KR101273331B1 - Non-contact remote detection method for the failures of hydraulic and pneumatic systems - Google Patents

Abstract

PURPOSE: A remote noncontact type malfunction sensing method is provided to senses factors of problems generated when the operation hydro-pneumatic pressure of a hydro-pneumatic device is not transmitted properly or the hydro-pneumatic pressure desired not to be operated is operated in advance. CONSTITUTION: A remote noncontact type malfunction sensing method is as follows. A grating having a predetermined interval is formed in a target(105) on the surface of a vibrating object. The micro vibration of the target is magnify-photographed by using a telescope in a remote area. Phases captured in a viewer of the telescope are re-magnified by using a microscope(101). The phases magnified by the microscope are photographed by a high-speed camera. The data of the images photographed by the high-sped camera is processed by a computer process device.

Description

Non-contact remote detection method for the failures of hydraulic and pneumatic systems}

The present invention relates to a remote non-contact failure detection system, and more particularly, to a system and a method for detecting a failure in a non-contact state with respect to a device in which vibration continuously occurs, such as a hydraulic device in a ship engine room.

In general, ships consist of a large number of hydraulic devices. Leakage of pneumatic pipes causes problems such as suspension of ship operation and flight delays, and it is difficult to properly identify the location and fault diagnosis where a failure has occurred, and expensive repair cost is consumed even after the ship is desperately required for abnormal diagnosis system of pneumatic system. In addition, it can be said that a non-contact vibration sensing system is necessary for diagnosing a place where a person is difficult to access. On the other hand, Figures 10 and 11 is a hydraulic pneumatic pipe experiment apparatus, and corresponds to the object for detecting the presence or absence of abnormality by the remote non-contact failure detection system of the present invention.

 In particular, the hydraulic system inside the engine room of the ship is designed to be monitored at all times because of its importance.However, even if the monitor shows that it is operating normally, the hydraulic pressure of the hydraulic components cannot be properly delivered due to the seizure of parts. In some cases, the operating hydraulic pressure may be transmitted when it should not.

In this case, it can be detected from the pressure element in the inlet line, but there is no countermeasure to attach the pressure sensor because the pipe diameter of the control hydraulic system is small.

In general, a method of measuring vibration is to attach a vibration sensor that can detect vibration on a vibrating object, and the detected vibration signal is sent to the signal analyzer through a wired cable, so that the maximum vibration value, average vibration value, and vibration spectrum are obtained. To be analyzed.

As the vibration sensor, a speed pickup for detecting the vibration speed and an acceleration pickup for detecting the vibration acceleration are used. The vibration sensor is attached to a part to measure the vibration to vibrate with the object.

The signal analyzer includes a filter for removing noise, an A / D converter for converting an analog signal detected by a sensor into a digital signal, an FFT module for obtaining a spectrum by frequency analysis, an integrating circuit for integrating or integrating a signal, and measured vibration. A display window for displaying a value, an input button for operating a signal analysis method, and the like are provided.

However, the vibration of a stationary object can be analyzed using the conventional vibration measuring device as described above. However, when measuring vibration of a rotating object or an object in a linear motion, a signal cable cannot be used. In addition, in a conventional vibration sensor such as acceleration pickup and speed pickup, the mass of the sensing unit receives acceleration such as centrifugal force, so that accurate measurement is not performed. Therefore, in this case, a non-contact sensor such as an eddy current sensor or a gap sensor is used.

However, even these sensors are often difficult to use the existing non-contact sensor because the diameter of the pipe is small or there are irregularities such as coupling bolts, keys.

The present invention has been made to solve the problems of the prior art, and an object of the present invention is to provide a system that can remotely detect whether or not to operate the hydraulic system of the hydraulic system in the vessel.

The problem solving means of this invention is as follows.

