Ball screw pair fault diagnosis system and method
By constructing a three-dimensional coordinate system and a dual-path verification mechanism, combined with the vibration signals and historical data of the ball screw, the problem of long and poor interpretability of ball screw fault diagnosis in the existing technology is solved, early fault identification and potential hidden danger warning are achieved, and diagnostic accuracy and equipment maintenance efficiency are improved.
Patent Information
- Application Number
- CN202510663029.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-22
- Publication Date
- 2025-08-01
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing ball screw pair fault diagnosis system cannot combine multi-dimensional parameters and historical data to locate the fault cause, resulting in long-term diagnosis and poor interpretability, making it difficult to identify early performance degradation and potential potential risks.
By reading the vibration signal on the surface of the ball screw nut, calculating the fault frequency and kurtitude ratio, building a three-dimensional coordinate system for fault types, combining historical cases and risk levels for diagnosis, and using a dual-path verification mechanism to improve diagnostic accuracy and adaptability.
It realizes sensitive identification and accurate diagnosis of early faults, improves the robustness and adaptability of diagnosis, can promptly warn of potential hidden dangers, and improves the maintenance efficiency and reliability of equipment.
Smart Images
Figure CN120404108A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of fault diagnosis, and particularly relates to a fault diagnosis system and method for a ball screw pair. Background Art
[0002] Due to advantages such as strong load-bearing capacity and high positioning accuracy, ball screws have become the most commonly used transmission components in automotive parts, precision machinery, industrial robots, and medical equipment. The operating state of ball screws has an important impact on characteristics such as the positioning accuracy, transmission efficiency, load capacity, and operating smoothness of machinery.
[0003] However, the fault diagnosis systems applied to ball screw pairs in the prior art still have the following deficiencies in the actual application process: After determining the abnormal state of the ball screw, it is impossible to locate the fault cause of the ball screw by combining multi-dimensional parameters and historical data of the ball screw, resulting in a long diagnostic process and poor interpretability; In addition, usually only intervene when obvious abnormalities occur in the equipment, such as when the threshold is exceeded and an alarm is issued. It is difficult to identify early performance degradation through trend analysis of continuous data during the normal operation stage, and it is impossible to achieve early warning of potential hidden dangers.
[0004] Therefore, a fault diagnosis system and method for a ball screw pair are introduced. Summary of the Invention
[0005] In view of this, the present invention provides a fault diagnosis system and method for a ball screw pair to solve the problems raised in the above background art.
[0006] The object of the present invention can be achieved by the following technical solutions: A fault diagnosis system for a ball screw pair, comprising: A state judgment module: reads the vibration signal on the surface of the nut corresponding to the ball screw, and analyzes to obtain the fault frequency f of the surface defect of the screw; after sending the fault frequency f of the surface defect of the screw to a technician to determine the filter, analyzes and judges the filtered vibration signal and the original vibration signal to obtain the kurtosis ratio ks; A state output module: receives the kurtosis ratio ks obtained by analyzing the ball screw, and uses the judgment logic to judge the state result of the current ball screw; wherein, is the judgment result of whether the current ball screw is abnormal, 1 indicates that the current ball screw is in an abnormal state, 0 indicates that the current ball screw is in a normal state, ks is the kurtosis ratio of the current sample, is the preset kurtosis threshold; Diagnostic positioning module: When the state result of the ball screw is in an abnormal state, use the pre-edited fault positioning calculation logic to locate the estimated cause corresponding to the abnormal state of the ball screw, and send it to the technician in combination with the risk level; the risk level includes low risk level, medium risk level, and high risk level.
[0007] In some embodiments, the specific calculation process of the fault frequency f is as follows: Using the formula Calculate the fault frequency f of the surface defect of the lead screw, where is the nominal diameter of the lead screw, is the derivative of the rotational angular displacement of the lead screw, is the ball diameter, r is the projection of the line connecting the ball center and the lead screw axis in the radial plane of the lead screw, is the contact angle between the ball and the lead screw.
[0008] In some embodiments, the specific calculation process of the kurtosis ratio ks is as follows: Using the formula Calculate the kurtosis ratio of the ball screw; where n is the signal length, is the i-th sampling point of the original signal, is the average value of the original signal, is the i-th sampling point of the filtered signal, is the average value of the filtered signal.
