which is measured by an empirical method and stored in advance on a storage medium)
The angular acceleration of the motor is calculated by dividing the rotational inertia by the motor rotational force (here, the rotational inertia is the inertial mass at the time of rotating the motor, which is measured by an empirical method and stored in advance in the storage medium)
The motor fault diagnosis system comprising: The method according to claim 1,
Wherein the second monitoring unit comprises:
A vibration detection module attached to the motor to detect motor vibrations in the axial and radial directions,
And a signal amplification module for linearly amplifying a signal according to the vibration value detected by the vibration detection unit and outputting the amplified signal. The method according to claim 1,
Wherein the third monitoring unit
A current detection module for sensing a current value output from the motor,
A signal processing module for converting the current value sensed by the current detection unit into a frequency domain signal and detecting a parameter for each item of the motor for each frequency band signal;
A database for storing item-specific parameters of a motor processed by the signal processing module into a frequency band signal;
And a comparison determination module for calculating a difference between the parameter of the motor detected by the signal processing module and the motor data preset in the database and outputting the result of each item of the motor corresponding to the difference, Fault diagnosis system. 5. The method of claim 4,
Wherein the current detection module comprises respective R, S and T sensors for sensing currents on R, S, T 3. 5. The method of claim 4,
Wherein the signal processing module performs fast Fourier transform on the current sensed by the current detection module to convert the current into a signal of a frequency band and outputs a parameter for each item of each frequency band. 5. The method of claim 4,
Wherein the data stored in the database includes a normal current, a normal low voltage, a frequency, a connection type, an RPM, a CT ratio, and a PT ratio. The method according to claim 1,
Wherein the failure discriminator comprises:
A neural network module in which a plurality of nodes are provided and which store learning data obtained as a result of iterative learning,
And a discrimination module for processing the algorithm stored in the neural network module and performing proportional integral derivative control to determine whether the learning data exceeds the reference value.

Images