Predictive maintenance method for device based on current waveform imbalance of multi-power unit

Abstract

Disclosed is a predictive maintenance method for a device based on current waveform imbalance of a multi-power unit, wherein real-time current waveforms are collected on the basis of values of currents consumed in respective power units of the device, and the collected real-time current waveforms are superimposed on each other to continuously detect waveform imbalance such that a mechanical abnormality in each of the power units can be detected, and various abnormalities such as defects in components associated with the power units of the device or errors in a control system for controlling the power units can also be detected.

Description

Predictive maintenance method for equipment through current waveform imbalance of multiple power units

[0001] The present invention relates to a predictive maintenance method for a device using current waveform imbalance of multiple power units, and more specifically, to a predictive maintenance method for a device using current waveform imbalance of multiple power units, which collects real-time current waveforms based on current values ​​consumed in each power unit of the device, and detects waveform imbalance continuously by superimposing the collected real-time current waveforms on each other, thereby detecting mechanical abnormalities of each power unit and simultaneously detecting various abnormalities such as defects in components related to the power unit of the device or errors in a control system that controls the power unit, thereby enabling stable predictive maintenance of a device including multiple power units that operate identically.

[0002] In general, various types of equipment are used in various industrial sites, and the stable operation of these equipment is very important.

[0003] If a single piece of equipment in a large-scale transport plant fails, the entire operation of the facility may be halted, resulting in enormous losses.

[0004] Therefore, recently, a predictive maintenance method for equipment has been developed to detect abnormal signs in the power unit (motor, pump, cylinder, compressor, etc.) that actually drives the equipment in advance and prevent huge losses due to sudden equipment failure.

[0005] However, since the conventional predictive maintenance method for equipment only detects abnormal signs in the power unit of the equipment, it is difficult to reliably detect abnormal signs in equipment where two or more power units must operate in the same manner. The reasons for this are as follows.

[0006] First, if an abnormality occurs in which one of the power units operates slightly faster or slower than the other power units due to an error in the control system while multiple power units that drive the device are operating normally, there is a problem in that the conventional predictive maintenance method for the device cannot detect the abnormality of the device because the power units are operating normally.

[0007] For example, if an abnormality such as the above occurs in the power unit that provides power to each wheel on both sides of a gantry crane, a safety accident may occur, such as the gantry crane tilting to one side and falling over or changing its direction of movement.

[0008] Second, while multiple power units that drive the device are operating normally, a problem with a component associated with one of the power units may cause the power unit to operate abnormally momentarily. However, there is a problem in that it is difficult to detect this phenomenon using conventional predictive maintenance methods for devices.

[0009] For example, if the rail that guides the wheels of a gantry crane is damaged or deformed, the power unit that provides power to the wheels passing over the damaged rail may momentarily perform abnormal movements, which may cause various safety accidents.

[0010] Therefore, there is a need to develop a predictive maintenance method for equipment that can reliably perform predictive maintenance on equipment that includes multiple power units that must operate identically.

[0011] The present invention has been proposed to solve the above-mentioned problems, and its purpose is to collect real-time current waveforms based on the current values ​​consumed in each power unit of the device, and to detect waveform imbalance continuously by superimposing the collected real-time current waveforms on each other, thereby detecting mechanical abnormalities in each power unit, and at the same time detecting various abnormalities such as defects in components related to the power unit of the device or errors in the control system controlling the power unit, thereby providing a predictive maintenance method for a device using current waveform imbalance of multiple power units capable of performing stable predictive maintenance of a device including a plurality of power units that operate identically.

[0012] In order to achieve the above object, the method for predictive maintenance of a device through current waveform imbalance of multiple power units according to the present invention is characterized by including: an information collection step (S10) for continuously measuring and collecting real-time current waveforms representing current values ​​consumed by each power unit over time during the operation of the device in real time when the device is operating; a detection step (S20) for continuously detecting waveform imbalance during the operation of the device by superimposing the real-time current waveforms of each power unit continuously measured and collected in the information collection step (S10); and a judgment step (S30) for setting a threshold value for the waveform imbalance and determining the device as being in an abnormal state and giving an alarm when the waveform imbalance continuously detected in the detection step (S20) exceeds the threshold value.

[0013] In addition, the method further includes a forced partitioning step (S40) of forcibly partitioning the real-time current waveform of each power unit measured and collected in the information collection step (S10) according to a set time interval, and setting the partitioned section as a unit partition section to form a plurality of unit partition sections on the real-time current waveform; wherein the detection step (S20) continuously detects waveform imbalance by overlapping the real-time current waveforms of each power unit, and continuously detects the waveform imbalance for the real-time current waveforms of the plurality of unit partition sections partitioned in the forced partitioning step (S40), and the judgment step (S30) is characterized in that when the waveform imbalance for the real-time current waveform of each unit partition section independently detected in the detection step (S20) exceeds a threshold value, the device is judged to be in an abnormal state and an alarm is generated.

