KR20230083630A - Anomaly diagnosis system basesd on unique for steel making bearing - Google Patents

Abstract

The present invention relates to an abnormality diagnosis system based on a specific sound of a steel tilting bearing and, more specifically, to a system for diagnosing abnormality based on a specific sound caused by a rotation sound of a low-speed bearing in a steel tilting bearing. According to the present invention, the abnormality diagnosis system based on the specific sound of the steel tilting bearing includes: a sensor collecting the generated sound by being installed in the tilting bearing; a system server collecting acoustic data; and an integrated server diagnosing an acoustic waveform and a pattern according to an angle (-180 to 180 degrees) through an ADC by converting a PLC current signal generated by the tilting bearing into a voltage signal and amplifying the voltage signal. The abnormality diagnosis system based on the specific sound of the steel tilting bearing can improve reliability of a prediction system by diagnosing the abnormality through acoustic analysis before a problem of the tilting bearing occurs as the abnormality diagnosis system analyzes the rotation sound of the low-speed bearing generated when tilting.

Description

Anomaly diagnosis system bases on unique for steel making bearing}

The present invention relates to a system for diagnosing an abnormality based on a specific sound of a steel-made tilt-rolling bearing, and more particularly, to a system for diagnosing an abnormality based on a specific sound generated from a rotational sound of a low-speed bearing of a steel-made tilt-rolled bearing.

The converter tilting device is a device that taps molten steel by tilting the converter, which is a facility that refines molten iron into steel at a steel mill, and is a key facility in a steelmaking plant that requires advanced design and manufacturing technology.

In general, in the working process of a converter for refining molten iron into molten steel, as shown in Figs. The molten iron is charged into the converter 1 using the charging crane 9, and oxygen is blown into the converter 1 through the oxygen lance 100 for a certain period of time, reacts with the molten iron in the converter 1, and converts the converter 1 ) in molten steel refining work. Then, for the tapping work of the refined molten steel, the converter 1 is repeatedly performed by the converter tilting device 4 installed on the side, and for the tilting work of the converter 1, the converter 1 Both sides are supported by tilting bearings (7).

In the process of tilting the converter 1, wear of the bearing 7 for fixation and tilting of the converter and whether or not there is an abnormality in the bearing 7 is performed by opening the converter bearing cover 13 using the furnace repair work period, The roller 10a of the bearing 7 and the retainer are checked for confirmation by inspection work. In the related art, inspection of the tilting bearing 7 is performed using an endoscope for inspecting the bearing. For reference, in the inspection work, the operator (converter master) listens to the sound of the bearing through a stethoscope and determines whether there is an abnormality before repair. In order to check the local damage and wear of the tilting bearing (7) of the converter (1) using an endoscope device for bearing inspection, the grease inside the tilting bearing (7) must be completely cleaned, and the wear and tear of the tilting bearing (7) Check the damaged area.

However, in the conventional converter tilting bearing inspection device as described above, it is difficult to accurately inspect due to the large size of the tilting bearing and the inability to tilt the converter during furnace maintenance work, and the operator's psychological burden is increased by checking the tilting bearing every time the furnace is repaired Above all, not only is it difficult to accurately inspect the tilting bearing, but in the conventional tilting bearing inspection method, a sensor for inspecting bearing vibration is assembled outside the tilting bearing housing to check the vibration of the tilting bearing, but the inside of the converter bearing Even if the bearing sensor is attached by installing the housing bush and the guide bush, there is a problem that it is impossible to check the vibration and the displacement of the converter due to thermal expansion of the converter generated during the converter blowing operation cannot be checked.

Patent Publication 2004-0036009 (2004.04.30)

The present invention was conceived to solve the above-mentioned problems, and by comparing the tilt signal (steel inclination angle) and the unique sound signal to diagnose the sound signal of the tilt bearing in real time, the problems of the steel tilt motion bearing are prevented in advance, and the converter An object of the present invention is to provide a system for diagnosing an abnormal sound based on a specific sound of a steel-making tilt-roll bearing that can reduce steel-making loss due to a failure of the steel-making.

In addition, when problems such as natural deterioration, poor maintenance, and foreign substances during use (bearing contact) occur, the purpose is to diagnose through acoustic analysis before problems occur in the tilting bearing based on the unique sound generated from the low-speed bearing rotation sound of the steelmaking tilting bearing. there is.

The system for diagnosing abnormalities based on unique acoustics of steel-making tilting bearings according to the present invention includes a sensor installed in a tilting bearing to collect generated sounds; a system server that collects sound data from the sensor; And by receiving the sound data from the system server, converting and amplifying the PLC current signal generated from the tilting bearing into a voltage signal, and then using the ADC to generate sound waveforms and patterns according to angles (-180 degrees to 180 degrees). A comprehensive server for diagnosing;

Preferably, the sensors are disposed on both sides of, or in front of, the drive shaft and the driven shaft of the tilting bearing.

