Predictive maintenance method of equipment using deviation

Abstract

The present invention relates to a predictive maintenance method for a device using deviation, wherein the configuration comprises: an information collection step (S10) for repeatedly measuring energy values ​​at set unit time intervals in an energy waveform representing information on changes in energy magnitude over time measured in a device in an operating state; a difference value collection step (S20) for repeatedly collecting difference values ​​for each energy value collected in the information collection step (S10) by calculating the difference between each energy value collected in the information collection step (S10) and a previous energy value; a setting step (S30) for setting a difference threshold for the difference values ​​of energy values ​​based on the difference values ​​collected in the difference value collection step (S20); and a detection step (S40) for repeatedly measuring energy values ​​at set unit time intervals in an energy waveform measured in a real-time operating state of the device, calculating a difference value for the measured energy value, and issuing an alarm if the calculated difference value exceeds the difference threshold to induce inspection management of the device. By calculating and collecting the difference between each energy value measured at unit time intervals and the previous energy value in the energy waveform over time measured by the device, and setting a threshold value based on the collected difference value information, and by detecting and alarming the device as somewhat unstable when the difference value for the energy value measured and collected in the real-time operating state of the device exceeds the threshold value, thereby inducing management such as replacement or repair in advance before the device fails, there is an effect of preventing economic losses caused by the suspension of equipment operation due to sudden device failure. In addition, the difference values ​​regarding the energy values ​​collected from the energy waveform measured by the device can effectively capture trends of abnormal signs, such as minute fluctuations in energy values ​​prior to device failure, thereby enabling the accurate and earlier detection of abnormal signs. In addition, by presenting various detection conditions to efficiently search for abnormal signs occurring in the device and detecting the device as an abnormal state when those conditions are satisfied, it is possible to detect abnormal signs occurring in the device with great precision and effectiveness, as well as ensure excellent reliability of the detection results.

Description

Technology Field

[0001] The present invention relates to a predictive maintenance method for a device using deviations, and more specifically, to a predictive maintenance method for a device using deviations that can prevent economic losses caused by the suspension of operation of equipment due to sudden device failure by calculating and collecting the difference between each energy value measured at unit time intervals and the previous energy value in an energy waveform according to the flow of time measured by the device, setting a threshold value based on the collected difference value information, and detecting and alarming the device as being in a somewhat unstable state when the difference value for the energy value measured and collected in the real-time operating state of the device exceeds the threshold value, thereby inducing management such as replacement or repair in advance before the device fails. Background Technology

[0002] In general, stable operation is very important for the various devices used in the automation process of facilities.

[0003] For example, in large-scale production facilities, dozens or hundreds of machines are installed and operate in conjunction with each other to continuously produce products. However, if any one of the machines fails, a massive situation may occur where the entire operation of the facility comes to a halt.

[0004] In such cases, significant losses are inevitable due to downtime caused by equipment failure, resulting not only in repair costs but also in wasted operating expenses and lost business performance during the shutdown.

[0005] According to recent data from the Ministry of Employment and Labor and the Korea Occupational Safety and Health Agency, the number of casualties from industrial safety accidents annually is estimated at around 100,000, and when converted into costs, this results in an annual loss of 18 trillion won.

[0006] The introduction of a predictive maintenance system is urgently needed to avoid such unexpected downtime costs. Although efforts are already being made to address issues under the guise of predictive maintenance, the development of more advanced methods is required for more efficient maintenance. Prior art literature

[65535] [Reference 0001] Republic of Korea Published Patent Application No. 10-2019-0108282 (September 24, 2019) [Reference 0002] Republic of Korea Published Patent Application No. 10-2019-0108273 (September 24, 2019) [Reference 0003] Republic of Korea Published Patent Application No. 10-2021-0044655 (April 23, 2021) The problem to be solved

[0007] The present invention is proposed to solve the aforementioned problems, and its purpose is to provide a predictive maintenance method for equipment using deviations that can prevent economic losses caused by the suspension of equipment operation due to sudden equipment failure by calculating and collecting the difference between each energy value measured at unit time intervals and the previous energy value in the energy waveform according to the flow of time measured by the equipment, setting a threshold value based on the collected difference value information, and detecting and alarming the equipment as somewhat unstable when the difference value for the energy value measured and collected in the real-time operating state of the equipment exceeds the threshold value, thereby inducing management such as replacement or repair in advance before the equipment failure occurs.

