A management method and system for parts of a five-axis CNC machine tool

Abstract

The present invention discloses a management method and system for parts of a five-axis numerical control machine tool, relating to the technical field of numerical control machine tool management, including: obtaining key parts and corresponding associated data, establishing a feature data set after extracting the features of the associated data; performing fuzzy processing on the feature data, formulating an evaluation set based on three wear probabilities, obtaining the historical monitoring values of the associated data from the historical operation records and extracting features; performing fuzzy marking on the feature data, establishing a quadratic parabolic membership function, determining the wear interval in combination with the numerical ranges of the feature data under normal and abnormal states, and determining the membership values of the feature data in each wear interval for each wear probability according to fuzzy rules; at the end of each monitoring cycle, obtaining a comprehensive wear probability evaluation value by combining all the feature data of the same part, judging whether wear quality inspection is required, and further determining in combination with the usage time after the last maintenance when it cannot be directly judged; realizing the precise management of parts of a five-axis numerical control machine tool through multi-dimensional data monitoring.

Description

Technical Field

[0001] The present invention relates to the technical field of numerical control machine tool management, and particularly to a management method and system for parts of a five-axis numerical control machine tool. Background Art

[0002] As the core equipment of modern high-end manufacturing equipment, the processing accuracy, efficiency and stability of five-axis numerical control machine tools directly determine the manufacturing quality of complex curved surface parts (such as aero-engine blades, medical implants, etc.); with the development of the manufacturing industry towards intelligence and precision, the reliability requirements of five-axis numerical control machine tools are increasing day by day; however, the core components of machine tools (such as spindle bearings, feed shaft screw nut pairs, tool magazine gears, etc.) are prone to problems such as wear and fatigue under long-term high-load operation, resulting in the accumulation of processing errors, increased downtime maintenance costs and even equipment scrapping;

[0003] The existing parts management methods for five-axis numerical control machine tools mostly rely on traditional regular maintenance strategies (such as preventive maintenance based on time or number of processing times) or passive maintenance modes (i.e., repair after a failure occurs), and it is difficult to achieve real-time status monitoring and precise predictive maintenance;

[0004] Secondly, the existing parts of five-axis numerical control machine tools usually only monitor a single parameter and cannot cover the complex failure mechanisms of multi-physical field coupling, which will lead to easy missed diagnosis of early faults; in addition, when evaluating part wear, the existing technology often treats the monitoring data as accurate numerical values for processing, ignoring the ambiguity and uncertainty existing in the actual data. The wear state of parts is not a clear boundary between either this or that, and there is a gradual process between slight wear and severe wear, while traditional methods are difficult to accurately describe this fuzzy state, resulting in a deviation between the evaluation result and the actual situation. Summary of the Invention

[0005] (1) Technical Problems to be Solved

[0006] In view of the technical problems in the background art, the present invention proposes a management method and system for parts of a five-axis numerical control machine tool. By obtaining key parts and corresponding associated data, and extracting features from the associated data; performing fuzzy processing on the feature data, and determining the membership values of the feature data in each wear interval for each wear probability according to fuzzy rules; at the end of each monitoring cycle, calculating the unit comprehensive evaluation value and judging whether wear quality inspection is required; thus solving the technical problems recorded in the background art.

[0007] (2) Technical Solutions

[0008] To achieve the above object, the present invention is realized through the following technical solutions:

[0009] A management method for parts of a five-axis numerical control machine tool includes:

[0010] Obtain the key parts in a five-axis CNC machine tool; set the monitoring cycle length, establish the associated data sets for each key part, and store the associated data corresponding to the key parts; perform feature extraction operations on the associated data of each key part respectively, and establish the feature data sets for each key part;

[0011] Perform fuzzy processing on each feature in the feature data set of each key part, and formulate an evaluation set based on three wear probabilities; obtain the historical operation records of the five-axis CNC machine tool, obtain the historical monitoring values of the associated data of each key part from them, and perform feature extraction;

[0012] Perform fuzzy marking on each feature data, and establish the corresponding quadratic parabolic membership function; determine the wear interval by combining the numerical intervals of feature data in normal and abnormal states, and based on the formulated fuzzy rules, determine the membership values of the feature data for each wear probability in each wear interval;

[0013] At the end of each monitoring cycle, obtain the membership values of each wear probability corresponding to each feature data, combine all the feature data of the same key part, obtain the comprehensive evaluation value for each wear probability, and obtain the unit comprehensive evaluation value after normalization; combine the unit comprehensive evaluation values of each wear probability to judge whether to perform wear quality inspection operations on each key part. If it cannot be directly judged, further judgment is made based on the usage time of the key part after the last maintenance.