The system server constantly monitors and records the sound in the forward and reverse directions of the tilting bearing and converts it into a database.

The comprehensive server processes the pattern of sound generated from the tilting bearing in any one of time domain, frequency domain, Bark Scale, LPCC (Linear Prediction Cepstral Coefficient), MFCC (Mel Frequency Cepstral Coefficient), Frequency_Bar Graph, and RLCR, Acoustic data processing module that analyzes the shape and size, receives the PLC current signal received wired or wirelessly from the tilting bearing and the sound signal of the system server, converts the tilting signal current into a voltage signal, and A diagnostic module for amplifying and diagnosing the acoustic waveform and pattern according to the angle (-180 to 180 degrees) through the ADC.

The comprehensive server processes the pattern of sound generated from the tilting bearing in any one of time domain, frequency domain, Bark Scale, LPCC (Linear Prediction Cepstral Coefficient), MFCC (Mel Frequency Cepstral Coefficient), Frequency_Bar Graph, and RLCR, Acoustic data processing module that analyzes the shape and size, receives the PLC current signal received wired or wirelessly from the tilting bearing and the sound signal of the system server, converts the tilting signal current into a voltage signal, and A diagnostic module for amplifying and diagnosing the acoustic waveform and pattern according to the angle (-180 to 180 degrees) through the ADC.

The integrated server processes the sound signal from the rotation of the steel-making tilt bearing in one of time domain, frequency domain, Bark Scale, LPCC (Linear Prediction Cepstral Coefficient), MFCC (Mel Frequency Cepstral Coefficient), Frequency_Bar Graph, RLCR, The PLC current signal is converted and amplified into a voltage signal, and the acoustic waveform and pattern according to the angle (-180 to 180 degrees) are diagnosed through the ADC.

The system according to an embodiment of the present invention analyzes the sound in which the sound of the rotation sound is heard differently when problems such as natural deterioration, poor maintenance, and foreign matter during use (bearing contact) occur as a cause of failure of low-speed bearing rotation noise generated during tilting. There is an effect of improving the reliability of the predictive system by diagnosing problems of the tilting bearing through acoustic analysis before they occur (prediction).

In addition, it is possible to improve the status of the domestic steel industry by reducing costs through economical repair, maintenance, and replacement of steel-making tilting bearings, and by developing a unique acoustic-based anomaly prediction system.

1 is a general converter molten steel refining work diagram,
2 is a generally used converter scrap iron charging diagram,
3 is a generally used converter molten steel charging diagram,
4 is a cross-sectional view of a generally used converter bearing.
5 is a view showing a system for diagnosing an abnormal sound based on a specific sound of a steel-making tilt-roll bearing according to an embodiment of the present invention;
6 is a view showing the sensor installation position of the tilting bearing of the unique sound-based abnormal diagnosis system for the steelmaking tilting bearing according to an embodiment of the present invention.
7 is a diagram showing a monitor configuration of a system according to an embodiment of the present invention.
8 is a diagram showing the analysis of acoustic data of a tilting bearing of the system according to this embodiment.
9 is a screen showing analysis of tilting signal data input to the acoustic sensor of the system according to the present embodiment.
10 is a diagram showing predicted normal data (top) and abnormal data (bottom) of the system according to the present embodiment.
11 is a diagram in which the average value by LPCC of the system according to the present embodiment is displayed in red.
12 is a diagram showing the waveform of the RLCR of the system according to the present embodiment.
13 is a diagram illustrating monitoring of a tilting signal, a driving shaft, and a tilting bearing signal of a driven shaft in the system according to the present embodiment.
14 is a view showing a monitoring screen in a state (12 mA) without tilt signal of the system according to the present embodiment.
15 shows the signal and the LPCC monitoring screen while the tilting signal changes to high, middle, and low.
16 shows a data signal and an LPCC monitoring screen that are changed as the tilt signal falls.
17 is a data signal, RLCR waveform, average value, and LPCC monitoring screen that change for the entire period of the tilt signal.
18 is a LPCC waveform of a tilting bearing with flaws (data investigated in a laboratory).
19 is an angle change monitoring screen of a tilt signal from normal (12 mA, 0 degree) to phase (20 mA, 90 degree).
20 is an LPCC monitoring screen in a state in which there is no tilting signal.
21 is a monitoring screen in which the data of the tilting signal (No. 3) is analyzed by 5 algorithms.
22 is a monitoring screen showing changes in 5 algorithms caused by external noise (No. 3).
23 is a monitoring screen showing changes in five algorithms based on a sound signal and a frequency spectrogram when there is a lot of external noise (No. 3).
24 is a monitoring screen on which a change in an incoming sound signal in which an external noise (No. 3) and a tilt signal are mixed together can be seen.
25 shows a monitoring screen of a tilt signal coming into a sensor in real time.
26 shows a monitoring screen in which the tilting signal (No. 3) data is analyzed by 5 algorithms.
27 shows a monitoring screen of signals coming into four sensors and tilting signals in real time.
28 is an angle change monitoring screen while the tilting signal changes from phase (20mA) to steady state (12mA).
29 is a monitoring screen of a waveform (2) in the time domain in a state (2) with a tilt signal.
30 is an angle change monitoring screen of a tilt signal from normal (12 mA, 0 degrees) to phase (20 mA, 90 degrees).
31 shows a graph showing the waveform of the LPCC when there is no tilt signal (No. 3).