[0008] Furthermore, the invention provides a predictive maintenance method for a device utilizing deviations, which enables the early and accurate detection of abnormal signs by effectively capturing trends of abnormal signs, such as minute fluctuations in energy values ​​prior to device failure, based on the difference values ​​collected from the energy waveforms measured by the device.

[0009] Furthermore, the invention provides a predictive maintenance method for a device using deviation that can not only detect abnormal signs occurring in the device with great precision and effectiveness but also ensure excellent reliability of the detection results by presenting various detection conditions to efficiently search for abnormal signs occurring in the device and detecting the device as an abnormal state when those detection conditions are satisfied. means of solving the problem

[0010] A predictive maintenance method for a device using deviation according to the present invention for achieving the above-mentioned purpose comprises: an information collection step (S10) for repeatedly measuring energy values ​​at set unit time intervals in an energy waveform representing information on changes in energy magnitude over time measured in a device in an operating state; a difference value collection step (S20) for repeatedly collecting difference values ​​for each energy value collected in the information collection step (S10) by calculating the difference between each energy value and a previous energy value; a setting step (S30) for setting a difference threshold for the difference values ​​of energy values ​​based on the difference values ​​for energy values ​​collected in the difference value collection step (S20); and a detection step (S40) for repeatedly measuring energy values ​​at set unit time intervals in an energy waveform measured in a real-time operating state of the device, calculating a difference value for the measured energy value, and issuing an alarm if the calculated difference value exceeds the difference threshold to induce inspection management of the device.The above difference value collection step (S20) sequentially arranges the difference values ​​for energy values ​​that are repeatedly collected over time and repeatedly connects other difference values ​​adjacent to the arranged difference values ​​with straight lines to repeatedly collect the slope of the straight line for the difference values; the above setting step (S30) sets a slope threshold for the slope of the difference values ​​based on the slope information for the difference values ​​collected in the above difference value collection step (S20); the above detection step (S40) repeatedly calculates and collects the difference values ​​for energy values ​​measured at set unit time intervals in the energy waveform measured in the real-time operating state of the device and arranges the collected difference values ​​for energy values ​​over time, and if the slope value of the straight line connecting the arranged difference values ​​and other difference values ​​adjacent to the device exceeds the slope threshold set in the above setting step (S30), an alarm is issued to induce inspection management of the device; and the above detection step (S40) includes at least two slopes of the straight line connecting the difference values ​​for energy values ​​that are repeatedly detected in the real-time operating state of the device, and the slope of the straight line connecting the difference values ​​for energy values ​​that are repeatedly detected in the real-time operating state of the device, and the slope of the straight line at unit time. The detection section is repeatedly partitioned, and if the average slope value obtained by averaging the straight slopes included in the partitioned slope detection section exceeds the slope threshold value set in the setting step (S30), an alarm is issued to induce inspection management of the device.

[0011] delete

[0012] delete

[0013] delete

[0014] delete

[0015] In addition, the difference value collection step (S20) repeatedly divides a distribution detection section of a unit time that includes a number of difference values ​​for energy values ​​that are repeatedly collected, and repeatedly constructs a difference value distribution map using the difference values ​​included in the divided distribution detection section.

[0016] delete

[0017] delete

[0018] delete

[0019] The above setting step (S30) arbitrarily sets a safe interval for the distribution probability of the difference value in the difference value distribution map collected and constructed in the above difference value collection step (S20), and sets one or more intervals selected from intervals other than the set safe interval as an alarm interval.

[0020] The detection step (S40) induces inspection management of the device by issuing an alarm if the difference value for the energy value measured at a set unit time interval in the energy waveform measured in the real-time operating state of the device corresponds to the alarm section of the difference value distribution map set in the setting step (S30); the setting step (S30) sets a distribution threshold for the distribution probability of the alarm section in the difference value distribution map based on the information of the difference value distribution map collected and constructed in the difference value collection step (S20); the detection step (S40) repeatedly divides a distribution detection section containing a number of difference values ​​of energy values ​​repeatedly detected in the real-time operating state of the device, and induces inspection management of the device if the distribution probability of the alarm section of the real-time difference value distribution map constructed based on the difference values ​​of the divided distribution detection section exceeds the distribution threshold set in the setting step (S30).