[0009] In some embodiments, the use of the pre-edited fault positioning calculation logic to locate the estimated cause corresponding to the abnormal state of the ball screw is specifically as follows: Extract the fault frequency f of the ball screw, calculate the ratio of the amplitude of the second harmonic to the amplitude of the fundamental frequency of the fault frequency f to obtain the frequency domain harmonic ratio, and at the same time obtain the fractal cone number and temperature data; Mark the frequency domain harmonic ratio, fractal cone number, and temperature data corresponding to the ball screw as W, P, and U respectively, set the frequency domain harmonic ratio threshold, fractal cone number threshold, and temperature threshold corresponding to W, P, and U, denoted as S, D, and R; calculate the ratios between W, P, U and the corresponding S, D, R respectively, denoted as harmonic threshold ratio, cone number threshold ratio, and temperature threshold ratio; that is, calculate the ratios with W, P, U as the numerators and S, D, R as the denominators respectively; Pre-construct a three-dimensional coordinate system, divide different fault type regions within the three-dimensional coordinate system as the division regions corresponding to each fault type, after normalizing the harmonic threshold ratio, cone number threshold ratio, and temperature threshold ratio of the ball screw, use the harmonic threshold ratio, cone number threshold ratio, and temperature threshold ratio as the values on the X-axis, Y-axis, and Z-axis corresponding to the three-dimensional coordinate points respectively, and construct three-dimensional coordinate points; Draw the position points corresponding to the three-dimensional coordinate points in the three-dimensional coordinate system, identify the divided area where the position points are located, and use the fault type in the divided area as the preliminary fault cause of the corresponding abnormal state of the ball screw.
[0010] In some embodiments, the positioning of the estimated cause of the abnormal state corresponding to the ball screw by using the pre-edited fault location calculation logic further includes: Extract each group of historical processing cases of the ball screw in the abnormal state from the storage database, where each group of historical processing cases includes the fault type, occurrence time point, processing technician, harmonic threshold ratio, cone number threshold ratio, and temperature threshold ratio of the ball screw; Mark the harmonic threshold ratio, cone number threshold ratio, and temperature threshold ratio corresponding to each group of historical processing cases as Mk, Nk, and Jk respectively; where k represents the case number of each group of historical processing cases; Calculate the similarity value Peri between Mk, Nk, Jk corresponding to each group of historical processing cases and the current ball screw W, P, U, that is, after normalizing Mk, Nk, Jk and W, P, U, substitute them into the formula Calculate and obtain; where Are preset weight coefficients respectively; After obtaining the similarity value Peri corresponding to each group of historical processing cases, select the historical processing case with a lower similarity value and extract its fault type as the confidence fault cause of the ball screw in the abnormal state; Correlate and match the preliminary fault cause and the confidence fault cause of the abnormal state corresponding to the ball screw. If the match is successful, directly use it as the estimated cause of the abnormal state corresponding to the ball screw; If the match fails, identify the regional position of the confidence fault cause in the three-dimensional coordinate system and perform position analysis with the regional position of the preliminary fault cause. If the two regional positions are adjacent, integrate the preliminary fault cause and the confidence fault cause as the estimated cause of the abnormal state corresponding to the ball screw. If the two regional positions are not adjacent, directly use the confidence fault cause as the estimated cause of the abnormal state corresponding to the ball screw.
[0011] In some embodiments, the specific confirmation steps of the risk level are as follows: Calculate the difference between the kurtosis ratio and the kurtosis threshold of the ball screw in the abnormal state, mark the calculated difference as the risk degree index, and input the marked risk degree index into the pre-constructed index set for matching. There are three groups of risk degree index intervals stored in the index set, denoted as A, B, and C, where A, B, and C correspond to low risk level, medium risk level, and high risk level respectively. The risk level corresponding to the matching result display interval is used as the risk level of the current ball screw state being abnormal.
[0012] In some embodiments, it further includes a health trend module: When the state result of the ball screw is in the normal state, the kurtosis ratio in the normal state is stored. When the ball screw has g consecutive state results in the normal state, the corresponding g groups of kurtosis ratios are extracted and trend analysis is performed, and the hidden danger level is output in combination with the running duration of the ball screw and sent to the technical personnel; wherein the hidden danger level includes a low hidden danger level, a medium hidden danger level, and a high hidden danger level.
[0013] In some embodiments, the extracting the corresponding g groups of kurtosis ratios and performing trend analysis specifically includes: Calculate the difference between each of the g groups of kurtosis ratios and the corresponding kurtosis threshold, and take the absolute value of the calculation result to obtain g groups of kurtosis differences; Arrange the g groups of kurtosis differences in chronological order, and calculate the difference between adjacent two groups of kurtosis differences for the arranged g groups of kurtosis differences. The difference calculation order is to subtract the kurtosis difference on the left from the kurtosis difference on the right in adjacent two groups of kurtosis differences; if the difference calculation result of a certain adjacent two groups of kurtosis differences is negative, take the absolute value and record it as a near abnormal value, and if the difference calculation result of a certain adjacent two groups of kurtosis differences is positive, record it as a deviation abnormal value; Sum the near abnormal values and deviation abnormal values of each group respectively to obtain a near total value and a deviation total value, and perform a ratio calculation with the near total value as the numerator and the deviation total value as the denominator to obtain the kurtosis trend index of the ball screw, denoted as h1.