Specific structural or functional descriptions presented in the embodiments of the present invention are merely exemplified for the purpose of explaining embodiments according to the concept of the present invention, and embodiments according to the concept of the present invention may be implemented in various forms. In addition, it should not be construed as being limited to the embodiments described in this specification, but should be understood to include all modifications, equivalents, or substitutes included in the spirit and scope of the present invention.

Meanwhile, in the present invention, terms such as first and/or second may be used to describe various elements, but the elements are not limited to the above terms. The above terms are used only for the purpose of distinguishing one component from other components, for example, within a range not departing from the scope of rights according to the concept of the present invention, a first component may be referred to as a second component, Similarly, the second component may also be referred to as the first component.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In describing the embodiments of the present invention, if it is determined that a description of a related known function or configuration may unnecessarily obscure the subject matter of the present invention, the description is omitted.

The system for diagnosing an abnormality based on the unique sound of a steel-making tilt-roll bearing according to an embodiment of the present invention

According to an embodiment of the present invention, the system for diagnosing an abnormality based on the specific sound of a steel-making tilting bearing is amplifying sound generated from the specific sound vibration of a tilting bearing in a steelmaking converter with a sensor (listening rod sensor, acoustic sensor, vibration sensor, MEMS sensor) and amplifying sound The signal is converted into digital and transmitted to the diagnostic device, and the received digital signal is stored remotely as an acoustic signal tailored to the tilting signal (tilt angle of steelmaking) and a general sound signal without tilting signal, and through sound pattern analysis When a change in pattern is detected by diagnosing fluctuations in real time, it is related to a unique sound-based abnormality diagnosis system technology that can reduce steelmaking loss due to failure of the tilting roller bearing in the converter by checking and taking action on the problem of the tilting roller bearing in steelmaking. .

5 is a diagram showing a system for diagnosing an abnormality based on specific sound of a steel-making tilt-roll bearing according to an embodiment of the present invention. As shown in FIG. 5 , the system 10 for diagnosing an abnormality based on the specific sound of a steel-making tilting bearing includes a sensor 410, a system server 420, a comprehensive server 430, and a monitor 440. In FIG. 5, there are cases in which workers (converter masters) directly touch and listen to the tilting bearing with a stethoscope. The heat of the converter reaches about 3,400 degrees and the worker listens while sweating.

The sensor 410 is installed on the tilting bearing to collect sound generated during tilting.

6 is a view showing the sensor installation position of the tilting bearing of the unique sound-based abnormal diagnosis system for the steelmaking tilting bearing according to an embodiment of the present invention. As shown in FIG. 5(6), the sensor according to the present embodiment is disposed on both sides or in front and behind the driving shaft and the driven shaft of the tilting bearing to collect sound generated during tilting. The sensor for performing this function includes a sound collector 411 that amplifies the low-speed bearing rotation sound generated during tilting, converts it into a digital signal through a filter, and transmits it to the system server to the outside. These sound collectors convert analog signals into digital signals through data sensing, amplification, filtering, ADC, sampling, and DSP.

The sensor according to this embodiment uses a capacitor microphone and a dynamic microphone that are very strong against signals. Here, the signal and noise signal are separated and amplified by the Instrumentation Amp. The amplified signal passes through a filter to remove unnecessary ambient noise frequencies, and the sensor transmits the signal through the ADC and MPU. The signal received here is analyzed in the time domain (RLCR) and frequency domain (Energy Square analysis) in the program to find the bearing signal as a spectrogram for each color.

The sensor according to this embodiment is a sensor comprising the AE Sensor3 (acoustic sensor 3) circuit shown in the same block diagram as the acoustic collector 411 of FIG. 5 . Acoustic sensors are used as four types of sensors (hearing rod sensor, acoustic sensor, vibration sensor, MEMS sensor), but a combination of sensors is used depending on the environment where the tilting bearing is located. Here, the sensor uses a MEMS sensor and a sound rod sensor, and uses the characteristics of the MEMS sensor and the sound rod sensor.

The listening rod sensor according to this embodiment is a sensor made by using a capacitor microphone in addition to the characteristics of the existing structural listening rod. That is, the listening rod sensor is a sensor to which a capacitor microphone is applied. The characteristic of the hearing rod sensor of this embodiment is that external noise is completely removed and only the sound part is heard.

This sensor uses a MEMS vibration sensor. The characteristics of the MEMS vibration sensor are that there is no hysteresis phenomenon, and sensitivity and linearity are excellent. The bearing signal detected by the MEMS vibration sensor is amplified and converted into an electrical signal through a filter, and the signal is ADC converted and transmitted as data.