[0021] delete

[0022] delete

[0023] delete Effects of the invention

[0024] As described above, according to the predictive maintenance method for a device using deviation according to the present invention, the difference between each energy value measured at unit time intervals in the energy waveform over time measured by the device and the previous energy value is calculated and collected, and a threshold value is set based on the collected difference value information. If the difference value for the energy value measured and collected in the real-time operating state of the device exceeds the threshold value, the device is detected and alarmed as being in a somewhat unstable state, thereby inducing management such as replacement or repair in advance before the device fails, so as to prevent economic losses caused by the suspension of operation of the equipment due to a sudden device failure.

[0025] In addition, the difference values ​​regarding the energy values ​​collected from the energy waveform measured by the device can effectively capture trends of abnormal signs, such as minute fluctuations in energy values ​​prior to device failure, thereby enabling the accurate and earlier detection of abnormal signs.

[0026] In addition, by presenting various detection conditions to efficiently search for abnormal signs occurring in the device and detecting the device as an abnormal state when those conditions are satisfied, it is possible to detect abnormal signs occurring in the device with great precision and effectiveness, as well as ensure excellent reliability of the detection results. Brief explanation of the drawing

[0027] FIG. 1 is a block diagram of a predictive maintenance method for a device using deviation according to an embodiment of the present invention. FIGS. 2 to 13 are drawings for explaining a predictive maintenance method of a device using the deviation shown in FIG. 1. Specific details for implementing the invention

[0028] A method for predictive maintenance of a device using deviation according to a preferred embodiment of the present invention will be described in detail with reference to the attached drawings. Detailed descriptions of known functions and configurations that are deemed to unnecessarily obscure the gist of the present invention are omitted.

[0029] FIGS. 1 to 13 illustrate a predictive maintenance method for a device using deviation according to an embodiment of the present invention, FIG. 1 is a block diagram of a predictive maintenance method for a device using deviation according to an embodiment of the present invention, and FIGS. 2 to 13 are drawings for explaining the predictive maintenance method for a device using deviation illustrated in FIG. 1.

[0030] As illustrated in FIG. 1, the predictive maintenance method (100) of a device using deviation according to an embodiment of the present invention includes an information collection step (S10), a difference value collection step (S20), a setting step (S30), and a detection step (S40).

[0031] The above information collection step (S10) is a step of repeatedly measuring energy values ​​at set unit time intervals in an energy waveform representing information on changes in energy magnitude over time measured by a device in an operating state.

[0032] Here, the energy measured and collected from the device operating in the information collection step (S10) may be selected from the current consumed for operating the device, vibration generated during device operation, noise generated during device operation, frequency of the power supplied to the device, and the temperature, humidity, and pressure of the device during device operation. In the predictive maintenance method (100) of the device using deviation of the present invention, the current consumed for operating the device is selected and used as the energy measured from the device, but it is not limited to selecting and using only such a current.

[0033] That is, in the above information collection step (S10), if information on the change in the magnitude (value) of the current consumed by the device in the operating state is collected over time, it can be collected as a current (energy) waveform as shown in FIG. 2, and the current value (energy value) is repeatedly measured and collected at unit time intervals from this current waveform.

[0034] For example, if the unit time interval is set to 1 minute, the current value is repeatedly measured and collected from the current waveform at 1-minute intervals, and the information of the current value collected in this way serves as the basis for calculating and collecting the difference value to detect abnormal signs of the device in the difference value collection step (S20) to be described later.

[0035] Here, it goes without saying that the unit time interval for measuring the current value can be set to various time intervals by considering various factors such as the type of device, the surrounding environment and conditions in which the device is used.

[0036] In addition, for the convenience of explanation, the current waveform measured and collected from the device in the information collection step (S10) is measured and collected from the device in a state where the current value continuously maintains a constant range, that is, when the operation of the device has started and the operation of the device has stabilized; however, it is obvious that the current waveform can be measured and collected in various states of the device.