[0014] In addition, in the information collection step (S10), instead of the current value measured and collected from the power unit, one of vibration, noise generated from the power unit, frequency of power supplied to the power unit, temperature, humidity, and pressure of the power unit is selected to measure and collect a real-time waveform such as a real-time current waveform, and the detection step (S20) detects a waveform imbalance based on the real-time waveform measured and collected from the power unit, and the judgment step (S30) is characterized in that it detects an abnormality sign of the device based on the waveform imbalance and threshold value detected in the detection step (S20).

[0015] As described above, according to the predictive maintenance method of a device using the current waveform imbalance of multiple power units according to the present invention, real-time current waveforms are collected based on the current values ​​consumed in each power unit of the device, and the waveform imbalance is continuously detected by superimposing the collected real-time current waveforms on each other, thereby detecting mechanical abnormalities of each power unit, and at the same time, various abnormalities such as defects in components related to the power unit of the device or errors in the control system that controls the power unit can be detected together, thereby having the effect of performing stable predictive maintenance of a device including multiple power units that operate identically.

[0016] FIG. 1 is a block diagram of a method for predictive maintenance of a device through current waveform imbalance of a multi-power unit according to an embodiment of the present invention.

[0017] Figures 2 to 8 are drawings for explaining a predictive maintenance method of a device through the current waveform imbalance of the multiple power units illustrated in Figure 1.

[0018] A method for predictive maintenance of a device using current waveform imbalance of multiple power units according to a preferred embodiment of the present invention is described in detail with reference to the attached drawings. Detailed descriptions of well-known functions and configurations that may unnecessarily obscure the gist of the present invention are omitted.

[0019] FIGS. 1 to 8 illustrate a predictive maintenance method for a device using a current waveform imbalance of a multi-power unit according to an embodiment of the present invention. FIG. 1 is a block diagram of a predictive maintenance method for a device using a current waveform imbalance of a multi-power unit according to an embodiment of the present invention, and FIGS. 2 to 8 are drawings each illustrating a predictive maintenance method for a device using a current waveform imbalance of a multi-power unit illustrated in FIG. 1.

[0020] As illustrated in FIG. 1, the predictive maintenance method (100) of a device using current waveform imbalance of a multi-power unit according to an embodiment of the present invention includes an information collection step (S10), a detection step (S20), and a judgment step (S30).

[0021] The above information collection step (S10) is a step for continuously measuring and collecting real-time current waveforms that represent the current value consumed by each power unit over time during the operation of a device including at least two power units that operate identically, when the device operates in real time.

[0022] Examples of devices that typically have two or more power units that operate identically include a gantry crane, which has a drive motor that provides equal power to each wheel on both sides for linear forward and backward movement, or a train, which has a brake device that provides braking power to each wheel on both sides of the train for braking.

[0023] Of course, in a gantry crane, the power unit would be the drive motor that provides power to the wheels, and in a train, the power unit would be the brake device.

[0024] For convenience of explanation, the method (100) for predictive maintenance of a device through current waveform imbalance of multiple power units of the present invention will be explained based on a gantry crane in which one power unit (drive motor) is formed on each side to provide power to the wheels on both sides, but the technology of the present invention is not limited to such devices.

[0025] That is, when real-time current waveforms are measured and collected from each power unit of the gantry crane in the real-time operation process in the above information collection step (S10), the real-time current waveforms of each power unit are continuously collected as shown in Fig. 2, and it can be seen that real-time current waveforms of similar shapes are measured due to the characteristic of each power unit operating in the same manner.

[0026] The above detection step (S20) is a step that continuously detects the waveform imbalance (Unbalance Score) while the device is performing an operation by superimposing the real-time current waveforms of each power unit that are continuously measured and collected in the above information collection step (S10).

[0027] Here, the waveform imbalance indicates the degree of similarity of real-time current waveforms that are overlapped with each other, and the degree of similarity is indicated by a number from 0 to 100 depending on the similarity of the real-time current waveforms. The lower the number, the higher the similarity of the real-time current waveforms, and the higher the number, the lower the similarity of the real-time current waveforms.

[0028] Accordingly, if the value of the waveform imbalance is "0", the real-time current waveforms are completely identical to each other, and if the value of the waveform imbalance is "100", the real-time current waveforms are completely different from each other.

[0029] Here, the numerical range representing the waveform imbalance is set to 0 to 100, with a lower number indicating high similarity and a higher number indicating low similarity, but it is of course possible to set the numerical range representing the waveform imbalance or the meaning of similarity of a low number or a high number in various ways.

