Method and device for detecting abnormality of an apparatus

Abstract

A method and apparatus for detecting equipment malfunctions. The apparatus for detecting equipment malfunctions includes: memory and a processor. The processor loads instructions from the memory to perform the aforementioned method for detecting equipment malfunctions. The method for detecting equipment malfunctions includes: providing the operating speed frequencies and operating power of a plurality of devices; calculating the raw scores of components based on the critical frequencies of the components of the devices; calculating a fitting model using curve fitting based on the operating speed frequencies, operating power, and raw scores of the components of the devices; providing the operating speed frequencies and operating power of a target device; calculating the raw scores of the components based on the critical frequencies of the components of the target device; and calculating an evaluation score of the components using the fitting model based on the operating speed frequencies, operating power, and raw scores of the components of the target device.

Description

Technical Field

[0001] This invention relates to a method and apparatus for detecting equipment malfunctions. More particularly, it relates to a method and apparatus for detecting equipment malfunctions using a physical model of vibration signals. Prior Technology

[0002] With the development of technology and the economy, various production equipment has been manufactured to provide diverse production functions to meet user needs. Generally, production equipment automatically detects abnormalities to alert users to damaged or malfunctioning parts. However, current inspection mechanisms mostly only issue warnings after parts have failed. For example, current inspection mechanisms can only detect abnormalities when parts have aged to the point of failure. However, when parts are severely aged but not yet damaged, the production equipment may still produce substandard products. The current inspection mechanisms cannot accurately assess the health of equipment parts, thus failing to meet user needs. Summary of the Invention

[0003] The purpose of this invention is to provide a method and apparatus for inspecting equipment malfunctions, which can assess the health of equipment parts so that users can understand whether the parts need to be repaired or replaced.

[0004] According to an embodiment of the present invention, a method for detecting equipment anomalies includes a modeling phase and an online evaluation phase. The modeling phase includes: providing a plurality of operating speeds and a plurality of operating powers for a plurality of devices; calculating a plurality of raw scores for a plurality of parts based on a plurality of key frequencies of a plurality of parts of the devices; and calculating a fitting model using curve fitting based on the operating speeds, operating powers, and raw scores of the parts. The online evaluation phase includes: providing the operating speeds and operating powers of a target device; calculating the raw scores of a target part based on the key frequencies of a target part of the target device; and calculating an evaluation score for one of the target parts using the aforementioned fitting model based on the operating speeds, operating powers, and raw scores of the target part of the target device.

[0005] In some embodiments, the original score of each of the above-mentioned parts and the original score of the target part are calculated using the following equation: in For raw fractions, This represents the energy values ​​corresponding to the top 50 frequencies in a given frequency spectrum. This represents the energy value corresponding to the key frequency.

[0006] In some embodiments, the fitting model is the following equation: Where n is the evaluation score, P is the operating power (kW), W is the operating speed and frequency (Hz), and SR is the raw score. These are the coefficients of the equation.

[0007] In some embodiments, when the target part is the main axis, the critical frequency is the frequency with the largest vibration energy value in a frequency range, wherein the error frequency range is the theoretical fundamental frequency of the target part ±3Hz.

[0008] In some embodiments, the above-described method for detecting equipment malfunctions further includes: determining whether the evaluation score of the target part is lower than a preset score threshold; and determining that the target part is malfunctioning when the evaluation score of the target equipment is lower than the preset score threshold.

[0009] According to an embodiment of the present invention, the above-described apparatus for detecting equipment malfunctions includes: a memory and a processor. The memory stores a plurality of instructions. The processor is electrically connected to the memory to load the instructions and perform a modeling phase and an online evaluation phase. The modeling phase includes: providing a plurality of operating speed frequencies and a plurality of operating power for a plurality of devices; calculating a plurality of raw scores based on a plurality of key frequencies of a plurality of parts of the devices; and calculating a fitting model using curve fitting based on the operating speed, operating power, and raw scores of parts of normal equipment. The online evaluation phase includes: providing the operating speed and operating power of a target device; calculating one raw score of a target part based on the key frequencies of a target part of the target device; and calculating one evaluation score of the target part using the above-described fitting model based on the operating speed, operating power, and raw scores of the target part of the target device.

[0010] In some embodiments, the original score of each of the above-mentioned parts and the original score of the target part are calculated using the following equation: in For raw fractions, This represents the energy values ​​corresponding to the top 50 frequencies in a given frequency spectrum. This represents the energy value corresponding to the key frequency.

[0011] In some embodiments, the fitting model is the following equation: Where n is the evaluation score, P is the operating power (kW), W is the operating speed and frequency (Hz), and SR is the raw score. These are the coefficients of the equation.

[0012] In some embodiments, when the target part is the main axis, the critical frequency is the frequency with the largest vibration energy value in a frequency range, wherein the error frequency range is the theoretical fundamental frequency of the target part ±3Hz.

