KR20210111502A - Pre-designated equipment of a facility and Predictive maintenance method using the same - Google Patents

Abstract

A predictive maintenance device according to one embodiment of the present invention comprises: an input part that inputs a type of facility or structure generating at least one among a noise and a vibration; a vibration frequency measuring part that measures a vibration frequency of the vibration for a preset time; a noise measuring part that measures the noise for a preset time; a deep learning self-diagnosis learning part that, after learning the noise and vibration generated in the normal state of the facility or structure, extracts a change in a periodic pattern of the measured noise and the measured vibration frequency based on the learning result, and predicts an occurrence of an abnormality in the facility or structure through a size change of the extracted periodic pattern; and an output part that outputs a prediction result of the deep learning self-diagnosis learning part in a voice or visual form. Therefore, the present invention is capable of having an advantage of being able to predict a safety abnormal condition in advance.

Description

Predictive maintenance apparatus and method {Pre-designated equipment of a facility and Predictive maintenance method using the same}

The present invention relates to a predictive maintenance apparatus and method.

In order to maintain safety and performance, investment in equipment and maintenance costs is increasing due to the recent large-scale damage caused by accidents or facility shutdowns to structures and facilities. In addition to routine maintenance performed to maintain safety and performance, preventive maintenance such as preliminary measures, planned periodic inspections, regular repairs, and regular replacements is performed. Although it is an advanced method compared to the break down maintenance method, which is repaired after a breakdown occurs, there is a problem in that unnecessary periodic replacement and unnecessary periodic inspection costs occur in planned preventive maintenance.

In addition, if the replacement cycle is applied as a standard for preventive maintenance based on the operating time of equipment, individual replacement cycle of parts, and operating environment, the relationship between each part that has a greater impact on the life cycle of parts, such as lubrication, etc. It is difficult to apply combinatorial correlations.

In addition, in the preventive maintenance method for conventional facilities, the vibration generated by repeated periodic rotation of an object with a rotating shaft, such as a motor, is pre-processed and Fourier transform is applied to the collected data using a vibration meter. After setting the pattern as a standard pattern, the same conditions and procedures are used for the object to be inspected, and the data-converted spectral pattern is analyzed and compared with the standard pattern to determine the state of the object.

This method requires environmental preparation to create a standard pattern using a unique vibration characteristic pattern of an object, and an expert intervention process to set filter conditions and sensor set values. However, there is also a limitation that environmental conditions for generating a standard pattern must be specified and maintained during inspection.

In the fields of non-destructive testing, such as structural stability verification, weld fatigue test, and crack occurrence inspection, vibration and noise inspection methods are used to collect vibration and sound signals generated by impacting an object with a force under a certain condition to determine the object's natural frequency and sound. By quantifying the resonance data, the stability of the structure and the state of the welds are determined by the method of comparative analysis with the reference data.

Such a conventional vibration test method requires a lot of measuring equipment and expertise in measurement in the process of generating a standard pattern, and there is a problem in that it is also necessary to check the diagnostic test result.

In another field, in the field of inspecting metal hardness, weld defects, cracks, etc. in a method that relies on the experience of the skilled person due to the worker's hearing by impacting with a simple handy hammer without a separate measuring device, the precision of the results where the skill of the worker is important There is a problem in determining

Accordingly, there is a problem in that non-uniform test results are obtained depending on the skill level of the skilled person and even the health condition.

As described above, there is a problem in that it is difficult to apply universally and generalized to various target objects and various inspection environments in the vibration inspection method by the conventional statistical rule-based program or by the repeated experience of the skilled person as described above.

Patent Publication No. 10-2009-0133160

An object of the present invention is to provide a predictive maintenance apparatus and method capable of solving the problems of the prior art.

Predictive maintenance apparatus according to an embodiment of the present invention for solving the above problems is an input unit for inputting a type of facility or structure generating at least one of noise and vibration; a vibration frequency measuring unit for measuring the vibration frequency of the vibration for a preset time; a noise measuring unit for measuring the noise for a preset time; After learning the noise and vibration occurring in the normal state of the facility or structure, the change in the periodic pattern of the measured noise and the measured vibration frequency is extracted based on the learning result, and the change size of the measured periodic pattern is used. a deep learning self-diagnosis learning unit for predicting abnormal occurrence of the facility or structure; and an output unit for outputting the prediction result of the deep learning self-diagnosis learning unit in voice or visual form.