[0014] In some embodiments, the outputting the hidden danger level in combination with the running duration of the ball screw specifically includes: Obtain the running duration and the set life of the ball screw, calculate the ratio of the running duration of the ball screw in the set life as the life additional index h2 of the ball screw; Perform normalization processing on the kurtosis trend index h1 and the life additional index h2 of the ball screw and then substitute them into the formula Perform weighted calculation to obtain the hidden danger assessment index Phd of the ball screw; wherein are the weight coefficients corresponding to the kurtosis trend index h1 and the life additional index h2 respectively; For the hidden danger assessment index Phd calculated for the ball screw, preset three groups of hidden danger assessment index intervals, and the three groups of hidden danger assessment index intervals respectively correspond to a low hidden danger level, a medium hidden danger level, and a high hidden danger level. Match the hidden danger assessment index Phd with the three groups of hidden danger assessment index intervals, and determine the hidden danger level of the ball screw based on the matching result.
[0015] A method for diagnosing faults of a ball screw pair includes: Fault frequency analysis: Read the vibration signal on the surface of the nut corresponding to the ball screw, and analyze to obtain the fault frequency f of the surface defect of the screw; Kurtosis ratio analysis: After sending the fault frequency f of the surface defect of the screw to the technician to determine the filter; analyze and judge the filtered vibration signal and the original vibration signal to obtain the kurtosis ratio ks; Status determination: Receive the kurtosis ratio ks analyzed for the corresponding ball screw, and judge the current status result of the ball screw; the status result includes abnormal status and normal status; Abnormal location: When the status result of the ball screw is in the abnormal state, use the pre-edited fault location calculation logic to locate the estimated cause of the abnormal state corresponding to the ball screw, and send it to the technician in combination with the risk level; the risk level includes low risk level, medium risk level, and high risk level; Normal prediction: When the status result of the ball screw is in the normal state, store the kurtosis ratio in the normal state. When the ball screw has g consecutive normal status results, extract the corresponding g groups of kurtosis ratios and perform trend analysis, and output the hidden danger level in combination with the running duration of the ball screw and send it to the technician; the hidden danger level includes low hidden danger level, medium hidden danger level, and high hidden danger level.
[0016] Compared with the prior art, the beneficial effects of the present invention are: The present invention extracts three types of parameters: frequency domain harmonic ratio, fractal cone number, and temperature, calculates the ratio with the corresponding threshold, constructs a three-dimensional coordinate system, normalizes the three types of threshold ratios and maps them to three-dimensional coordinate points. Through the preset fault type area division, directly locate the preliminary fault cause. At the same time, extract historical cases from the database, calculate the similarity value between the current parameters and the historical cases, select the case with a low similarity value as the confidence fault cause, associate and match the preliminary fault cause with the confidence fault cause. If the match is successful, cross-verify to enhance the confidence level; if the match fails, through the adjacent analysis of the three-dimensional space area, integrate or preferentially select the confidence cause, realizing a dual-path verification mechanism driven by the model and data, reducing misjudgment caused by the limitations of a single method, and improving the diagnostic robustness; The present invention stores the kurtosis ratio in the normal state, analyzes the kurtosis difference trend when there are g consecutive groups of normal data, calculates the kurtosis trend index and the life addition index, and fuses them to obtain the hidden danger assessment index. Match the low, medium, and high hidden danger levels through the hidden danger assessment index to realize the timely warning of the potential risks of the ball screw; The present invention calculates the difference between the kurtosis ratio and the threshold in the abnormal state, matches the preset interval to determine the low, medium, and high risk levels, and the kurtosis threshold is automatically adjusted according to the equipment life cycle to improve the diagnostic adaptability. Brief description of the drawings
[0017] In the following description of exemplary embodiments in conjunction with the accompanying drawings, more details, features, and advantages of the present application are disclosed. In the drawings: Figure 1 is a principle block diagram of the present invention; Figure 2 is a flowchart of the present invention. Detailed implementation manners
[0018] Several embodiments of the present application will be described in more detail below with reference to the accompanying drawings so that those skilled in the art can implement the present application. The present application can be embodied in many different forms and for many different purposes and should not be limited to the embodiments set forth herein. These embodiments are provided so that this application will be thorough and complete, and will fully convey the scope of the application to those skilled in the art. The embodiments do not limit the present application.