[0013] In some embodiments, the above-described method for detecting equipment malfunctions further includes: determining whether the evaluation score of the target part is lower than a preset score threshold; and determining that the target part is malfunctioning when the evaluation score of the target equipment is lower than the preset score threshold. Simple Explanation of the Diagram

[0014] To make the above and other objects, features, advantages and embodiments of the present invention more apparent and understandable, the detailed description of the accompanying drawings is as follows: Figure 1 is a flowchart illustrating a method for detecting equipment malfunctions according to an embodiment of the present invention; Figure 2 illustrates the relationship between the critical frequency and the theoretical fundamental frequency of the spindle according to an embodiment of the present invention; Figure 3 illustrates the relationship between the critical frequency and the theoretical fundamental frequency of the balls in a bearing according to an embodiment of the present invention. Figure 4 is a schematic diagram illustrating the fitting steps according to an embodiment of the present invention; as well as Figure 5 illustrates an apparatus for detecting equipment malfunctions according to an embodiment of the present invention. Implementation

[0015] The following is a detailed description of the embodiments in conjunction with the accompanying drawings. However, the embodiments provided are not intended to limit the scope of the invention, and the description of the structural operation is not intended to limit the order of execution. Any structure resulting from the recombination of elements and producing a device with equivalent functionality is within the scope of the invention. Furthermore, the drawings are for illustrative purposes only and are not drawn to their original dimensions.

[0016] Please refer to Figure 1, which is a flowchart illustrating a method 100 for inspecting equipment malfunctions according to an embodiment of the present invention. The method 100 for inspecting equipment malfunctions includes a modeling stage and an online evaluation stage. The modeling stage includes steps 110-130 to establish a physical model (hereinafter referred to as the fitting model) representing the vibration signal. The online evaluation stage includes steps 140-160 to perform malfunction inspection on the target equipment and determine the health status (hereinafter referred to as the evaluation score) of the target component in the target equipment. Thus, the user can use the evaluation score to determine whether to replace / repair the target component in the target equipment.

[0017] In step 110, the operating speed and power of a plurality of devices are provided. In embodiments of the invention, the devices may include devices that operate normally without severely aged parts and / or devices that operate abnormally and have severely aged parts. In some embodiments, the devices include parts that vibrate during operation, such as gears, bearings, spindles, and rotors.

[0018] In step 120, the original score of a component is calculated based on the critical frequencies of the components of the device. In an embodiment of the invention, a normal component refers to a component without severe aging. The original score of a component is calculated using the following equation: in For raw fractions, This represents the energy values ​​corresponding to the top 50 frequencies in a given frequency spectrum. This represents the energy value corresponding to the key frequency. In some embodiments, a Fast Fourier Transform (FFT) can be used to obtain the corresponding spectrum of the vibration signal of the component, but the embodiments of the present invention are not limited thereto.

[0019] In this embodiment, the critical frequency is the actual fundamental frequency of the component's vibration signal. Specifically, this embodiment uses the theoretical fundamental frequency of the component's vibration signal as the center, supplemented by an error tolerance frequency band to find the actual fundamental frequency of the component. For example, if the component is a spindle, then ±3Hz of the theoretical fundamental frequency is taken as the frequency range for finding the critical frequency, where ±3Hz is the error tolerance frequency band, as shown in Figure 2. Figure 2 illustrates the relationship between the critical frequency and the theoretical fundamental frequency for the spindle, where the theoretical fundamental frequency is the operating speed / 60, for example, 80Hz. The critical frequency of the spindle is the frequency with the largest vibration energy value (i.e., the largest amplitude) within the frequency range of 80Hz±3Hz. For example, the frequency 79Hz corresponds to the largest amplitude value, therefore, the frequency 79Hz is the critical frequency of the spindle in this embodiment.

[0020] For example, if the component is a rotor, its allowable frequency range is ±3Hz from 0.5 to 5.5 times the theoretical fundamental frequency, such as 0.5 times ±3Hz, 1.5 times ±3Hz, 2.5 times ±3Hz, 3.5 times ±3Hz, 4.5 times ±3Hz, and 5.5 times ±3Hz. Specifically, the frequency values ​​with the maximum energy in the above six frequency bands are detected respectively to obtain six maximum energy frequency values, and the critical frequency is the average of the six maximum energy frequency values.

[0021] For example, if the part is a gear, its allowable frequency range is: meshing frequency ± (theoretical fundamental frequency / 2). For instance, if the gear has 48 teeth and the rotational frequency (theoretical fundamental frequency) is 80Hz, then the meshing frequency (GMF) is 3840Hz (48*80), and the allowable frequency range is 3840 ±40Hz. The critical frequency is the frequency value with the maximum energy within 3840 ±40Hz.