Predictive maintenance method according to an embodiment of the present invention for solving the above problems comprises the steps of: inputting a type of facility or structure in which at least one of noise and vibration occurs in an input unit; Measuring the vibration frequency of the vibration for a preset time in the vibration frequency measuring unit; measuring the noise for a preset time by a noise measuring unit; After learning the noise and vibration that occur in the normal state of the facility or structure in the deep learning self-diagnosis learning unit, extract the change in the periodic pattern of the measured noise and the measured vibration frequency based on the learning result, and predicting the occurrence of abnormalities in the facility or structure through the change size of the periodic pattern; and outputting the prediction result of the deep learning self-diagnosis learning unit in voice or visual form in the output unit.

By using the equipment predictive maintenance apparatus and method according to an embodiment of the present invention, there is an advantage that it is possible to predict in advance the location of defects in equipment and structures generating vibrations and sounds, whether cracks occur, and safety abnormalities, etc. .

1 is a block diagram showing a predictive maintenance apparatus according to an embodiment of the present invention.
FIG. 2 is a detailed configuration diagram of the vibration frequency measuring unit shown in FIG. 1 .
3 is an exemplary diagram of a deep learning network.
4 is a flowchart illustrating a predictive maintenance method according to an embodiment of the present invention.

Hereinafter, embodiments of the present specification will be described with reference to the accompanying drawings. However, it is to be understood that the technology described herein is not limited to specific embodiments, and includes various modifications, equivalents, and/or alternatives of the embodiments of the present specification. . In connection with the description of the drawings, like reference numerals may be used for like components. In the present specification, expressions such as “have,” “may have,” “include,” or “may include” indicate the presence of a corresponding characteristic (eg, a numerical value, function, operation, or component such as a part). and does not exclude the presence of additional features.

In this specification, expressions such as "A or B," "at least one of A and/and B," or "one or more of A or/and B" may include all possible combinations of the items listed together. . For example, "A or B," "at least one of A and B," or "at least one of A or B" means (1) includes at least one A, (2) includes at least one B; Or (3) it may refer to all cases including both at least one A and at least one B.

As used herein, expressions such as "first," "second," "first," or "second," can modify various elements, regardless of order and/or importance, and refer to one element. It is used only to distinguish it from other components, and does not limit the components. For example, the first user equipment and the second user equipment may represent different user equipment regardless of order or importance. For example, without departing from the scope of the rights described herein, a first component may be referred to as a second component, and similarly, the second component may also be renamed as a first component.

A component (eg, a first component) is "coupled with/to (operatively or communicatively)" to another component (eg, a second component); When referring to "connected to", it should be understood that the certain element may be directly connected to the other element or may be connected through another element (eg, a third element). On the other hand, when it is said that a component (eg, a first component) is "directly connected" or "directly connected" to another component (eg, a second component), the component and the It may be understood that other components (eg, a third component) do not exist between other components.

As used herein, the expression "configured to (or configured to)" depends on the context, for example, "suitable for," "having the capacity to ," "designed to," "adapted to," "made to," or "capable of." The term “configured (or configured to)” may not necessarily mean only “specifically designed to” in hardware. Instead, in some circumstances, the expression “a device configured to” may mean that the device is “capable of” with other devices or parts.

For example, the phrase “a processor configured (or configured to perform) A, B, and C” refers to a dedicated processor (eg, an embedded processor) for performing the operations, or by executing one or more software programs stored in a memory device. , may mean a generic-purpose processor (eg, a CPU or an application processor) capable of performing corresponding operations.

The terms used herein are used only to describe specific embodiments, and may not be intended to limit the scope of other embodiments. The singular expression may include the plural expression unless the context clearly dictates otherwise. Terms used herein, including technical or scientific terms, may have the same meanings as commonly understood by one of ordinary skill in the art described herein. Among the terms used herein, terms defined in a general dictionary may be interpreted with the same or similar meaning as the meaning in the context of the related art, and unless explicitly defined in the present specification, have ideal or excessively formal meanings. is not interpreted as In some cases, even the terms defined in this specification cannot be construed to exclude the embodiments of the present specification.