The signals received here are analyzed by time domain (RLCR) and frequency domain (Energy Square analysis) in the comprehensive server to find bearing signals with spectrograms for each color, and also analyzed by Bark scale and MFCC (Mel Frequency Cepstral Coefficients) to , and displays the frequency in real time with LPCC (Linear Prediction Cepstral Coefficients).

The entire system that detects the signal of the tilting bearing with the above-mentioned sensor, system and program has been explained. Here, the spectrogram analysis (patented) method of the frequency domain for the filter is a method of analyzing data by setting the frequency band below in three steps.

- Low spectral energy frequency:0-20KHz

- Medium spectral energy frequency:20-50KHz

- High spectral energy frequency: 50-100KHz was set. This setting can be changed according to the environment in the program (refer to the formula below).

[formula]

The system server 420 is a component that collects acoustic data from sensors. This system server constantly monitors and records the sound of the forward and reverse directions of the tilting bearing and converts it into a database. On the other hand, the integrated server 430 receives the PLC current signal of the tilting bearing generated in the system of the tilting bearing by wire or wirelessly. For reference, the tilting signal of the tilting bearing is not generated from the system server, but the tilting signal generated from the tilting bearing system is received and applied by the comprehensive server.

The integrated server 430 is a component for receiving sound data from the system server and analyzing the pattern of sound generated from the tilting bearing. This comprehensive server includes a sound data processing module that analyzes the pattern of sound generated from the tilting bearing in the time domain, frequency domain, shape, and size. In addition, the integrated server receives the tilting signal current generated by the tilting bearing system itself wired or wirelessly, receives the sound signal from the system server, converts the PLC current into voltage, and converts the analog to digital and converts the tilting angle through an amplifier. It includes a diagnostic module (I-V-ADC) that displays and diagnoses to predict the failure of the converter in advance by finding problems in the tilting bearing through acoustic waveforms and patterns. This diagnosis module converts and amplifies the PLC current signal into a voltage signal, finds problems in the tilting bearing through the sound waveform and pattern according to the angle (-180~0 degree ~ 180 degree) through the ADC, and Predict circuit failure.

The monitor 440 receives the acoustic data and tilting angle display data processed by the comprehensive server and monitors whether or not an abnormal diagnosis has been made. In addition, if there is an abnormality, the converter specialist listens to the sound of the tilting bearing through the speaker of FIG. 5 and reconfirms whether the bearing has a problem.

The steel-making tilt-rolling bearing specific sound-based abnormality diagnosis system according to this embodiment is a sensing technology, and the sound by rotation of the steel-making tilt-rolling bearing is measured in the time domain, frequency domain, Bark Scale, LPCC (Linear Prediction Cepstral Coefficient), MFCC (Mel Frequency Cepstral Coefficient), Frequency_Bar Graph, and RLCR, a comprehensive analysis was applied, where the PLC current signal was converted and amplified into a voltage signal, and the sound according to the angle (-180 degrees ~ 0 degrees ~ 180 degrees) through the ADC. By directly listening to the acoustic sound with waveform and pattern remotely, it is possible to find the problem of the tilting bearing and predict the failure of the converter in advance.

In addition, the system according to the present embodiment uses a sound signal standardization technology for rotational sound of steel-making tilting bearings and a data transmission method using wired and wireless networks. Here, real-time transmission and communication technology by embedded program is extended to remote diagnosis and monitoring system. These operations are described as follows.

First, the unique sound-based abnormality diagnosis system for steel-making tilt-rolling bearings collects actual data of steelmaking tilt-rolling bearings through sensors disposed on both sides or in the front and rear of the drive shaft and driven shaft of the tilt motion bearing. The system server collects actual data of steel-made tilt-roll bearings, classifies steel-made tilt-roll bearing data, and secures a database.

Next, in the comprehensive server, the pattern of sound generated in the tilting bearing through the separation of the noise and the signal of the steelmaking tilting bearing including the tilting signal current is analyzed in the time domain, frequency domain, Bark Scale, LPCC (Linear Prediction Cepstral Coefficient), MFCC ( Mel Frequency Cepstral Coefficient), Frequency_Bar Graph, and RLCR to analyze the shape and size in the acoustic data processing module. The diagnostic module (I-V-ADC) of the comprehensive server receives the Kyung-Dong signal current (PLC current signal generated from the Kyung-Dong bearing itself) received from the Kyung-Dong bearing system by wire or wirelessly and the acoustic signal from the system server. (PLC current signal) is converted into a voltage signal and amplified to diagnose the acoustic waveform and pattern according to the angle (-180 degree ~ 0 degree ~ 180 degree) through ADC.

The monitor according to this embodiment performs monitoring by displaying data information of the system according to month, date, and time through the embedded S/W of the steel-making tilting bearing sound-based abnormal diagnosis system (or steelmaking tilting bearing sound-based prognosis system). do. Through this monitor configuration, it is possible to control diagnosis by setting a diagnosis level for each system. In addition, the monitor performs the function of monthly data storage and monitoring system of each system, and performs the function of data collection and analysis of each system.