[0022] For example, if the component is a bearing, the critical frequency needs to consider three components: the outer ring, the inner ring, and the balls. Specifically, suppose a device rotates at 4800 rpm, and its internal bearing specifications are 9 balls, a contact angle of 0 degrees, a bearing pitch diameter of 53.5 mm, and a ball diameter of 11.112 mm. The calculated outer ring coefficient is 3.565. The outer ring frequency of this device (which can be considered the theoretical fundamental frequency of the bearing) = 3.565 * (4800 / 60) = 285.2 Hz. Therefore, when determining the error range, the nearest integer to 3.565 is taken as 3, with an error tolerance of... The allowable range is calculated as follows: bearing outer ring frequency ± |bearing outer ring coefficient - 3| * (4800 / 60) / 5 = 285.2 ± |3.565 - 3| * 16 = 285.2 ± 9.04, meaning the allowable error range is 276.16 Hz to 294.24 Hz. As shown in Figure 3, regarding the relationship between the critical frequency of the balls in the bearing and the outer ring frequency (theoretical fundamental frequency), the frequency value with the maximum energy in the range of 276.16 Hz to 294.24 Hz is located at 280 Hz.

[0023] Furthermore, in this embodiment, to ensure that the original score of the part can exclude the influence of machine error (i.e., differences in the equipment body), the calculation of the original score uses the energy values ​​corresponding to the top 50 frequencies of the vibration signal spectrum during part operation as a reference (e.g., as the denominator of the above equation (1)). For example, in one example, the energy value corresponding to the critical frequency of the part is 0.03, and it is in a healthy state. In another example, the energy value corresponding to the critical frequency of the part is 0.52, also in a healthy state. To integrate the above examples, the sum of the energy values ​​corresponding to the top 50 frequencies of the vibration signal spectrum is used as the denominator for reference calculation to obtain the ratio of the energy values ​​of the critical frequencies. Thus, the original scores calculated in the two examples above will be similar, for example, both equal to 0.022. This avoids the influence of machine error.

[0024] However, the embodiments of the present invention are not limited thereto. In other embodiments of the present invention, the frequency values ​​of the top 40 or 60 frequencies can also be used as a reference.

[0025] Please refer back to Figure 1. In step 130, a fitting model is calculated using curve fitting based on the operating speed, power, and original scores of several normal and abnormal devices. Please refer to Figure 4, which illustrates the fitting process of step 130 according to an embodiment of the present invention. As shown in Figure 4, devices 410 and 420 are pumps with different speeds and powers. After steps 110-120 above, the original scores of the normal internal parts can be obtained. For example, the original scores of the normal parts of device 410 are between 0.007 and 0.009, corresponding to 10 data points within a day. Similarly, the original scores of the normal parts of device 420 are between 0.013 and 0.016, also corresponding to 10 data points within a day. The fitting model of this embodiment uses the following equation: (2) Where n is the evaluation score of the part, P is the operating power (kW), W is the operating speed and frequency (Hz), and SR is the raw score. These are the coefficients of the equation.

[0026] In this embodiment, the evaluation score for each piece of data is given based on the original score. For example, a higher original score will result in a lower evaluation score, and a lower original score will result in a higher evaluation score. For instance, if the original score is 0.007, the evaluation score will be 95. Or, for example, if the original score is 0.008, the evaluation score will be 93. In some embodiments, considering that only normal part data from normal equipment may be available during the modeling phase, the given evaluation score will not be lower than a preset threshold. In this embodiment, the preset threshold is 80, but the embodiments of the present invention are not limited to this.

[0027] The coefficients of the equation are obtained in step 130. Then, a fitting model for assessing the health of the component can be obtained. This fitting model assesses the health of the component based on the energy value corresponding to the critical frequency of the component, while avoiding the influence of differences between different equipment on the health assessment.

[0028] Please refer back to Figure 1. After establishing the fitting model in the modeling stage, the online evaluation stage can then be performed, including steps to perform anomaly checks on the target equipment and determine the evaluation score of the target parts in the target equipment. In step 140, the operating speed and operating power of the target equipment to be evaluated are provided. Then, in step 150, the original score of the target part is calculated based on the key frequency of the target part of the target equipment. The method of calculating the original score in step 150 is similar to that in step 120, so it will not be repeated here. Next, step 160 is performed to calculate the evaluation score of the target part using the aforementioned fitting model based on the operating speed, operating power, and original score of the target part of the target equipment. Specifically, when the operating speed, operating power, and original score of the target part of the target equipment are as described above, the corresponding evaluation score can be calculated using the aforementioned formula (2) so that the user can understand the health status of the target parts of the target equipment.