Hereinafter, a predictive maintenance apparatus and method according to an embodiment of the present invention will be described in more detail based on the accompanying drawings.

1 is a block diagram showing a predictive maintenance apparatus according to an embodiment of the present invention, FIG. 2 is a detailed configuration diagram of the vibration frequency measuring unit shown in FIG. 1, and FIG. 3 is an exemplary diagram of a deep learning network.

As shown in Fig. 1, the predictive maintenance apparatus 100 according to an embodiment of the present invention is a state that is always attached to a facility or structure generating vibration and noise, and is in a state of being constantly attached to the vibration and sound trend occurring in the facility and structure in real time. After learning the normal and defective state of facilities and structures based on the deep learning algorithm, a predictive maintenance warning sound is generated when the vibration and noise generated in the facility reaches the minimum threshold of the learned normal state vibration and noise. It may be an invention characterized in that it is provided to the manager.

More specifically, the predictive maintenance device 100 includes an input unit 110 , a setting unit 120 , a vibration frequency measuring unit 130 , a noise measuring unit 140 , a deep learning self-diagnosis learning unit 150 , and an output unit. (160).

The input unit 110 may be configured to input a type of facility or structure in which at least one of noise and vibration is generated.

In addition, the input unit 110 provides a user-touch-based input UI and/or a voice recognition-based input UI (User-Interface) that is linked to a display unit of the output unit 150 to be described later.

The setting unit 120 may be configured to set a measurement range of vibration frequency and a measurement range of noise required for self-diagnosis learning according to information (type) of a facility or structure input from the input unit 110 .

In addition, the setting unit 120 is configured to set the frequency filter value of the vibration frequency measurement unit 130 and the decibel filter value of the noise measurement unit to measure information corresponding to the set vibration frequency measurement range and noise measurement range can

In addition, the setting unit 120 may be configured to set a learning pattern in the deep learning self-diagnosis learning unit 150 according to the number of local measurement ranges of equipment or structures.

Next, the vibration frequency measuring unit 130 measures the vibration frequency of the vibration for a preset time.

The vibration frequency measurement unit 130 includes a frequency amplifier 131 , a signal conversion unit 132 , a filter unit 133 , and an FFT analysis unit 134 .

The frequency amplifying unit 131 may be configured to amplify a fine vibration (tremor) signal generated on the surface of a facility or structure.

The signal conversion unit 131 is configured to convert the amplified analog vibration signal into a digital signal. In general, when converting an analog signal to a digital signal, if the highest frequency of the analog signal is less than half of the sampling frequency, there is no problem, but the analog If the highest frequency of the signal is greater than half of the sampling frequency, distortion may occur in the digitally converted signal.

Accordingly, in the present application, by sequentially performing the anti-aliasing analog filter, quantization, oversampling, anti-aliasing digital filter, and downsampling of the filter unit 133, in the conversion process It is configured to prevent signal distortion.

The FFT (Fast Fouier Transform) analyzer 134 analyzes the digitally converted vibration signal into a time domain and a frequency domain. Time domain analysis generates RMS, Peak, Crest Factor, and Kurtosis data through a low pass filter, a high pass filter, a band pass filter, a differentiator, and an integrator.

Frequency domain analysis is to obtain the magnitude of the component of the vibration signal as a function of frequency, and it is derived using mechanical and electrical filters to calculate the numerical value in the measured waveform of the vibration signal.

The noise measuring unit 140 may be configured to measure the noise generated in the facility or structure for a preset time for a preset time.

On the other hand, the noise measuring unit 140 converts the noise generated in the facility or structure for a preset time into a voltage signal according to the frequency, and then passes only the voltage signal of the predetermined frequency band based on the voltage signal of the predetermined frequency band. It may be a configuration for calculating the sound pressure value.

Next, the deep learning self-diagnosis learning unit 150 learns the noise and vibration generated in the normal state of the facility or structure, and then changes the periodic pattern of the measured noise and the measured vibration frequency based on the learning result. may be configured to extract and predict the occurrence of abnormalities in the facility or structure through the magnitude of change in the extracted periodic pattern.