[0019] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It will be further understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning that is consistent with their meaning in the relevant art and / or the context of this specification, and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
[0020] Embodiment 1 Please refer to Figure 1 as shown. A ball screw pair fault diagnosis system includes a state judgment module, a state output module, a diagnosis and positioning module, and a health trend module; The state judgment module is used to read the vibration signal on the surface of the nut corresponding to the ball screw, and analyze to obtain the fault frequency f of the surface defect of the screw; send the fault frequency f of the surface defect of the screw to the technician to determine the filter; determine the filter based on the maximum correlation kurtosis algorithm and the fault frequency f; analyze and judge the filtered vibration signal and the original vibration signal to obtain the kurtosis ratio ks; Supplementary description: The system includes traditional components and sensors. The transmission component is mainly composed of a screw and a nut. Among them, the power is transmitted between the screw and the nut with steel balls as rolling elements. The sensor is used to sense the vibration signal on the surface of the nut; Determine the filter based on the maximum correlation kurtosis algorithm and the fault frequency f. The fault frequency f provides the "target coordinates" for the filter, enabling it to accurately separate the fault signal; the maximum correlation kurtosis algorithm endows the filter with "self-adaptability", maximizing the detectability of fault features through parameter optimization. The combination of the two ensures the sensitivity and accuracy of the diagnostic system to early faults; The specific calculation process of the fault frequency f is: using the formula The fault frequency f of the surface defects of the lead screw is calculated, where is the nominal diameter of the lead screw, is the derivative of the rotational angular displacement of the lead screw, is the diameter of the ball, r is the projection of the line connecting the center of the ball and the axis of the lead screw on the radial plane of the lead screw, is the contact angle between the ball and the lead screw; The specific calculation process of the kurtosis ratio ks is as follows: where n is the signal length, is the i-th sampling point of the original signal, is the average value of the original signal, is the i-th sampling point of the filtered signal, is the average value of the filtered signal; The status output module is used to receive the kurtosis ratio ks obtained from the analysis of the corresponding ball screw, and use the judgment logic to judge the status result of the current ball screw; where is the judgment result of whether the current ball screw is abnormal. 1 indicates that the current ball screw is in an abnormal state, 0 indicates that the current ball screw is in a normal state, ks is the kurtosis ratio of the current sample, is the preset kurtosis threshold; The kurtosis threshold is automatically updated according to the different states of the equipment during the running-in period, normal operation period, and aging period. For example, the threshold is relaxed during the running-in period and tightened during the aging period to improve the diagnostic adaptability;
[0021] The diagnostic and positioning module is used to, when the status result of the ball screw is in an abnormal state, use the pre-edited fault location calculation logic to locate the estimated cause of the abnormal state corresponding to the ball screw, and send it to the technical personnel in combination with the risk level; the risk level includes low risk level, medium risk level, and high risk level; Specifically: Calculate the difference between the kurtosis ratio and the kurtosis threshold of the ball screw in the abnormal state. Mark the calculated difference as the risk degree index, and input the marked risk degree index into the pre-constructed index set for matching. There are three groups of risk degree index intervals stored in the index set, represented by A, B, and C. Among them, A, B, and C correspond to the low risk level, medium risk level, and high risk level respectively. The risk level corresponding to the matching result display interval is used as the risk level of the current ball screw status result being in an abnormal state; Supplementary explanation: Determining the risk level corresponding to the abnormal state can evaluate the urgency of fault handling and provide clear priority maintenance guidelines for technical personnel; Extract the fault frequency f of the ball screw, calculate the ratio of the amplitude of the second harmonic of the fault frequency f to the amplitude of the fundamental frequency to obtain the frequency-domain harmonic ratio, and at the same time obtain the fractal cone number and temperature data; Supplementary note: The temperature data is collected through pre-deployed temperature sensors. Among them, the frequency-domain harmonic ratio reflects the degree of energy distribution concentration, the fractal cone number quantifies the irregularity of the vibration signal, reflects the surface damage complexity, and the temperature data reflects whether there is abnormal lubrication or contact stress; Mark the frequency-domain harmonic ratio, fractal cone number, and temperature data corresponding to the ball screw as W, P, and U respectively. Set the frequency-domain harmonic ratio threshold, fractal cone number threshold, and temperature threshold corresponding to W, P, and U, denoted as S, D, and R; calculate the ratios between W, P, and U and the corresponding S, D, and R respectively; that is, calculate the ratios with W, P, and U as the numerators and S, D, and R as the denominators respectively, denoted as the harmonic threshold ratio, cone number threshold ratio, and temperature threshold ratio; Pre-construct a three-dimensional coordinate system, divide different fault type regions within the three-dimensional coordinate system as the divided regions corresponding to each fault type. After normalizing