[0014] Specifically, the key parts include spindle bearings, feed axis lead screw nut pairs, tool magazine rotation drive gears, servo motor windings, cutting tools, and swiveling heads, where:

[0015] The associated data of spindle bearings include vibration data, temperature data, and sound data; the associated data of feed axis lead screw nut pairs include vibration data, temperature data, and pressure data; the associated data of tool magazine rotation drive gears include vibration data, sound data, and temperature data; the associated data of servo motor windings include vibration data, current data, and temperature data; the associated data of cutting tools include vibration data, sound data, and temperature data; the associated data of swiveling heads include vibration data, swing angle data, and temperature data.

[0016] Specifically, the feature data of vibration data include the mean and standard deviation of the real-time comprehensive vibration amplitude, the amplitude peak value, the energy ratio of each frequency band, the spectrum peak value, and the spectrum bandwidth; the real-time comprehensive vibration amplitude is obtained through operations on the vibration amplitudes x, y, and z in each of the X, Y, and Z directions obtained;

[0017] The feature data of temperature data include the temperature mean, the temperature standard deviation, and the change rate of temperature within the sliding window;

[0018] The characteristic data of the pressure data include the mean pressure, the peak pressure, the standard deviation of the pressure, the pressure change rate, and the spectral peak;

[0019] The characteristic data of the sound data include the mean sound pressure level, the standard deviation of the sound pressure level, the power spectral density, the spectral bandwidth, and the spectral amplitude of each frequency component;

[0020] The characteristic data of the current data include the maximum value of the current, the peak current, the mean current, the standard deviation of the current, the spectral peak, and the spectral bandwidth;

[0021] The characteristic data of the swing angle data include the maximum value of the swing angle, the peak swing angle, the mean value and the standard deviation of the peak swing angles during the monitoring period.

[0022] Specifically, the evaluation set is {low wear probability, medium wear probability, high wear probability};

[0023] Obtain the historical operation records of the five-axis CNC machine tool, and obtain the historical monitoring values of the associated data of each key component from the historical operation records;

[0024] Summarize the historical record data corresponding to the operating status data of "normal" and "abnormal" respectively, that is, at the end of each monitoring period, judge the associated data monitoring values during the monitoring period corresponding to the operating status of the key components;

[0025] Perform characteristic data extraction operations on the historical monitoring values of the associated data with the operating status data of "normal" and "abnormal" respectively, and obtain the historical characteristic data of a number of historical monitoring values of the associated data.

[0026] Furthermore, use the linear normalization method to normalize the characteristic data sets of each key component, and the historical characteristic data of each key component in the normal and abnormal states;

[0027] Obtain the numerical interval [X 2ij , X 5ij of each characteristic data of each key component in the normal state, and the upper numerical interval [X 1ij , X 3ij and the lower interval [X 4ij , X 6ij of each characteristic data of each key component in the abnormal state, where X ij represents the historical value of the jth characteristic data of the ith key component;

[0028] Perform fuzzy marking on each characteristic data, specifically marked as small, medium, and large; establish a quadratic parabolic membership function.

[0029] Further, substitute each feature data value obtained from the analysis of each monitoring period into the membership function to calculate its membership degree for each fuzzy label; if only one of the calculated membership degrees of the fuzzy labels is non-zero, assign the corresponding membership degree to the membership value of the corresponding wear probability respectively;

[0030] If there is more than one calculated membership degree of the fuzzy labels, that is, two membership degrees are non-zero, then determine the membership value of each wear probability based on the magnitude relationship of the membership degrees of each fuzzy label. Specifically, assign the maximum value of the two membership degrees to the membership value of the low wear probability or the high wear probability.