The monitor configuration according to this embodiment can perform real-time monitoring and sound output of the collected system and implement a perfect signal through high-resolution sampling. In addition, it performs replay and replay functions of acoustic diagnosis data and data storage. In addition, it controls the measurement data time series analysis / frequency series analysis function and realizes data expansion/reduction and editing functions to diagnose and display abnormal data. The monitor according to the present embodiment can control the overall system and perform automatic DB configuration and DB input/output design using database functions and MS-SQL.

7 is a diagram showing a monitor configuration of a system according to an embodiment of the present invention. As shown in FIG. 7, as a screen for monitoring the tilting sound signal coming into the sensor in real time, the monitor according to this embodiment is composed of a total of four screens (here, screens can be added depending on the sensor) It is a real-time screen being analyzed. Waveform analysis for each sensor (upper end of drive shaft) and Bark Scale (lower end of driven shaft) are added to monitor drive and driven axes in real time, and abnormal signals are displayed as events.

8 is a diagram showing the analysis of acoustic data of a tilting bearing of the system according to this embodiment. As shown in FIG. 8, when a signal is input to the sensor, data is displayed as a time waveform in real time on the upper side as shown on the screen on the left, and data is displayed in real time Bark Scale on the lower side. The monitor displays a total of 8 screens in real time, a screen by 2 drive shaft sensors and data by 2 driven shaft sensors (screens can be added depending on the sensor). In FIG. 7, the upper part of the sensor screen ① is a real-time waveform, and the lower part is a screen displaying data in Bark Scale.

9 is a screen showing analysis of the dynamic sound signal data input to the acoustic sensor of the system according to the present embodiment. Screen “1” of FIG. 9 displays the frequency of the signal as a color (spectrum) when the dynamic sound signal data is received. The frequency is divided into low, middle, and high frequency bands and displayed in color according to the formula below. Here, red (spectrum) indicates the maximum energy of each band frequency and displays it as a color, and low energy is classified as a blue or white spectrum.

Screen “2” of FIG. 9 displays the frequency as a bar (bar graph) on the scale of the Bark Scale (acoustic scientist in Germany in 1961) when the dynamic acoustic signal data is received. The Bark Scale categorizes the loudness on a scale of 1 to 24. This corresponds to the 24 critical bands of hearing, and the frequencies set by Bark are 20, 100, 200 ~9500, 12000, 15500 (Hz), with a total of 24 frequencies (the frequencies set by Bark). It is indicated by a bar. In this screen, 24 bars are reduced to 14 and displayed. For reference, you can check which frequency band data is coming in by looking at the height of the bar.

Screen “3” of FIG. 9 displays the frequency of the signal as a bar (bar graph) when the dynamic sound signal data is received. Here, Bar (bar graph) displays the pure bar (bar graph) of the frequency to which the algorithm of the “1” screen description is applied, and checks which frequency appears high. ) Here, the algorithm of “1” screen description displays the frequency of the signal as a color (spectrum) when the dynamic signal data is received. The frequency is divided into low, middle, and high frequency bands and displayed in color according to the formula below. Here, red (spectrum) indicates the maximum energy of each band frequency and is displayed as a color, and low energy is classified as a blue or white spectrum.

Screen “4” of FIG. 9 displays the frequency of the signal by LPCC (Linear Prediction Cepstral Coefficients) when the dynamic acoustic signal data is received.

LPCC (Linear Prediction Cepstral Coefficients) is called a linear prediction filter coefficient, and based on past samples, predicts the current value of real number value time sequence (time series) x to generate FIR (Finite impulse response filter, that is, input signal This means that filtering is performed with only constant values of , and the series of data is estimated by obtaining the coefficients of the p-order linear predictor, which is a filter.

The expression is expressed as [a, g] = 1pc(x, p). That is, the predicted signal and the incoming tilting acoustic bearing signal are compared by floating 256 samples.

10 is a diagram showing predicted normal data (top) and abnormal data (bottom) of the system according to the present embodiment. Referring to FIG. 10, the normal and abnormal data predicted by LPCC are clearly distinguished.

Here, it is divided into normal (upper waveform) and abnormal (lower two waveforms) predicted by LPCC.

11 is a diagram in which the average value by LPCC of the system according to the present embodiment is displayed in red. As shown in FIG. 11, the red line in the middle is the average value of LPCC, and displays a simple reference line that can confirm the normal state of the current tilting bearing.