That is, the target is attached to the vibrating object and the target has a grid having a predetermined interval on the surface;

A telescope for remotely capturing fine vibrations of the target;

A microscope for re-enlarging an image captured by the telescope viewer;

A high speed camera for capturing an image magnified by the microscope at high speed;

Computer processing apparatus for processing image data captured by the high speed camera

, ≪ / RTI &

The target is attached to a grid having a predetermined interval,

The microscope is arranged in a structure that is eyepiece with the telescope, the distance between the telescope 102 viewer and the bottom end of the microscope 101 is set to obtain a clear image in consideration of the magnification and focal length of the microscope,

One side of the microscope 101 is directly connected to the high-speed camera 100, shooting with a high-speed camera to secure a plurality of images for a certain time, to check the amount of pixel movement between the first and second images of the image, the observed image The amount of image movement between the first and next chapters is 2 pixels so that it can be used for PIV calculation processing.

In the host computer 103, the secured image data processing is performed,

Processing of photographed image data is performed by using the PIV method, and the frequency domain of the vibration object is derived by converting the brightness information of the spatial domain into the information of the frequency domain in the image processing, and the integral value of the frequency domain is converted into a risk threshold. A remote non-contact failure detection system and method for determining whether a vibration object is detected by setting the area is presented.

The present invention exhibits the effect of detecting the large and small problem factors of the ship in advance due to the operation of the pneumatic pressure that is not properly delivered or the operation of the pneumatic device in the vessel due to the cohesion or the like.

In particular, by using a combination of a telescope and a microscope, a mechanism or a variety of structures, such as a long distance away by using the technology proposed in the present invention also has a function to prevent accidents caused by vibration.

A conceptual diagram of particle image velocimetry (PIV) is shown in FIG. 1, and FIG. 2 shows a principle of a gray level correlation method using image data of two frames.
3 shows the devices of the present invention.
4 is a view showing an image of a target to be observed in a high speed camera after passing through a microscope through a telescope,
5 corresponds to an image monitored by the host computer 103 after passing through the high speed camera 100.
6 is a photograph of a state in which a microscope and a high speed camera are installed in the telescope.
7 is a photograph taken only in the microscope in FIG.
8 and 9 are scenes of processing video image data using a self-developed program using a PIV technique.
FIG. 10 is a photograph of a pneumatic test apparatus, and FIG. 11 is an enlarged view of a target on which a lattice of FIG. 10 is drawn.
12 to 27 correspond to various Fourier transform graphs for verifying the reliability of the remote non-contact failure detection system of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings.

However, the scope of the present invention is not limited to the embodiment, it is made clear in advance that all inventions of the claims and equivalent scope.

The present invention describes an apparatus and method for sensing and analyzing a vibrating object remotely in a non-contact manner using PIV (Particle Image Velocimetry).

1. Principles applied to the invention

The local velocity of the flow field is easily obtained by knowing the direction and direction of the small straight line traveled by the tracer particles passing through any point.

를 실제의 유동속도에 근사시키려면 이동변위가 충분히 작아야 한다. 다시 말하면 입자가 그리는 궤적은 직선성과 등간격성이 보장되어야 한다.
In other words, it can be obtained from the relationship between the time interval and the vector displacement required for arbitrary particle motion as in the case of particle dynamics. The basic principle is simple

To approximate the actual flow velocity, the displacement must be small enough. In other words, the trajectories that the particles draw should ensure linearity and equidistance.

The conceptual diagram of PIV is shown in FIG. 1, and FIG. 2 shows the principle of the gradation value correlation method using image data of two frames.

The reason why the gray scale correlation method is used is that the vibration region of the target shows the micro pixel unit when the vibration object is to be measured at a long distance. In order to solve such a problem, a method of finding the end point of a vector by comparing the distribution characteristics of grayscale values is required instead of finding the centroid coordinates of individual particles.

Assuming that the gradation pattern of the particle image of the flow field does not change significantly during the micro time, the gradation value distribution in the correlation region in the first frame and the gradation value distribution in the second frame after the micro time show similar characteristics. Therefore, if the correlation coefficient is obtained by comparing the same magnitude on the second frame with the same area on the second frame with respect to the arbitrary position of the first frame, the largest value is to be regarded as the movement position of the same particle. It becomes possible.