[0030] That is, the above detection step (S20) can continuously detect waveform imbalance by superimposing real-time current waveforms measured and collected from each power unit of the gantry crane during a real-time operation process. As shown in FIG. 3, if the real-time current waveforms measured and collected from each power unit are similar to each other, the numerical value of the continuously detected waveform imbalance is detected to be very low.

[0031] Of course, in theory, since each power unit of the gantry crane operates identically, the same real-time current waveform should be measured and collected, and the waveform imbalance value should be detected as "0". However, due to various variables such as the period of use, operating status, and operating conditions of each power unit, the value may be detected as various values.

[0032] If the real-time current waveforms measured and collected from each power unit, as shown in Fig. 4, are dissimilar to each other, a high value of waveform imbalance is detected.

[0033] The above judgment step (S30) is a step for setting a threshold value for the waveform imbalance, and if the waveform imbalance continuously detected in the detection step (S20) exceeds the threshold value, the device is judged to be in an abnormal state and an alarm is issued.

[0034] That is, as illustrated in Fig. 5, in the judgment step (S30), if the real-time current waveforms measured and collected from each power unit of the gantry crane during the real-time operation process are similar to each other and the continuously detected waveform imbalance is formed to be less than the threshold value, each power unit of the gantry crane will be judged to be normal.

[0035] Here, the threshold value is arbitrarily set to a value of 30 to determine the status of the device, but it is obvious that this threshold value can be set to various values ​​considering the type of device, working environment, usage conditions, etc.

[0036] If, as shown in Fig. 6, real-time current waveforms measured and collected from each power unit of the gantry crane during a real-time operation process are dissimilar to each other and the numerical value of any continuously detected waveform imbalance exceeds a threshold value, the power unit of the gantry crane is judged to be abnormal and an alarm is sent to the surroundings to immediately stop the gantry crane in operation so that rapid inspection and management of the gantry crane can be performed.

[0037] Here, various abnormal signs can be suspected and detected as alarms through the above judgment step (S30).

[0038] First, when various mechanical problems occur in the power unit, abnormal current waveforms are measured and collected from the power unit with the problem, and the waveform imbalance is formed with a high degree of imbalance, so abnormal signs in the power unit can be quickly detected.

[0039] Second, if deformation or damage occurs in the rail that guides the wheels of the gantry crane, an abnormal load is generated on the wheels that are transported along the rail, and an abnormal current waveform is measured and collected from the power unit that provides power to the wheels, so that abnormal signs in the rail can be quickly detected through inspection.

[0040] Third, as shown in Fig. 7, even if the power unit is in a normal state but operates slightly faster or slower than other power units due to a control system (program) error, the waveform imbalance will be high, so that the occurrence of an error in the control system can be quickly detected.

[0041] Therefore, the predictive maintenance method (100) of the device through the current waveform imbalance of the multi-power unit of the present invention can easily detect various abnormal signs that may occur in the device, the components related to the operation of the device (for example, a rail guiding the movement of a gantry crane), and the control system that controls the device, thereby inducing very stable operation management of the device.

[0042] Meanwhile, in the information collection step (S10), the real-time current waveform of each power unit measured and collected is forcibly divided into a set time interval, and the divided section is set as a unit division section to form a plurality of unit division sections on the real-time current waveform, and a forced division step (S40) is further included.

[0043] The above detection step (S20) continuously detects waveform imbalance by superimposing real-time current waveforms of each power unit, and continuously detects waveform imbalance for real-time current waveforms of multiple unit sections divided in the above forced division step (S40).

[0044] The above judgment step (S30) determines that the device is in an abnormal state and issues an alarm if the waveform imbalance of the real-time current waveform of each unit section independently detected in the above detection step (S20) exceeds the threshold value.

[0045] Here, if the power unit operates without rest, the current value consumed from the power unit will be continuously measured, and the real-time current waveform will be formed for a very long time. When detecting waveform imbalance for the long-formed real-time current waveform, even if the current waveforms are dissimilar to each other in a short period of time in some waveforms, if the entire current waveforms are similar to each other, the dissimilar portion is only a part of the whole, and thus a problem may arise in which the dissimilarity for the current waveforms formed dissimilarly in a short period of time is not properly reflected.

[0046] In order to solve this problem, as illustrated in FIG. 8, the partition step (S40) forcibly partitions the real-time current waveform continuously formed from each power unit of the gantry crane operating without a pause according to a set time interval to construct a plurality of unit partition sections, and the detection step (S20) independently continuously detects the waveform imbalance of the real-time current waveform for each of the plurality of unit partition sections constructed as described above, and the judgment step (S30) determines the gantry crane to be in an abnormal state and issues an alarm if any waveform imbalance among the waveform imbalances of the real-time current waveforms of each unit partition section independently detected in the detection step (S20) exceeds a threshold value, thereby effectively detecting abnormal signs of the equipment operating without a pause.