[0029] In some embodiments, the method 100 for detecting equipment malfunctions may further include: determining whether the evaluation score of a target part is lower than the aforementioned preset score threshold (e.g., 80 points); if the evaluation score of the target part is lower than the aforementioned preset score threshold, the target part is determined to be malfunctioning and requires repair. Conversely, if the evaluation score of the target part is higher than or equal to the aforementioned preset score threshold, the target part is determined to be normal and does not require repair.

[0030] Please refer to Figure 5, which illustrates an apparatus 500 for detecting equipment malfunctions according to an embodiment of the present invention. The apparatus 500 for detecting equipment malfunctions in this embodiment is a computer device, such as a personal computer, which includes a memory 510 and a processor 520. The memory 510 stores a plurality of instructions, and the processor 520 is electrically connected to the memory 510 to load these instructions to perform the aforementioned method 100 for detecting equipment malfunctions. Thus, the apparatus 500 for detecting equipment malfunctions can assess the health of target components of a target device.

[0031] Although the present invention has been disclosed above with reference to several embodiments, it is not intended to limit the present invention. Anyone with ordinary knowledge in the technical field to which this invention pertains may make various modifications and refinements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of this invention shall be determined by the appended claims.

[0032] 100: Methods for detecting equipment malfunctions 110~160: Steps 200: Relationship between key frequency and theoretical fundamental frequency 300: The relationship between the critical frequency of the balls in a bearing and the outer ring frequency 410: Equipment 420: Equipment 500: Device for detecting equipment malfunctions 510: Memory 520: Processor

Claims

1. A method for detecting equipment malfunctions, applicable to an apparatus for detecting equipment malfunctions, comprising: performing a modeling phase, comprising: providing a plurality of operating speed frequencies and a plurality of operating powers for a plurality of devices; calculating a plurality of raw scores for a plurality of parts based on a plurality of key frequencies of the plurality of parts of the devices; and calculating a fitting model using curve fitting based on the operating speed frequencies, the operating powers, and the raw scores of the parts of the devices; and performing an online evaluation phase, comprising: providing an operating speed frequency and an operating power for a target device; calculating a raw score for a target part based on a key frequency of a target part of the target device, wherein the raw score of each of the parts and the raw score of the target part are calculated using the following equation: where is the raw score, is the energy value corresponding to the top 50 frequencies in a corresponding spectrum, and is the energy value corresponding to the key frequency; and calculating an evaluation score for the target part using the fitting model based on the operating speed frequency, the operating power, and the raw score of the target part of the target device.

2. The method for inspecting equipment malfunctions as described in claim 1, wherein the fitting model is the following equation: where n is the evaluation score, P is the operating power, W is the operating speed frequency, SR is the original score, and is the coefficient of the equation.

3. The method for inspecting equipment malfunctions as described in claim 1, wherein when the target part is the main shaft, the critical frequency is the frequency with the largest vibration energy value in a frequency range, wherein the frequency range is the theoretical fundamental frequency of the target part ±3Hz.

4. The method for detecting equipment abnormalities as described in claim 1 further includes: determining whether the evaluation score of the target part is lower than a preset score threshold; and determining that the target part is abnormal when the evaluation score of the target equipment is lower than the preset score threshold.

5. An apparatus for detecting equipment malfunctions, comprising: a memory for storing a plurality of instructions; and a processor electrically connected to the memory to load the instructions and perform: a modeling phase comprising: providing a plurality of operating speed frequencies and a plurality of operating powers for a plurality of devices; calculating a plurality of raw scores based on a plurality of key frequencies of a plurality of components of the devices; and calculating a fitting model using curve fitting based on the operating speed frequencies, the operating powers, and the raw scores of the components of the devices; and an online evaluation phase comprising: providing an operating speed frequency and an operating power of a target device; calculating a raw score of a target component based on a key frequency of a target component of the target device; and calculating an evaluation score of the target component using the fitting model based on the operating speed frequency, the operating power, and the raw score of the target component of the target device, wherein the raw score of each of the components and the raw score of the target component are calculated using the following equation: Where is the original fraction, is the energy value of the top 50 frequencies in a corresponding spectrum, and is the energy value of the key frequency.

6. The apparatus for detecting equipment malfunction as described in claim 5, wherein the fitting model is the following equation: where n is the evaluation score, P is the operating power, W is the operating speed frequency, SR is the original score, and is the coefficient of the equation.

7. The apparatus for inspecting equipment malfunctions as described in claim 5, wherein when the target part is the main axis, the critical frequency is the frequency with the largest vibration energy value in a frequency range, wherein the frequency range is the theoretical fundamental frequency of the target part ±3Hz.

8. The apparatus for detecting equipment malfunction as described in claim 5, wherein the processor further performs the following: determining whether the evaluation score of the target part is lower than a preset score threshold; and determining that the target part is malfunctioning when the evaluation score of the target equipment is lower than the preset score threshold.

Images