[0037] The difference value collection step (S20) is a step of repeatedly collecting difference values ​​for each energy value by calculating the difference between each energy value collected in the information collection step (S10) and the previous energy value.

[0038] That is, if the current values ​​collected at unit time intervals set in the information collection step (S10) in FIG. 3 are sequentially referred to as the first current value, the second current value, ..., the nth current value, the first current value is excluded because there is no previous current value, the difference between the second current value and the previous first current value is calculated and the difference value is collected, the difference between the third current value and the previous second current value is calculated and the difference value is collected, and the difference values ​​for the fourth current value, the fifth current value, ..., the nth current value are each calculated and collected, and the difference values ​​of each current value calculated and collected in this way serve as the basis for the difference threshold value set to detect abnormal signs of the device in the setting step (S30).

[0039] Here, the difference value for each current value calculated and collected in the difference value collection step (S20) may be a positive or negative value, but for the convenience of explanation, all are converted into absolute values ​​and collected, and in the setting step (S30) to be described later, a difference threshold value is set for the difference value converted into an absolute value.

[0040] The above setting step (S30) is a step of setting a difference threshold for the difference value of energy values ​​based on the difference value of energy values ​​collected in the above difference value collection step (S20).

[0041] That is, the difference threshold for the difference value is set based on the difference value of the current value that changes abnormally before the device fails, for example, in situations such as device deterioration, aging, or load caused by foreign matter obstruction, based on the difference value information for each current value collected through the information collection step (S10) and the difference value collection step (S20).

[0042] Here, it goes without saying that the threshold can be set by distinguishing between at least two thresholds, such as an alarm threshold and a danger threshold, to create various levels of alarms, thereby enabling more precise warning of abnormal signs in the device.

[0043] The detection step (S40) is a step of repeatedly measuring energy values ​​at set unit time intervals from an energy waveform measured in the real-time operating state of the device, calculating a difference value for the measured energy values, and issuing an alarm if the calculated difference value exceeds the difference threshold value to induce inspection management of the device.

[0044] That is, as illustrated in FIG. 4, the current value is repeatedly measured and detected at set unit time intervals from the current (energy) waveform measured and collected from the device in real time, and the difference value for the detected current value is repeatedly calculated and collected. If the calculated difference value does not exceed the difference threshold value set in the setting step (S30), the device is detected in a stable state. Conversely, if the real-time difference value of the device exceeds the difference threshold value, the device is detected and alarmed as being in a somewhat unstable state, thereby inducing management such as replacement or repair before the device fails, so that economic losses caused by the suspension of operation of the equipment due to a sudden device failure can be prevented in advance.

[0045] Here, the difference value for each current value represents the difference between the current value and the previous current value. This effectively reflects the unstable state where the current (energy) value fluctuates slightly before the device fails, thereby enabling the early and accurate detection of abnormal signs prior to device failure through the predictive maintenance method described above.

[0046] Meanwhile, the difference value collection step (S20) sequentially arranges the difference values ​​for the energy values ​​that are repeatedly collected over time, and repeatedly connects other difference values ​​adjacent to the arranged difference values ​​with a straight line to repeatedly collect the slope of the straight line for the difference values.

[0047] That is, as shown in Fig. 5, a predetermined slope value can be obtained by connecting the difference values ​​of each current value with a straight line. This slope value may include an upward slope value (positive) where the slope rises and a downward slope value (negative) where the slope falls, but all slope values ​​are collected by quantifying them as absolute values.

[0048] The slope information regarding the difference values ​​of each current value collected in this way serves as the basis for the slope threshold value set to detect abnormal signs of the device in the setting step (S30).

[0049] Then, the setting step (S30) sets a slope threshold for the slope of the difference value based on the slope information for the difference value collected in the difference value collection step (S20), and

[0050] The detection step (S40) repeatedly calculates and collects difference values ​​for energy values ​​measured at set unit time intervals from energy waveforms measured in the real-time operating state of the device, arranges the collected difference values ​​according to the flow of time, and then, if the slope value of a straight line connecting the arranged difference value and other adjacent difference values ​​exceeds the slope threshold set in the setting step (S30), an alarm is issued to induce inspection management of the device.