[0047] That is, as described above, by dividing a real-time current waveform formed over a long period of time into a number of unit segments through the segmentation step (S40), and detecting the waveform imbalance for the real-time current waveform of each of the segmented unit segments, the dissimilarity of current waveforms formed dissimilarly in a relatively short period of time can be accurately reflected as the waveform imbalance, thereby inducing more effective detection of abnormal signs of the device.

[0048] For example, when a gantry crane moves along a partially damaged rail, a current waveform different from that of other power units will be generated momentarily in the power unit that provides power to the wheels moving along the rail. The waveform imbalance for this momentary dissimilar current waveform can be easily reflected, allowing the manager to quickly and accurately recognize and respond to damage to the rail.

[0049] Meanwhile, in the above information collection step (S10), instead of the current value measured and collected from the power unit, one of the vibrations generated from the power unit, noise, frequency of the power supplied to the power unit, temperature, humidity, and pressure of the power unit is selected to measure and collect real-time waveforms such as real-time current waveforms.

[0050] The above detection step (S20) detects waveform imbalance based on real-time waveforms measured and collected from the power unit, and the above judgment step (S30) detects abnormal signs of the device based on the waveform imbalance and threshold value detected in the above detection step (S20), but it is of course possible to detect abnormal signs of the device as described above.

[0051] The predictive maintenance method (100) of the present invention for a device using a current waveform imbalance of multiple power units, which is performed through the above process, collects real-time current waveforms based on the current values ​​consumed in each power unit of the device, and detects waveform imbalances continuously by superimposing the collected real-time current waveforms on each other, thereby detecting mechanical abnormalities of each power unit, and at the same time detecting various abnormalities such as defects in components related to the power unit of the device or errors in the control system that controls the power unit, thereby having the effect of performing stable predictive maintenance of a device including multiple power units that operate identically.

[0052] While the present invention has been described with reference to the embodiments illustrated in the accompanying drawings, these are merely illustrative and are not limited to the above-described embodiments. Those skilled in the art will readily appreciate that various modifications and equivalent embodiments are possible. Furthermore, it should be understood that modifications are possible by those skilled in the art without departing from the spirit and scope of the present invention. Therefore, the scope of the claims in the present invention is not limited to the detailed description, but rather by the claims set forth below and their technical spirit.

[0053] The present invention can be used in the predictive maintenance industry of equipment.

Claims

1. In a predictive correction method for a device including at least two power units that operate identically, When the device operates in real time, an information collection step (S10) that continuously measures and collects real-time current waveforms that represent the current values ​​consumed in each power unit over time during the operation of the device; A detection step (S20) for continuously detecting waveform imbalance (Unbalance Score) while the device is performing operation by superimposing real-time current waveforms of each power unit continuously measured and collected in the above information collection step (S10); and A predictive maintenance method for a device using current waveform imbalance of a multi-power unit, characterized by including a judgment step (S30) of setting a threshold value for the waveform imbalance and determining that the device is in an abnormal state and giving an alarm when the waveform imbalance continuously detected in the detection step (S20) exceeds the threshold value.

2. In paragraph 1, In the above information collection step (S10), the real-time current waveform of each power unit measured and collected is forcibly divided according to a set time interval, and the divided section is set as a unit division section to form a plurality of unit division sections on the real-time current waveform, and further includes a forced division step (S40); The above detection step (S20) continuously detects waveform imbalance by superimposing real-time current waveforms of each power unit, and continuously detects waveform imbalance for real-time current waveforms of multiple unit sections divided in the above forced division step (S40). A method for predictive maintenance of a device through current waveform imbalance of a multi-power unit, characterized in that the above judgment step (S30) determines the device to be in an abnormal state and issues an alarm when the waveform imbalance of the real-time current waveform of each unit section independently detected in the above detection step (S20) exceeds a threshold value.

3. In paragraph 1, In the above information collection step (S10), instead of the current value measured and collected from the power unit, one of the vibration, noise generated from the power unit, the frequency of the power supplied to the power unit, the temperature, humidity, and pressure of the power unit is selected to measure and collect a real-time waveform such as a real-time current waveform. A method for predictive maintenance of a device through current waveform imbalance of multiple power units, characterized in that the detection step (S20) detects waveform imbalance based on real-time waveforms measured and collected from a power unit, and the judgment step (S30) detects abnormal signs of the device based on the waveform imbalance detected in the detection step (S20) and a threshold value.

Images