the harmonic threshold ratio, cone number threshold ratio, and temperature threshold ratio of the ball screw, use the harmonic threshold ratio, cone number threshold ratio, and temperature threshold ratio as the values corresponding to the X-axis, Y-axis, and Z-axis of the three-dimensional coordinate points respectively to construct three-dimensional coordinate points; Plot the position points corresponding to the three-dimensional coordinate points within the three-dimensional coordinate system, identify the divided region where the position points are located, and use the fault type of the divided region as the preliminary fault cause of the corresponding abnormal state of the ball screw; Supplementary note: Map the three types of threshold ratios to the X-axis, Y-axis, and Z-axis of the three-dimensional coordinate system respectively. Each coordinate point corresponds to the real-time state of the ball screw. By pre-dividing the spatial regions of different fault types (such as "wear region", "overload region", "poor lubrication region", etc.), the complex multi-parameter combination can be transformed into a position judgment in geometric space, intuitively reflecting the corresponding relationship between the fault type and the multi-parameter combination; Extract each group of historical treatment cases corresponding to the abnormal state of the ball screw from the storage database. Each group of historical treatment cases includes the fault type of the ball screw, the occurrence time point, the treatment technician, the harmonic threshold ratio, the cone number threshold ratio, and the temperature threshold ratio; Mark the harmonic threshold ratio, cone number threshold ratio, and temperature threshold ratio corresponding to each group of historical treatment cases as Mk, Nk, and Jk respectively; where k represents the case number of each group of historical treatment cases; Calculate the similarity value Peri between Mk, Nk, and Jk corresponding to each group of historical treatment cases and the current ball screw W, P, and U, that is, after normalizing Mk, Nk, Jk and W, P, U, substitute them into the formula Calculate and obtain; where are preset weight coefficients respectively; After obtaining the similarity value Peri corresponding to each group of historical processing cases, the historical processing case with the lower similarity value is selected and its fault type is extracted as the confident fault cause of the ball screw corresponding to the abnormal state; In addition, by calculating the multi-dimensional parameter similarity between the current abnormal state and historical cases, we can directly match historical fault types with similar characteristics. This "learning from history" reasoning model avoids repeated analysis and is particularly suitable for fault scenarios under complex working conditions that are difficult to directly determine using theoretical models. Correlate and match the preliminary fault cause and the confident fault cause corresponding to the abnormal state of the ball screw. If the match is successful, it will be directly used as the estimated cause of the abnormal state of the ball screw; If the matching fails, the regional position of the confident fault cause in the three-dimensional coordinate system is identified and compared with the regional position of the preliminary fault cause for position analysis. If the two sets of regional positions are adjacent, the preliminary fault cause and the confident fault cause are integrated as the estimated cause of the abnormal state of the ball screw. If the two sets of regional positions are not adjacent, the confident fault cause is directly used as the estimated cause of the abnormal state of the ball screw. As a side note, traditional fault diagnosis typically relies on a single technology, while this solution forms a "double insurance" mechanism through bidirectional verification of preliminary causes and confidence reasons. If the two successfully match, the confidence level is greatly improved through cross-validation of different technical paths, and can be directly used as a high-reliability diagnosis result. If the match fails, instead of directly denying any result, spatial location analysis (regional proximity) is used to explore potential associations to prevent missed diagnosis due to the limitations of a single method (such as the model not covering edge scenarios and case library data bias). The health trend module is used to store the kurtosis ratio under normal conditions when the ball screw status result is normal. When the ball screw status results are normal for g consecutive times, the module extracts the corresponding g groups of kurtosis ratios and performs trend analysis. The module also outputs the potential hazard level based on the running time of the ball screw and sends it to the technician. The potential hazard level includes low potential hazard level, medium potential hazard level and high potential hazard level. The value of g is preset by the technician and g>5. Specifically: Calculate the difference between the kurtosis ratio of group g and the corresponding kurtosis threshold, and take the absolute value of the calculation result to obtain the kurtosis difference of group g; Arrange the g groups of kurtosis differences in chronological order, and calculate the difference between the two adjacent groups of kurtosis differences for the arranged g groups of kurtosis differences. The difference calculation order is to subtract the kurtosis difference on the left from the kurtosis difference on the right of the two adjacent groups of kurtosis differences. If the difference between the two adjacent groups of kurtosis differences is a negative value, then take the absolute value and record it as a proximal outlier. If the difference between the two adjacent groups of kurtosis differences is a positive value, then record it as a deviating outlier. Sum each group of adjacent outliers and deviating outliers to obtain the adjacent total value and the deviating total value. Calculate the ratio with the adjacent total value as the numerator and the deviating total value as the denominator to obtain the kurtosis trend index of the ball screw, which is recorded as h1.