If such operations can be performed at high speed on a computer, the end point of the vector can be easily obtained with respect to the virtual central particle (the start point of the velocity vector). The correlation coefficient between two consecutive frames is obtained by Equation 1.

Distribute particles with excellent traceavility in the flow field, record the instantaneous positions of these particles with a small time interval, and calculate the displacement of each particle. However, manual and artificial methods cause problems such as accuracy and processing capacity. . In order to solve such a problem, a computer having a large capacity for high speed processing may be used.

 Fourier transform is used to obtain the frequency domain, which converts the brightness information of the spatial domain into the information of the frequency domain in image processing. For convenience, only one-dimensional case will be considered. A pair of discrete Fourier transforms for the continuous function f (x) of the one-dimensional real number x is obtained using Equation (2).

If f (x) and g (x) are FT of F (u) and G (u), the convolution is paired as in the Fourier transform in the frequency domain analysis, and the relationship of Hold.

Discrete convolution is given by equation (4).

Correlation Equation is similar to synthesis, except that it takes a complex conjugate on one function, and the correlation rules for Fourier transformation apply as in synthesis. The difference in taking a complex conjugate from a function is different, and the same rules for Fourier transform apply as in synthesis.

은 상관의 표시이며, * 는 공액복소수를 의미한다.Above

Denotes the correlation and * denotes the conjugate complex number.

In addition, discrete correlation is the same as Equation (5).

In Equation (6), when f (x) and g (x) are the same, an auto-correlation function is obtained, and when it is different, a cross-correlation function is obtained. Equations (4) and (6) are used in Equation (3) and Equation (5) when the sum of the calculation equations is large due to discrete spaces or large spatial regions applied to the image image. As described above, the discrete fast FT technique may be applied to the calculation of the Fourier transform in the frequency domain.

2. Experiment apparatus, data processing and result analysis

In order to confirm the reliability of the invention, the experiment was performed by replacing the vibration source with a mobile phone. Furthermore, the vibration analysis of the pipe was carried out using the time-dependent velocity and Fourier transform of the source.

3 is a picture of the components of the monitoring system for confirming the vibration state by observing the vibrating object of the present invention.

The target 105 attached to the vibration object as a whole, the telescope 102 for observing the target, and the microscope 101 for magnifying the image formed on the telescope viewer once again are coupled to the telescope.

The high speed camera 100 is directly connected to the microscope to continuously photograph an image formed on the microscope at a high speed. On the other hand, since the laser 104 serves to facilitate the observation of the target target 105, the target may be illuminated using other light sources.

Referring to FIG. 3, the vibrating object to be observed becomes a target 105, and a lattice of 1 millimeter is attached to the target. This grating size can be adjusted according to the specifications of the telescope 102 and the distance between the target and the telescope.

In this experiment, when the distance to the target is 12 meters and the size of the target grid is 1 millimeter, the image formed on the telescope can be enlarged to 94 pixels by 94 pixels. (See FIG. 5) The reason why the distance to the target is 12 meters is according to the specification of the minimum focal length of the telescope.

In the telescope 102, the image is formed in the viewer by enlarging 100 times the size of the target, and of course, the magnification can be added or subtracted.

4 shows an image of a target to be observed after passing through a microscope through a telescope and formed on a high speed camera, and FIG. 5 corresponds to an image monitored by the host computer 103 through the high speed camera 100. .

It is very suitable to observe the movement of a fine target by vibration with a telescope. The specifications of the telescopes used in the table below are described, but not necessarily limited thereto (using the MEADE LX200ACF model).

Specifications

8"
Optical system

Advanced Coma Free
Telescope diameter

203mm (8)
Focal Length / Aspect Ratio
2000 mm

f / 10 (8)
Magnification

600X (8)
Eyepiece Series 4000
26mm Super
Plossl
Main material

Pyrex glass
Calibration plate material

water white glass

FIG. 6 is a photograph of a state in which a microscope and a high speed camera are installed in a telescope, and FIG. 7 is a photograph of only a microscope in FIG. 6.