[0051] That is, as illustrated in FIG. 6, the difference values ​​for each current value measured and collected from the real-time device are sequentially arranged over time, and the arranged difference values ​​are connected to each other in a straight line to detect the slope value of the straight line that repeatedly connects the difference values. When the detected slope value of the straight line exceeds the slope threshold value set in the setting step (S30), the device is detected as being in a somewhat unstable state and an alarm is issued, thereby inducing management such as replacement or repair before the device fails, so that economic losses caused by the suspension of operation of the equipment due to a sudden device failure can be prevented in advance.

[0052] Here, it goes without saying that the slope values ​​for the difference between each current value measured and collected from the real-time device are all detected as absolute values ​​and compared with the slope threshold.

[0053] In addition, as illustrated in FIG. 7, the detection step (S40) can repeatedly divide the slope detection section of a unit time, which includes at least two slopes of a straight line connecting the difference values ​​of energy values ​​that are repeatedly detected in the real-time operating state of the device, and if the average slope value obtained by averaging the straight line slopes included in the divided slope detection section exceeds the slope threshold value set in the setting step (S30), an alarm can be issued to induce inspection management of the device.

[0054] Here, for the sake of convenience of explanation, the average slope value of the slope detection section is compared with a slope threshold value pre-set in the setting step (S30) to detect abnormal signs of the device, but it is obvious that a separate slope threshold value can be independently set and used in the setting step (S30) so as to be applied only to the average slope value.

[0055] Meanwhile, the difference value collection step (S20) sequentially arranges the difference values ​​for the energy values ​​that are repeatedly collected over time, and repeatedly divides the area detection section of a unit time that includes at least two of the arranged difference values, and detects and collects the integrated area value of the divided area detection section.

[0056] That is, as shown in FIG. 8, the area inside the waveform connecting the difference values ​​included in the area detection section is measured, and the integrated area value of the area detection section is repeatedly detected and collected.

[0057] The integrated area value information of the area detection section collected in this way serves as the basis for the area threshold value set to detect abnormal signs of the device in the setting step (S30).

[0058] Then, the setting step (S30) sets an area threshold for the integrated area value based on the integrated area value information for the difference value of the area detection section collected in the difference value collection step (S20).

[0059] The detection step (S40) repeatedly calculates and collects difference values ​​for energy values ​​measured at set unit time intervals from energy waveforms measured in the real-time operating state of the device, arranges the collected difference values ​​according to the flow of time, and then, if the integrated area value of an area detection section containing at least two of the arranged difference values ​​exceeds the area threshold value set in the setting step (S30), an alarm is issued to induce inspection management of the device.

[0060] That is, as illustrated in FIG. 9, an area detection section containing at least two difference values ​​of each current value measured and collected from a real-time device is repeatedly partitioned, and if the integrated area value of the partitioned real-time area detection section exceeds the area threshold value set in the setting step (S30), the device is detected and alarmed as being in a somewhat unstable state, thereby inducing management such as replacement or repair before a device failure occurs, so as to prevent economic loss caused by the suspension of operation of the equipment due to a sudden device failure.

[0061] Meanwhile, the difference value collection step (S20) repeatedly divides a distribution detection section of a unit time that includes a number of difference values ​​for energy values ​​that are repeatedly collected, and repeatedly constructs a difference value distribution map using the difference values ​​included in the divided distribution detection section.

[0062] That is, if a distribution map is constructed based on a plurality of difference values ​​included in the distribution detection interval, a difference value distribution map as shown in FIG. 10 can be constructed.

[0063] Here, the difference value distribution map for the distribution detection section is repeatedly constructed, and the constructed difference value distribution map information serves as the basis for setting safety and alarm sections of the difference value distribution map to detect abnormal signs of the device in the setting step (S30).

[0064] Then, the setting step (S30) arbitrarily sets a section with a high probability of distribution of difference values ​​in the difference value distribution map collected and constructed in the difference value collection step (S20) as a safe section, and sets one or more sections selected from sections other than the set safe section as an alarm section.