The deep learning self-diagnosis learning unit 150 mixes learning data including periodic patterns of noise and vibration frequencies generated when the facility or structure is in an abnormal state by applying it to a Mix-augmentation method. It is possible to derive (generate) learning data of noise and vibration frequencies that occur in an abnormal state.

Meanwhile, the deep learning self-diagnosis learning unit 150 may learn a neural network or derive a result based on a network structure configured through learning. The structure of the neural network may form a network structure composed of one or more layers, and such a network structure includes an artificial neural network and a convolutional neural network.

Network), a recurrent neural network, a bidirectional neural network, and the like may be included. Furthermore, the network structure may be configured by reflecting various methods such as convolution, subsampling, activation, drop out, softmax, and normalization.

In addition, the deep learning self-diagnosis learning unit 150 determines the periodic pattern characteristics of the frequencies of noise and vibration in an abnormal state based on the output value of the network node. Since the neural network outputs results through the combination of the output values of numerous nodes, the activated node may vary according to the periodic pattern characteristics of the input signal, that is, the noise and vibration frequencies in an abnormal state.

On the other hand, the deep learning self-diagnosis learning unit 150 of the present invention can be adjusted according to the noise and vibration frequency range generated in the applied equipment or structure, so that the layer of the network structure, for example, the input layer, the hidden layer ( hidden layer) and variable number of neurons in the output layer, and may include a function of adjusting the learning progress rate.

For example, the bias of the network is initialized to 0, and since the weight of each neuron connection affects the learning result, the range of initialization values for all connection strengths of the input, hidden, and output layers is determined and initialized.

At this time, if the number of input neurons is set to n in the range of the initialization value, the weight is initialized by generating a random number generator in the range [-/n, +1/n]. can do.

Next, the output unit 160 may output the prediction result of the deep learning self-diagnosis learning unit 150 in voice or visual form.

The output unit 160 may include a display unit and an alarm unit, and the display unit may include a liquid crystal display (LCD), a thin film transistor liquid crystal display (TFT LCD), and an organic light emitting diode (OLED). It may include at least one of a light-emitting diode (OLED), a flexible display, a three-dimensional display (3D display), an e-ink display, and a light emitting diode (LED).

4 is a flowchart illustrating a method for predictive maintenance of a facility or structure according to an embodiment of the present invention.

Referring to FIG. 4 , in the predictive maintenance method ( S700 ) of a facility or structure according to an embodiment of the present invention, the type of facility or structure in which at least one of noise and vibration is generated is input in the input unit 110 ( S710 ) Then, the vibration frequency measuring unit 130 measures the vibration frequency of the vibration for a preset time (S720), and the noise measuring unit 140 measures the noise for a preset time (S730).

Thereafter, after learning the noise and vibration generated in the normal state of the facility or structure in the deep learning self-diagnosis learning unit 150, the change in the periodic pattern of the measured noise and the measured vibration frequency based on the learning result After extracting and predicting the occurrence of abnormalities in the facility or structure through the magnitude of change in the extracted periodic pattern (S740), the output unit 160 displays the prediction result of the deep learning self-diagnosis learning unit 150 by voice or visual to output (S750).

Therefore, using the predictive maintenance apparatus and the predictive maintenance method using the predictive maintenance apparatus according to an embodiment of the present invention, predicts and detects abnormalities in the conditions of facilities and structures in which vibrations and sounds are generated, and based on this, the facilities and structures are notified to the manager It has the advantage of being able to warn (alarm) the status abnormality.

Through this, it provides the advantage of being able to minimize the cost of post-management due to facility management and structure management.

"~" used in an embodiment of the present invention may be implemented as a hardware component, a software component, and/or a combination of a hardware component and a software component. For example, devices and components described in the embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), It may be implemented using one or more general purpose or special purpose computers, such as a programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For convenience of understanding, although one processing device is sometimes described as being used, one of ordinary skill in the art will recognize that the processing device includes a plurality of processing elements and/or a plurality of types of processing elements. It can be seen that can include For example, the processing device may include a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as parallel processors.