Screen “5” of FIG. 9 displays the frequency of the signal as RLCR (Relative Level Crossing Rate) when the dynamic signal data is received. RLCR (Relative Level Crossing Rate) is set based on the incoming signal when the normal signal (normal operation) comes in when the Kyungdong Acoustic data signal comes in. , it displays how much the signal crosses with the set standard (adjusted on the program screen) as a numerical value (Level Crossing Rate). So, are there any current problems with numerical values? You can also check if there is no The formula is:

[formula]

12 is a diagram showing the waveform of the RLCR of the system according to the present embodiment. In Fig. 12, it is indicated from the upper left (4.5 sec) to the upper right (8.9 sec). That is, during 8.9-4.5=4.4 seconds, RLCR shows the number 3.9 at the right end of (a) screen, which indicates that this number crossed the relative level crossing rate 3.9 times. On the other hand, when it is abnormal, (b) 89 and (c) 43 crossover rates are shown. As such, the normal value is small, but the abnormal value is displayed as a high value.

13 is a view showing the monitoring of the tilting signal, the driving shaft, and the tilting bearing sound signal of the driven shaft of the system according to the present embodiment. As shown in FIG. 13, looking at the monitoring screen of the tilting signal coming into the sensor in real time, two sensors were connected to the driving shaft and the driven shaft to monitor (1, 2) with the time domain waveform and Bark Scale. At this time, when a sensor is added, a screen is also added. Analysis is performed by frequency spectrum, Bark Scale, Bar Graph, LPCC, and RLCR. And the current signal of the tilting signal was displayed as before (20mA), stop (12mA), and reverse rotation (4mA) of the tilting signal as shown in (3) by the hardware system through voltage and ADC. This makes it possible to accurately analyze the condition of the tilting bearing by the tilting signal.

The system according to this embodiment can accurately identify the problem of the tilt bearing by analyzing the sound data by the tilt signal and the LPCC.

14 is a view showing a monitoring screen in a state (12 mA) without tilt signal of the system according to the present embodiment. If the tilting signal is normal (12mA, No. 3), you can see that the data signal of the tilting bearing of the drive shaft and the driven shaft and the average signal of the LPCC (red line in the middle) are constantly displayed.

15 shows the signal and the LPCC monitoring screen while the tilting signal changes to high, middle, and low. When the tilting signal is down (-)90 degrees, normal 0 degrees, and up 90 degrees, you can see the graph where the position of the tilting bearing changes from up (over 90 degrees) to mid (0 degrees) down (over (-) 90 degrees). can (3 times). Here, the waveform of LPCC changes when moving from top to bottom (external noise and sound of tilt signal), but it can be seen that the average signal of LPCC (red line in the middle) is constantly displayed. In other words, it means that there is no problem with the tilting bearing.

16 shows a data signal and an LPCC monitoring screen that are changed as the tilt signal falls. You can see the data signal and the change value of LPCC shown as the tilt signal falls. Here, the large LPCC means that the tilt signal is constantly changed.

17 is a data signal, RLCR waveform, average value, and LPCC monitoring screen that change for the entire period of the tilt signal. It is monitoring that the angle of the tilt signal continuously changes (all sections). Here, '4' represents the data signal and value of RLCR, and represents the average value of about 8 to 10. This means that the LPCC is kept constant and there is no problem with the tilting bearing.

18 is a LPCC waveform of a tilting bearing with flaws (data investigated in a laboratory). The LPCC waveform of the wrong signal (bearing problem, bearing defect in the laboratory) was displayed again by the tilt signal. Summarizing the contents described above, when there is no problem, the RLCR value maintains the average value (8 to 10), and the LPCC average curve is also displayed constantly, so when a graph similar to FIG. 18 is displayed, there is a problem with the tilting bearing can predict

As shown in FIG. 13, the system for monitoring the tilting signal and the signals of the tilting bearings of the drive shaft and the driven shaft according to this embodiment connects two sensors each to the drive shaft and the driven shaft and monitors the waveform in the time domain and the Bark Scale. (1, 2). Analysis is performed by frequency spectrum, Bark Scale, Bar Graph, LPCC, and RLCR. And the current signal of the tilting signal was displayed as before (20mA), stop (12mA), and reverse rotation (4mA) of the tilting signal as shown in (3) by the hardware system through voltage and ADC. This makes it possible to accurately analyze the condition of the tilting bearing by the tilting signal. And on the monitoring screen, the signal waveform input to the sensor and RLCR (a numerical value that intersects based on the set value of the reference frequency of the signal, generally displays a value of 7.0 to 14.0). If the RLCR value is less than 6 but displays a high value of 15 or more, it can be confirmed that there is an error in the tilting bearing. At this time, it can be known clearly through the signal analysis screen.

19 is an angle change monitoring screen of a tilt signal from normal (12 mA, 0 degree) to phase (20 mA, 180 degree). It is possible to accurately identify the problem of the tilting bearing by analyzing the sound data by the tilting signal and LPCC. In FIG. 19, the tilting signal is maintained in the case of a stop state (12mA, No. 3), and the time and angle of the tilting bearing moving forward (20mA, No. 3) are displayed. It can be seen that the shape of the LPCC graph and the average signal (red line in the middle) are constantly displayed when the data signal is operating normally (normal if the average value is between 7 and 12).