[0065] Here, as shown in FIG. 11, the difference value corresponding to the safe zone with a high probability of distribution in the difference value distribution diagram can be seen as a value indicating a somewhat stable state of the device, and the difference value corresponding to the alarm zone with a low probability of distribution can be seen as a value indicating a somewhat unstable state of the device.

[0066] Therefore, if the above difference value distribution is divided into a safe zone and an alarm zone, the safe zone is the area where difference values ​​are distributed when the device is in a stable state, and the alarm zone is the area where difference values ​​are distributed when the device is in a somewhat unstable state.

[0067] Here, the above warning section is selected as all sections other than the above safety section, that is, sections on both sides of the above safety section; however, it is not the case that the above warning section is set to be limited to the sections selected in this way.

[0068] Then, the detection step (S40) induces inspection management of the device by issuing an alarm if the difference value for the energy value measured at a set unit time interval in the energy waveform measured in the real-time operating state of the device corresponds to the alarm section of the difference value distribution map set in the setting step (S30).

[0069] That is, as illustrated in FIG. 12, if the difference value for each current value measured and collected from the device in real time corresponds to the alarm section of the difference value distribution map set in the setting step (S30), the device is detected and alarmed as being in a somewhat unstable state, thereby inducing management such as replacement or repair before the device fails, so as to prevent economic losses caused by the suspension of operation of the equipment due to a sudden device failure.

[0070] In addition, the setting step (S30) sets a distribution threshold for the distribution probability of the alarm section in the difference value distribution map based on the information of the difference value distribution map collected and constructed in the difference value collection step (S20).

[0071] Here, the distribution threshold is a value for triggering an alarm when the distribution probability of the alarm interval constructed from the difference value distribution map increases abnormally, and can be set to various probability values ​​considering conditions such as the type of device, usage environment, and lifespan.

[0072] Then, the detection step (S40) repeatedly divides a distribution detection section containing multiple difference values ​​of energy values ​​that are repeatedly detected in the real-time operating state of the device, and if the distribution probability of the alarm section of the real-time difference value distribution map constructed based on the difference values ​​of the divided distribution detection section exceeds the distribution threshold set in the setting step (S30), it induces inspection management of the device.

[0073] That is, as illustrated in FIG. 13, a real-time difference value distribution map is constructed based on the difference value within the distribution detection interval in the real-time operating state of the device, and the real-time difference value distribution map is constructed repeatedly at unit time intervals of the repetitive distribution detection interval. When the distribution probability for the alarm interval of the real-time difference value distribution map constructed at this time exceeds the distribution threshold set in the setting step (S30), the device is detected and alarmed as being in a somewhat unstable state, thereby inducing management such as replacement or repair in advance before a device failure occurs, so as to prevent economic losses caused by the suspension of operation of the equipment due to a sudden device failure.

[0074] For example, Fig. 13 shows that the distribution threshold is set to 10%, and abnormal signs of the device are detected by comparing the distribution probability of the alarm section of the real-time difference value distribution of the device with respect to the set distribution threshold.

[0075] The predictive maintenance method (100) of the present invention, which predicts abnormal signs of the device through the above process, calculates and collects the difference between each energy value measured at unit time intervals and the previous energy value in the energy waveform according to the flow of time measured by the device, and sets a threshold value based on the collected difference value information, so that if the difference value for the energy value measured and collected in the real-time operating state of the device exceeds the threshold value, the device is detected and alarmed as being in a somewhat unstable state, thereby inducing management such as replacement or repair in advance before the device fails, so as to prevent economic loss caused by the suspension of operation of the equipment due to a sudden failure of the device.

[0076] In addition, the difference values ​​regarding the energy values ​​collected from the energy waveform measured by the device can effectively capture trends of abnormal signs, such as minute fluctuations in energy values ​​prior to device failure, thereby enabling the accurate and earlier detection of abnormal signs.

[0077] In addition, by presenting various detection conditions to efficiently search for abnormal signs occurring in the device and detecting the device as an abnormal state when those conditions are satisfied, it is possible to detect abnormal signs occurring in the device with great precision and effectiveness, as well as ensure excellent reliability of the detection results.