The software may comprise a computer program, code, instructions, or a combination of one or more thereof, which configures a processing device to operate as desired or is independently or collectively processed You can command the device. The software and/or data may be any kind of machine, component, physical device, virtual equipment, computer storage medium or device, to be interpreted by or to provide instructions or data to the processing device. , or may be permanently or temporarily embody in a transmitted signal wave. The software may be distributed over networked computer systems, and stored or executed in a distributed manner. Software and data may be stored in one or more computer-readable recording media.

The method according to the embodiment of the present invention may be implemented in the form of program instructions that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be specially designed and configured for the embodiment, or may be known and available to those skilled in the art of computer software. Examples of the computer readable recording medium include magnetic media such as hard disks, floppy disks and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floppy disks. - includes magneto-optical media, and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language codes such as those generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

Those of ordinary skill in the art to which the present invention pertains may modify and modify the above-described contents without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed in the present invention are not intended to limit the technical spirit of the present invention, but to explain, and the scope of the technical spirit of the present invention is not limited by these embodiments. The protection scope of the present invention should be construed by the following claims, and all technical ideas within the scope equivalent thereto should be construed as being included in the scope of the present invention.

100: predictive maintenance device
110: input unit
120: setting unit
130: vibration frequency measurement unit
131: frequency amplification unit
132: signal conversion unit
133: filter unit
134: FFT analysis unit
140: noise measuring unit
150: Deep learning self-diagnosis learning unit
160: output unit

Claims (6)

an input unit for inputting a type of facility or structure generating at least one of noise and vibration;
a vibration frequency measuring unit for measuring the vibration frequency of the vibration for a preset time;
a noise measuring unit for measuring the noise for a preset time;
After learning the noise and vibration occurring in the normal state of the facility or structure, the change in the periodic pattern of the measured noise and the measured vibration frequency is extracted based on the learning result, and the change size of the extracted periodic pattern is used. Deep learning self-diagnosis learning unit for predicting abnormal occurrence of the facility or structure; and
Predictive maintenance apparatus including an output unit for outputting the prediction result of the deep learning self-diagnosis learning unit in voice or visual. According to claim 1,
Pre-maintenance apparatus further comprising a setting unit for setting the measuring range (hz) of the vibration frequency measuring unit and the measuring range (db) of the noise measuring unit according to the type of facility or structure input from the input unit. 3. The method of claim 2,
Predictive maintenance apparatus, characterized in that the setting unit sets the frequency filter value of the vibration frequency measurement unit and the decibel filter value of the noise measurement unit to measure information corresponding to the set vibration frequency measurement range and noise measurement range. According to claim 1,
The vibration frequency measuring unit
a frequency amplifier for amplifying a fine vibration (tremor) signal generated on the surface of the facility or structure;
a signal converter converting the amplified analog vibration signal into a digital signal;
a filter unit that sequentially performs an anti-aliasing analog filter, quantization, oversampling, anti-aliasing digital filter, and downsampling to prevent signal distortion in the digital signal conversion process; and
Predictive maintenance device including a FFT (Fast Fouier Transform) analyzer that analyzes the digitally converted vibration signal into a time domain and a frequency domain. According to claim 1,
The noise measuring unit
After converting the noise generated in the facility or structure for a preset time into a voltage signal according to the frequency, only the voltage signal of a predetermined frequency band passes through, and the sound pressure value is calculated based on the voltage signal of the predetermined frequency band. Predictive maintenance device. inputting a type of facility or structure in which at least one of noise and vibration is generated in the input unit;
Measuring the vibration frequency of the vibration for a preset time in the vibration frequency measuring unit;
measuring the noise for a preset time in a noise measuring unit;
After learning the noise and vibration generated in the normal state of the facility or structure in the deep learning self-diagnosis learning unit, extract the change in the period pattern of the measured noise and the measured vibration frequency based on the learning result, and Predicting the occurrence of abnormalities in the facility or structure through the change size of the periodic pattern; and
Predictive maintenance method comprising the step of outputting the prediction result of the deep learning self-diagnosis learning unit by voice or visual in an output unit.

Images