20 is an LPCC monitoring screen in a state in which there is no tilting signal. You can see a graph showing the waveform of the LPCC in the case of no tilt signal (No. 3). Here, it can be seen that the shape of the LPCC graph and the average signal (red line in the middle) are constantly displayed (average values of 7 to 12 are normal). In other words, it means that there is no problem with the tilting bearing.

21 is a monitoring screen in which the data of the tilting signal (No. 3) is analyzed by 5 algorithms. The tilting signal (No. 3) is shown at regular intervals. It can be confirmed that the tilting signal in red and dark bars is also shown on the frequency spectrogram screen (No. 1). And you can check all the screens analyzed with the frequency spectrum, Bark Scale, Bar Graph, LPCC, and RLCR values. Here, too, it can be confirmed that the LPCC value and the RLCR value are within the range of the set values (refer to 1 and 2 above). (Here, number 1 is the frequency spectrogram and Bark Scale screen, number 2 is the Bar Graph and LPCC analysis screen, and number 3 is the sound signal and the tilting signal in the enlarged waveform.)

22 is a monitoring screen showing changes in 5 algorithms caused by external noise (No. 3). It is a screen displayed by 5 algorithms by external noise (No. 3). Here, the frequency spectrogram is displayed in red as a whole, and it can be seen that a lot of energy in the frequency spectrogram is displayed (dark red) in areas where there is noise. there is. In this way, there is external noise, but there is no change in LPCC and RLCR, and noise and problems of tilting bearings are clearly distinguished.

23 is a monitoring screen showing changes in five algorithms based on a sound signal and a frequency spectrogram when there is a lot of external noise (No. 3). It is a screen displayed by 5 algorithms when there is a lot of external noise (No. 3). Here, you can see that the frequency spectrogram (No. 1) is displayed in a wide red color and a lot of energy in the frequency spectrogram is displayed (dark red). there is.

24 is a monitoring screen on which a change in an incoming sound signal in which an external noise (No. 3) and a tilt signal are mixed together can be seen. The external noise (No. 3) and the tilting signal are mixed together. Here, the waveform of LPCC changes a lot. However, the average value shows 9.1 (meaning there is no problem). In addition, the energy of the tilting signal can be confirmed (dark bar-shaped red) in the frequency spectrogram. As described above, the first analysis screen by the real-time monitoring screen (sound signal and Bark Scale) and the second analysis program by the LPCC and RLCR algorithm are displayed on the tilting signal, and finally, the five algorithms (frequency spectrum, Bark Scale, Bark Scale) Graph, LPCC, RLCR value), the original sound signal, and the enlarged sound signal were displayed and analyzed to finally confirm that the problem of the tilting bearing can be accurately identified in advance.

On the other hand, as shown in FIG. 25, the current signal of the tilting signal is displayed as before (20mA), stop (12mA), and reverse rotation (4mA) of the tilting signal as shown in (3) by the hardware system through voltage and ADC. This makes it possible to accurately analyze the condition of the tilting bearing by the tilting signal. And on the monitoring screen, the signal waveform input to the sensor and RLCR (a numerical value that intersects based on the set value of the reference frequency of the signal, generally displays a value of 7.0 to 14.0). If the RLCR value is less than 6 but displays a high value of 15 or more, it can be confirmed that there is an error in the tilting bearing. At this time, it can be known clearly through the signal analysis screen.

26 shows a monitoring screen obtained by analyzing data of a tilting signal (No. 3) by 5 algorithms. The sound of the tilting bearing can be heard as it is as shown in FIG. When there is a tilting signal, among the sound signals shown in No. 3 (waveform amplified by the original signal), the bar-like part marked at regular intervals indicates the operation time of the tilting bearing. Also, if you look at number 1 in the frequency spectrogram, a dark red bar appears in the middle of the frequency spectrum, which shows the operating state of the tilting bearing. If you listen to this sound signal, you can clearly distinguish between an operating state and a non-operating state of the tilting bearing.

27 shows a monitoring screen of signals coming into four sensors and tilting signals in real time. "No. 1" is the screen that shows the signals coming from 4 sensors (2 drive axes, 2 driven axes) in real time in time domain waveform and Bark Scale. In addition, an event ("No. 2") was generated for the signal of the start (moving the tilting bearing) and the end (stopping the tilting bearing) by the tilting signal. When this event signal is clicked and analyzed, the screens shown in FIGS. 30 and 31 can be seen.

"3" in FIG. 27 indicates a tilting signal and indicates time and angle. Here, the tilting signal converts the PLC current signal into a voltage, and displays the before (20mA), stop (12mA), and reverse rotation (4mA) of the tilting signal as shown in “3” by the data through the ADC. This makes it possible to accurately analyze the condition of the tilting bearing by the tilting signal. And on the monitoring screen, the signal waveform input to the sensor and RLCR (a numerical value that intersects based on the set value of the reference frequency of the signal, generally displays a value of 7.0 to 14.0). If the RLCR value is less than 6 but displays a high value of 15 or more, it can be confirmed that there is an error in the tilting bearing. At this time, it can be known clearly through the signal analysis screen.