[0022] Supplementary note: if the total deviation value is 0, it will automatically be rounded to the integer one; Through the time series analysis of g consecutive groups of kurtosis differences, the changing trend of kurtosis value relative to the threshold is captured; the difference calculation of adjacent kurtosis differences (deviating outliers, adjacent outliers) is introduced; The deviation from the outlier value reflects that the kurtosis difference deviates from the abnormal state, and the proximity to the outlier value reflects that the kurtosis difference approaches the abnormal state; Obtain the operating time and set life of the ball screw, and calculate the proportion of the ball screw operating time to the set life as the life additional index h2 of the ball screw; The kurtosis trend index h1 and life additional index h2 of the ball screw are normalized and then inserted into the formula Perform weighted calculation to obtain the hidden danger assessment index Phd of the ball screw; are the weight coefficients corresponding to the kurtosis trend index h1 and the life additional index h2 respectively; For the hidden danger assessment index Phd calculated for the ball screw, three groups of hidden danger assessment index intervals are preset, and the three groups of hidden danger assessment index intervals correspond to low hidden danger levels, medium hidden danger levels, and high hidden danger levels, respectively. The hidden danger assessment index Phd is matched with the three groups of hidden danger assessment index intervals, and the hidden danger level of the ball screw is determined based on the matching results; To supplement, hidden danger assessment refers to the upgrade from "fault detection" to "hidden danger prediction" achieved by PhD through multi-dimensional data fusion (trend + life), dynamic trend capture (h1) and a graded early warning mechanism. This significantly improves the maintenance efficiency and reliability of key equipment such as ball screws, and is particularly suitable for industrial scenarios with high requirements for continuous operation stability.
[0023] Example 2 See also Figure 2As shown, a ball screw pair fault diagnosis system provided according to Embodiment 1 of the present application, and Embodiment 2 of the present application proposes a ball screw pair fault diagnosis method. Embodiment 2 is merely a preferred manner of Embodiment 1, and the implementation of Embodiment 2 will not affect the independent implementation of Embodiment 1.
[0024] Specifically, the difference of a ball screw pair fault diagnosis method provided by Embodiment 2 of the present application is that it includes: Fault frequency analysis: Read the vibration signal on the surface of the nut corresponding to the ball screw, and analyze to obtain the fault frequency f of the surface defect of the screw. Kurtosis ratio analysis: After sending the fault frequency f of the surface defect of the screw to the technician to determine the filter, analyze and judge the filtered vibration signal and the original vibration signal to obtain the kurtosis ratio ks. Status determination: Receive the kurtosis ratio ks obtained by analyzing the corresponding ball screw, and judge the current status result of the ball screw; the status result includes an abnormal status and a normal status. Abnormal location: When the status result of the ball screw is an abnormal status, use the pre-edited fault location calculation logic to locate the estimated cause of the abnormal status corresponding to the ball screw, and send it to the technician in combination with the risk level; where the risk level includes a low risk level, a medium risk level, and a high risk level. Normal prediction: When the status result of the ball screw is a normal status, store the kurtosis ratio in the normal status. When the ball screw has g consecutive normal status results, extract the corresponding g groups of kurtosis ratios and perform a trend analysis, and output the hidden danger level in combination with the running time of the ball screw and send it to the technician; where the hidden danger level includes a low hidden danger level, a medium hidden danger level, and a high hidden danger level.
[0025] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the present invention to only the specific implementation manners. Obviously, many modifications and changes can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principle and practical application of the present invention, so that those skilled in the art in the relevant technical field can understand and utilize the present invention well. The present invention is only limited by the claims and their full scope and equivalents.
Claims
1. A fault diagnosis system for a ball screw pair, characterized in that, Including: Status judgment module: Read the vibration signal on the surface of the nut corresponding to the ball screw, and analyze to obtain the fault frequency f of the surface defect of the screw; After sending the fault frequency f of the surface defect of the screw to the technician to determine the filter, analyze and judge the filtered vibration signal and the original vibration signal to obtain the kurtosis ratio ks; Status output module: Receive the kurtosis ratio ks obtained from the analysis corresponding to the ball screw, and use the judgment logic to judge the status result of the current ball screw; among them, is the judgment result of whether the current ball screw is abnormal. 1 indicates that the current ball screw is in an abnormal state, 0 indicates that the current ball screw is in a normal state, ks is the kurtosis ratio of the current sample, is the preset kurtosis threshold; Diagnosis and location module: When the status result of the ball screw is an abnormal state, use the pre-edited fault location calculation logic to locate the estimated cause corresponding to the abnormal state of the ball screw, and send it to the technician in combination with the risk level; The risk level includes low risk level, medium risk level and high risk level.