Referring to FIG. 6, the viewer of the telescope 102 faces upward, and the microscope is arranged in a structure that is eyepiece with the telescope. The distance between the telescope 102 viewer and the lowest end of the microscope 101 was set to 35 millimeters because the magnification and focal length of the microscope are the most appropriate positions. In this experiment, the magnification of the microscope was set to 60 times.

The high speed camera 100 is directly connected to the upper portion of the microscope 101. The camera shoots images with a high speed camera, and obtains 500 images per second for a total of 2000 images for 4 seconds and checks the amount of pixel movement between the first and second images. (See FIG. 5 for pixels)

The amount of image movement between the first and next chapters of the observation image requires 2 pixels. The reason why the amount of movement of pixels is important is that the amount of movement of 2 pixels or more must be secured in order to calculate the PIV, and in order to satisfy this condition (that is, to secure the amount of movement of 2 pixels or more), enlargement of the target is required. .

The host computer 103 is responsible for processing the secured image data. In this experiment, image data was processed using a self-developed program using the PIV method (see FIGS. 8 and 9).

8 and 9, the pixel shift amount in the X-axis direction, the pixel shift amount in the Y-axis direction, the pixel shift amount in the XY axis, the Fourier transform in the X direction, the Fourier transform in the Y direction, and the Fourier transform in the XY axis Is derived.

 By using direct near-field imaging, the vibrational Hz of the object was found through Fourier transformations of the object during vibration and no vibration, and it was confirmed whether the same Hz could be obtained through long-distance photography. In other words, the calculation method used in the experiments below uses PIV (Particle Image Velocimetry), which is a direct method using a high speed camera 100 and a long distance shooting using a telescope (102, telescope) and a microscope (101). The verification process was performed using.

By directly photographing near-field, vibration Hz of the target object was found through Fourier Transform (FT) during vibration and no vibration, and it was confirmed whether the same Hz could be obtained through long-distance photographing.

3. System verification result

Experiments for verifying the reliability of the system of the present invention were performed on mobile phone vibration.

In more detail, the comparative experiment was to take a short distance direct shot using only a high-speed camera, but the vibration frequency of the mobile phone was found through the Fourier transformation of the object during vibration and no vibration, and through the configuration device of the present invention as shown in FIG. The reliability verification of the present invention is performed by checking whether the vibration frequency of the mobile phone found in the above-mentioned direct shooting can be found even in the long distance shooting by using a long distance shooting (using a microscope, a telescope and a high speed camera).

case vibration
conditions
Object
camera
(fps) time
(sec) A X
cell phone
(Cellular phone) 500 4 B 0 500 4 C X 500 4 D 0 500 4 One case five times
( AB : Direct shooting, CD : Long distance shooting)

In Table 2, four cases, namely cases A and B, are near-field direct photographing, and cases C and D are far-field photographing data.

More specifically, case A has a target in a vibration-free state, case B is in a vibration state, case C is in a vibration-free state, and case D is in a vibration state.

In the short distance shooting, unlike a long distance measurement, a 500-mm high speed camera is used to attach a 105 mm lens and to shoot at a distance of 250 mm from a target.

The reason is that the target object is enlarged to the maximum and the point where the image is clear is 250mm between the camera lens and the target.

A total of four seconds were taken and 2000 pictures were stored (500 shots per second) to perform PIV calculations. The data extracted through this calculation is stored in an Excel file, and the graphs are graphs of FIG. 12 and below.

FIG. 12 shows a moving path in an xy coordinate system as a result of PIV calculation of a representative point in a region in case A, and FIG. 13 shows a Fourier Transform (FT) calculation result of an X-direction component in Case A. FIG. . In the graph, the x direction is Hz and the y direction is FT value.

Fig. 14 shows the FT (Fourier Transform) calculation results of the Y-direction component in Case A. In the graph, the x direction is Hz, the y direction is an FT value, and FIG. 15 shows the results of FT (Fourier Transform) calculation of the entire XY component in Case A. FIG. In the graph, the x direction is Hz and the y direction is FT value.

In this case, it can be seen that case A measures the vibration of the surrounding buildings in the absence of vibration of the mobile phone.