28 is an angle change monitoring screen while the tilting signal changes from phase (20mA) to steady state (12mA). 27 is a monitoring screen, and displays the time and angle of the tilting bearing in which the tilting signal moves from phase (20mA) to steady state (12mA). Here, by displaying a value within 8.03 to 12.3 of the RLCR value, it can be confirmed that the condition of the tilting bearing is normal. More detailed analysis can be confirmed by viewing the data signal of "No. 2" on the analysis screen.

29 is a monitoring screen of a waveform (2) in the time domain in a state (2) with a tilt signal. FIG. 29 is an enlarged view of the driven shaft sensor parts S3 and 4 of FIG. 28 . Here, when the tilting signal proceeds (No. 2), the tilting signal sound is displayed in red (1) in the waveform of the time domain. If you listen to the waveform sound in this time domain, you can accurately hear the sound of the tilting bearing. What is displayed in red is that the waveform is blue when it is normal, but when there is a tilt signal, it is displayed in red because the sound is higher than the reference value. And it can be confirmed that the tilting bearing is operated by this red color. And the RLCR value is indicating 8.3. That is, it confirms that there is no problem in the operation of the tilting bearing. In other words, it means that there is no problem with the tilting bearing.

30 is an angle change monitoring screen of a tilt signal from normal (12 mA, 0 degrees) to phase (20 mA, 180 degrees). When the tilting signal is stopped (12mA, No. 3), it is maintained, and the time and angle of the tilting bearing moving forward (20mA, No. 3) are displayed. It can be seen that the shape of the LPCC graph and the average signal (red line in the middle) are constantly displayed when the data signal is operating normally (normal if the average value is between 7 and 12).

31 shows a graph showing the waveform of the LPCC when there is no tilt signal (No. 3). Here, it can be seen that the shape of the LPCC graph and the average signal (red line in the middle) are constantly displayed (average values of 7 to 12 are normal). (In other words, it means that there is no problem with the tilting bearing).

Low-speed bearing rotation noise generated during tilting causes problems such as natural deterioration, poor maintenance, and foreign substances during use (bearing contact) as the cause of failure. The system according to this embodiment has the effect of improving the reliability of the predictive system by analyzing the sound in which the sound of the rotational sound is heard differently and diagnosing through acoustic analysis before the problem of the tilting bearing occurs (prediction). In addition, it is possible to improve the status of the domestic steel industry by reducing costs through economical repair, maintenance, and replacement of steel-making tilting bearings, and by developing a unique acoustic-based anomaly prediction system.

The present invention described above is not limited by the above-described embodiments and the accompanying drawings, and various substitutions, modifications, and changes are possible without departing from the technical spirit of the present invention will be apparent to those skilled in the art.

Claims (5)

A sensor installed on the tilting bearing to collect generated sound;
a system server that collects sound data from the sensor; and
A comprehensive server that receives sound data from the system server, converts and amplifies the PLC current signal generated from the tilt bearing into a voltage signal, and diagnoses the acoustic waveform and pattern according to the angle through the ADC. Characterized in that the steel-making tilt-dong bearing unique sound-based abnormality diagnosis system. According to claim 1,
The sensor is a steel-making tilt bearing unique sound-based abnormal diagnosis system, characterized in that disposed on both sides or in front of the drive shaft and the driven shaft of the tilt bearing. According to claim 1,
The system server constantly monitors and records the sound in the forward and reverse directions of the tilting bearing and converts it into a DB. According to claim 1,
The comprehensive server processes the pattern of sound generated from the tilting bearing in any one of time domain, frequency domain, Bark Scale, LPCC (Linear Prediction Cepstral Coefficient), MFCC (Mel Frequency Cepstral Coefficient), Frequency_Bar Graph, and RLCR sound data processing module that analyzes shape and size by
The PLC current signal received by wire or wirelessly from the tilting bearing and the sound signal of the system server are received, and the tilting signal current is converted into a voltage signal and amplified to obtain an angle (-180 to 180 degrees) through the ADC. A steel-making tilt-rolling bearing specific sound-based abnormality diagnosis system, characterized in that it comprises a diagnosis module for diagnosing the acoustic waveform and pattern according to FIG. According to claim 1,
The comprehensive server processes the sound signal by any one of the time domain, frequency domain, Bark Scale, LPCC (Linear Prediction Cepstral Coefficient), MFCC (Mel Frequency Cepstral Coefficient), Frequency_Bar Graph, and RLCR for the sound caused by the rotation of the steel-making tilt bearing. , Steel-making tilt-rolling bearing specific sound-based abnormal diagnosis system, characterized by converting and amplifying the PLC current signal into a voltage signal and diagnosing the acoustic waveform and pattern according to the angle (-180 to 180 degrees) through the ADC.

Images