2. The fault diagnosis system for a ball screw pair according to claim 1, characterized in that, The specific calculation process of the fault frequency f is as follows: Using the formula calculate the fault frequency f of the surface defects of the lead screw, where is the nominal diameter of the lead screw, is the derivative of the rotational angular displacement of the lead screw, is the ball diameter, r is the projection of the line connecting the ball center and the lead screw axis on the radial plane of the lead screw, is the contact angle between the ball and the lead screw.
3. The fault diagnosis system of a ball screw pair according to claim 2, characterized in that, The specific calculation process of the kurtosis ratio ks is as follows: Using the formula the kurtosis ratio of the ball screw is calculated; where n is the signal length, is the i-th sampling point of the original signal, is the average value of the original signal, is the i-th sampling point of the filtered signal, is the average value of the filtered signal.
4. The fault diagnosis system for a ball screw pair according to claim 3, characterized in that, The use of the pre-edited fault location calculation logic to locate the estimated cause corresponding to the abnormal state of the ball screw is specifically: Extract the fault frequency f of the ball screw, and calculate the ratio of the second harmonic amplitude to the fundamental frequency amplitude of the fault frequency f to obtain the frequency domain harmonic ratio. At the same time, obtain the fractal cone number and temperature data; Mark the frequency domain harmonic ratio, fractal cone number and temperature data corresponding to the ball screw as W, P, U respectively, and set the frequency domain harmonic ratio threshold, fractal cone number threshold and temperature threshold corresponding to W, P, U, denoted as S, D, R; Calculate the ratios between W, P, U and the corresponding S, D, R respectively, denoted as harmonic threshold ratio, cone number threshold ratio and temperature threshold ratio; That is, calculate the ratios with W, P, U as the numerators and S, D, R as the denominators respectively; Pre-construct a three-dimensional coordinate system, divide different fault type areas in the three-dimensional coordinate system as the divided areas corresponding to each fault type. After normalizing the harmonic threshold ratio, cone number threshold ratio and temperature threshold ratio of the ball screw, use the harmonic threshold ratio, cone number threshold ratio and temperature threshold ratio as the values on the X-axis, Y-axis and Z-axis corresponding to the three-dimensional coordinate points respectively to construct three-dimensional coordinate points; Plot the position points corresponding to the three-dimensional coordinate points in the three-dimensional coordinate system, identify the divided area where the position points are located, and use the fault type of the divided area as the preliminary fault cause corresponding to the abnormal state of the ball screw.
5. The fault diagnosis system for a ball screw pair according to claim 4, characterized in that, The use of the pre-edited fault location calculation logic to locate the estimated cause corresponding to the abnormal state of the ball screw also includes: Extract each group of historical processing cases corresponding to the abnormal state of the ball screw from the storage database, where each group of historical processing cases includes the fault type, occurrence time point, processing technician, harmonic threshold ratio, cone number threshold ratio and temperature threshold ratio of the ball screw; Mark the harmonic threshold ratio, cone number threshold ratio and temperature threshold ratio corresponding to each group of historical processing cases as Mk, Nk, Jk respectively; where k represents the case number of each group of historical processing cases; Calculate the similarity value Peri between Mk, Nk, Jk corresponding to each group of historical processing cases and the current ball screw W, P, U, that is, after normalizing Mk, Nk, Jk and W, P, U, substitute them into the formula Calculated; where are respectively preset weight coefficients; After obtaining the similarity value Peri corresponding to each group of historical processing cases, select the historical processing cases with lower similarity values and extract their fault types as the confidence fault causes for the corresponding abnormal states of the ball screw; Correlate and match the preliminary fault causes and confidence fault causes for the corresponding abnormal states of the ball screw. If the match is successful, directly use it as the estimated cause for the corresponding abnormal state of the ball screw; If the match fails, identify the regional position of the confidence fault cause in the three-dimensional coordinate system and perform a position analysis with the regional position of the preliminary fault cause. If the two regional positions are adjacent, integrate the preliminary fault cause and the confidence fault cause as the estimated cause for the corresponding abnormal state of the ball screw. If the two regional positions are not adjacent, directly use the confidence fault cause as the estimated cause for the corresponding abnormal state of the ball screw.