Knowing that the vibration of the mobile phone is 40Hz as described above, the result of using the device (Fig. 3) of the present invention was also checked to see if it is 40Hz.

FIG. 20 shows a moving path in an x-y coordinate system as a result of PIV calculation of a representative point in a region in case C. FIG. Fig. 21 shows the FT (Fourier Transform) calculation result of the X-direction component in Case C. In the graph, the x-direction is Hz and the y-direction is FT value.

Fig. 22 shows Fourier Transform calculation results of the Y-direction component when Case C is used. In the graph, the x direction is Hz, the y direction is a Fourier transform value, and FIG. 23 shows the results of the Fourier transform calculation of the entire X-Y component in Case C. FIG. In the graph, the x direction is Hz and the y direction is a Fourier transform (FT) value.

In this case, in case C, the FT value is amplified by the mobile phone's non-vibration remote measurement, but it can be seen that it catches the vibration inside the building.

FIG. 23 shows a movement path in an x-y coordinate system as a result of PIV calculation of a representative point in a region in case D. FIG. Fig. 24 shows the FT (Fourier Transform) calculation result of the X-direction component in Case D. In the graph, the x-direction is Hz and the y-direction is FT value.

Fig. 25 shows the FT (Fourier Transform) calculation results of the Y-direction component in Case D. In the graph, the x direction is Hz and the y direction is FT value. Fig. 26 shows the results of FT (Fourier Transform) calculation of the entire XY component in Case D. In the graph, the x direction is Hx and the y direction is FT value.

In the case of the remote measurement using the telescope 102 and the microscope 101 as described above, when the mobile phone vibrated, the vibration of 40 Hz could be observed. Based on this fact, it is obvious that observing the vibration in the ship's pneumatic pipe is also possible to measure remotely. It is meaningful that it is a non-contact vibration detection method using a camera for the leakage of the pneumatic pipe installed in the ship engine room.

Reliability was verified by comparing the data of the telescope 102 and the microscope 101 with the remote measurement method and the near field measurement method when the mobile phone was a vibration source.

Through experiments, vibration was measured using the PIV technique, and through Fourier transform analysis, the relevant frequency domain was found, and the integral value at that frequency domain was set as the critical limit region, and it was used as a failure detection system of various vibration bodies such as hydraulic and pneumatic devices in ships. It can exert the function of.

The above embodiments of the present invention should not be construed as limiting the technical idea of the present invention. Those skilled in the art to which the present invention pertains can modify and change the technical idea of the present invention in various forms. Accordingly, such improvements and modifications will fall within the scope of the present invention as long as they are obvious to those skilled in the art.

100: high speed camera 101: microscope
102: telescope 103: computer
104: Laser 105: Target

Claims (3)

In the failure detection method for detecting the presence of abnormal vibration of the vibration object by remote contactless,
Forming a lattice having a predetermined spacing on a target of a vibrating object surface;
Remotely capturing a minute vibration of the target using a telescope;
Re-expanding the image captured by the telescope viewer through a microscope;
Photographing the image magnified by the microscope with a high speed camera;
Processing the image data photographed by the high speed camera by a computer processing apparatus;
The target is attached to a grid having a predetermined interval,
The microscope is arranged in a structure that is eyepiece with the telescope, the distance between the telescope 102 viewer and the bottom end of the microscope 101 is set to obtain a clear image in consideration of the magnification and focal length of the microscope,
One side of the microscope 101 is directly connected to the high-speed camera 100, shooting with a high-speed camera to secure a plurality of images for a certain time, to check the amount of pixel movement between the first and second images of the image, the observed image The amount of image movement between the first and next chapters is 2 pixels, so that it is used for PIV calculation processing.
In the host computer 103, the secured image data processing is performed,
Processing of photographed image data is performed by using the PIV method, and the frequency domain of the vibration object is derived by converting the brightness information of the spatial domain into the information of the frequency domain in the image processing, and the integral value of the frequency domain is converted into a risk threshold. It is possible to determine whether the vibration object is detected by setting the area.
And the distance between the target and the telescope is at least the minimum focal length of the telescope. delete delete

Images