[0078] The present invention has been described with reference to embodiments illustrated in the accompanying drawings, but this is illustrative and is not limited to the embodiments described above. Those skilled in the art will understand that various modifications and equivalent embodiments are possible therefrom. Furthermore, modifications by those skilled in the art are possible within the scope that does not deviate from the spirit of the present invention. Accordingly, the scope of the rights claimed in the present invention is not limited to the scope of the detailed description, but is limited by the claims set forth below and the technical concept thereof. Explanation of the symbols

[0079] S10. Information collection stage S20. Difference value collection step S30. Setup Step S40. Detection step 100. Predictive maintenance method for equipment using deviation

Claims

Claim 1 A predictive maintenance method for equipment used in various facilities comprises: an information collection step (S10) for repeatedly measuring energy values ​​at set unit time intervals in an energy waveform representing information on changes in energy magnitude over time measured in the equipment in an operating state; a difference value collection step (S20) for repeatedly collecting difference values ​​for each energy value collected in the information collection step (S10) by calculating the difference between each energy value and a previous energy value; a setting step (S30) for setting a difference threshold for the difference values ​​of energy values ​​based on the difference values ​​for energy values ​​collected in the difference value collection step (S20); and a detection step (S40) for repeatedly measuring energy values ​​at set unit time intervals in an energy waveform measured in the real-time operating state of the equipment, calculating a difference value for the measured energy value, and issuing an alarm if the calculated difference value exceeds the difference threshold to induce inspection management of the equipment.The above difference value collection step (S20) sequentially arranges the difference values ​​for energy values ​​that are repeatedly collected over time and repeatedly connects other difference values ​​adjacent to the arranged difference values ​​with straight lines to repeatedly collect the slope of the straight line for the difference values; the above setting step (S30) sets a slope threshold for the slope of the difference values ​​based on the slope information for the difference values ​​collected in the above difference value collection step (S20); the above detection step (S40) repeatedly calculates and collects the difference values ​​for energy values ​​measured at set unit time intervals in the energy waveform measured in the real-time operating state of the device and arranges the collected difference values ​​for energy values ​​over time, and if the slope value of the straight line connecting the arranged difference values ​​and other difference values ​​adjacent to the device exceeds the slope threshold set in the above setting step (S30), an alarm is issued to induce inspection management of the device; and the above detection step (S40) includes at least two slopes of the straight line connecting the difference values ​​for energy values ​​that are repeatedly detected in the real-time operating state of the device, and the slope of the straight line connecting the difference values ​​for energy values ​​that are repeatedly detected in the real-time operating state of the device, and the slope of the straight line at unit time. A predictive maintenance method for a device using deviation, characterized by repeatedly dividing detection sections and averaging the linear slopes included in the divided slope detection sections to induce inspection management of the device by issuing an alarm when the average slope value exceeds a slope threshold set in the setting step (S30). Claim 2 delete Claim 3 delete Claim 4 delete Claim 5 In claim 1, the difference value collection step (S20) repeatedly divides a distribution detection section of a unit time that includes a number of difference values ​​for energy values ​​collected repeatedly, and repeatedly constructs a difference value distribution map using the difference values ​​included in the divided distribution detection section; the setting step (S30) arbitrarily sets a safety section for the distribution probability of difference values ​​in the difference value distribution map collected and constructed in the difference value collection step (S20), and sets one or more sections selected from sections other than the set safety section as an alarm section; the detection step (S40) induces inspection management of the device by issuing an alarm if the difference value for energy values ​​measured at set unit time intervals in the energy waveform measured in the real-time operating state of the device corresponds to the alarm section of the difference value distribution map set in the setting step (S30); the setting step (S30) sets a distribution threshold value for the distribution probability of the alarm section in the difference value distribution map based on the information of the difference value distribution map collected and constructed in the difference value collection step (S20); and A predictive maintenance method for a device using deviations, characterized in that the detection step (S40) repeatedly divides a distribution detection section containing a number of difference values ​​of energy values ​​that are repeatedly detected in the real-time operating state of the device, and if the distribution probability of the alarm section of the real-time difference value distribution map constructed based on the difference values ​​of the divided distribution detection section exceeds the distribution threshold set in the setting step (S30). Claim 6 delete Claim 7 delete

Images