6. The fault diagnosis system of a ball screw pair according to claim 5, characterized in that The specific steps for confirming the risk level are as follows: Calculate the difference between the kurtosis ratio and the kurtosis threshold for the ball screw in the abnormal state. Mark the calculated difference as the risk degree index. Input the marked risk degree index into a pre-constructed index set for matching. The index set stores three groups of risk degree index intervals, denoted as A, B, and C, where A, B, and C correspond to low risk level, medium risk level, and high risk level respectively. Take the risk level corresponding to the displayed interval of the matching result as the risk level of the current abnormal state result of the ball screw.
7. A ball screw pair fault diagnosis system according to claim 6, characterized in that, It also includes a health trend module: When the ball screw state result is in the normal state, store the kurtosis ratio in the normal state. When the ball screw has g consecutive state results in the normal state, extract the corresponding g groups of kurtosis ratios and perform a trend analysis, and output the hidden danger level in combination with the running duration of the ball screw and send it to the technical staff; the hidden danger level includes low hidden danger level, medium hidden danger level, and high hidden danger level.
8. A ball screw pair fault diagnosis system according to claim 7, characterized in that, The specific operation of extracting the corresponding g groups of kurtosis ratios and performing a trend analysis is as follows: Calculate the difference between each of the g groups of kurtosis ratios and the corresponding kurtosis threshold, and take the absolute value of the calculation result to obtain g groups of kurtosis differences; Arrange the g groups of kurtosis differences in chronological order. Calculate the difference between adjacent two groups of kurtosis differences for the arranged g groups of kurtosis differences. The difference calculation order is to subtract the kurtosis difference on the left from the kurtosis difference on the right in adjacent two groups of kurtosis differences; if the difference calculation result of a certain adjacent two groups of kurtosis differences is negative, take the absolute value and record it as a near abnormal value. If the difference calculation result of a certain adjacent two groups of kurtosis differences is positive, record it as a deviation abnormal value; Sum up each group of near abnormal values and deviation abnormal values respectively to obtain the near total value and the deviation total value. Calculate the ratio with the near total value as the numerator and the deviation total value as the denominator to obtain the kurtosis trend index of the ball screw, denoted as h1.
9. The fault diagnosis system for a ball screw pair according to claim 8, wherein, The specific operation of outputting the hidden danger level in combination with the running duration of the ball screw is as follows: Obtain the running duration and the set life of the ball screw, and calculate the proportion of the running duration of the ball screw in the set life as the life additional index h2 of the ball screw; Normalize the kurtosis trend index h1 and the life additional index h2 of the ball screw and then substitute them into the formula Perform weighted calculation to obtain the hidden danger assessment index Phd of the ball screw; where are the weight coefficients corresponding to the kurtosis trend index h1 and the life additional index h2 respectively; For the hidden danger assessment index Phd calculated for the ball screw, three groups of hidden danger assessment index ranges are preset, and the low hidden danger level, medium hidden danger level, and high hidden danger level are respectively corresponding to the three groups of hidden danger assessment index ranges. The hidden danger assessment index Phd is matched with the three groups of hidden danger assessment index ranges, and the hidden danger level of the ball screw is determined based on the matching result.
10. A fault diagnosis method for a ball screw pair, applied to the ball screw pair fault diagnosis system according to any one of the above claims 1-9, characterized in that, Including: Fault frequency analysis: Read the vibration signal on the surface of the nut corresponding to the ball screw, and analyze to obtain the fault frequency f of the surface defect of the screw. Kurtosis ratio analysis: After sending the fault frequency f of the surface defect of the screw to the technician to determine the filter; Analyze and judge the filtered vibration signal and the original vibration signal to obtain the kurtosis ratio ks. Status determination: Receive the kurtosis ratio ks analyzed for the ball screw, and judge the current status result of the ball screw; The status result includes an abnormal status and a normal status. Abnormal location: When the status result of the ball screw is an abnormal status, use the pre-edited fault location calculation logic to locate the estimated cause of the abnormal status of the ball screw, and send it to the technician in combination with the risk level; The risk level includes a low risk level, a medium risk level, and a high risk level. Normal prediction: When the status result of the ball screw is a normal status, store the kurtosis ratio in the normal status. When the ball screw has g consecutive normal status results, extract the corresponding g groups of kurtosis ratios and perform trend analysis, and output the hidden danger level in combination with the running duration of the ball screw and send it to the technician; The hidden danger level includes a low hidden danger level, a medium hidden danger level, and a high hidden danger level.