[0031] Further, denote the membership values of each feature data for the low wear probability, medium wear probability, and high wear probability as a 1ij 、a 2ij 、a 3ij ;

[0032] Combine all the feature data belonging to the same key part to obtain the comprehensive evaluation values b 1i 、b 2i 、b 3i of each wear probability. The expression is:

[0033] where M i represents the total number of feature data of the i-th key part, k ij represents the weight coefficient of the j-th feature data in the i-th key part, and

[0034] Specifically, perform a normalization operation on the comprehensive evaluation values b 1i 、b 2i 、b 3i to obtain the unit comprehensive evaluation values c 1i 、c 2i 、c 3i of each wear probability;

[0035] After the end of each monitoring period, combine the unit comprehensive evaluation values c 1i 、c 2i 、c 3i of each wear probability. If c 1i ≥μ*(c 2i +c 3i ), it means that the i-th key part does not need to undergo wear quality inspection; where μ represents the precision index and μ > 1;

[0036] If c 3i >c 1i and c 3i >c 2i , it means that the i-th key part needs to undergo wear quality inspection;

[0037] Otherwise, it indicates that it is necessary to further determine whether the i-th key part needs to be inspected for wear.

[0038] Specifically, the further determination is as follows:

[0039] Obtain the historical operation record and historical maintenance record of the key part, obtain the usage duration and usage time data of the key part from the historical operation record, and obtain the historical maintenance time data of the key part from the historical maintenance record;

[0040] Combine the historical operation record and historical maintenance record to obtain the usage duration data of the key part in every two adjacent historical maintenance records; perform a mean operation on all the usage duration data of the key part in two adjacent maintenance records to obtain the unit usage duration data T of the key part i ;

[0041] Combine the current usage time t of the key part after the last maintenance i , if t i ≥λ*T i , it indicates that the key part needs to be inspected for wear. Here, λ represents the time correction index, and λ > 0.7.

[0042] A management system for parts of a five-axis CNC machine tool, comprising:

[0043] A data acquisition and processing module, used to acquire key parts in a five-axis CNC machine tool; set the monitoring cycle length, establish an associated data set for each key part, and store the associated data of the corresponding key part; perform feature extraction operations on the associated data of each key part respectively, and establish a feature data set for each key part;

[0044] A fuzzy algorithm processing module, used to perform fuzzy processing on each feature in the feature data set of each key part, and formulate an evaluation set based on three wear probabilities; obtain the historical operation record of the five-axis CNC machine tool, obtain the historical monitoring value of the associated data of each key part from it and perform feature extraction; perform fuzzy marking on each feature data, and establish a corresponding quadratic parabolic membership function; determine the wear interval in combination with the numerical intervals of feature data in normal and abnormal states, and determine the membership value of the feature data in each wear interval for each wear probability based on the formulated fuzzy rules;

[0045] The wear judgment module is used to, at the end of each monitoring cycle, obtain the membership values of each wear probability corresponding to each feature data, combine all the feature data of the same key part to obtain the comprehensive evaluation value of each wear probability, and obtain the unit comprehensive evaluation value after normalization; combine the unit comprehensive evaluation values of each wear probability to judge whether to perform wear quality inspection operations on each key part. If it cannot be directly judged, further judgment is made based on the usage time of the key part after the previous maintenance.

[0046] (III) Beneficial effects

[0047] The present invention provides a management method and system for parts of a five-axis CNC machine tool, having the following beneficial effects:

[0048] 1. For the characteristics of each key part, the data types to be monitored, the sensor placement positions, and the sensor selection types are carefully planned, and a comprehensive data monitoring system is established; a reasonable monitoring cycle is set, and the associated data of each key part is collected and stored in real time, providing a rich and accurate data basis for subsequent data analysis and wear assessment, helping to promptly discover potential problems of the parts, ensuring the stable operation of the machine tool, and improving the processing quality and efficiency;

[0049] 2. By means of fuzzy processing, a factor set and an evaluation set for wear assessment are established, and historical feature data is extracted using historical operation records and operation status data; the feature data is normalized and fuzzily marked, and combined with a quadratic parabolic membership function and fuzzy rules, a quantitative assessment of the wear probability of each feature data is realized; this method can comprehensively consider the influence of various factors on part wear, avoid the limitations of single-index assessment, improve the accuracy and reliability of wear assessment, and provide a scientific basis for subsequent wear judgment and decision-making;

[0050] 3. Wear judgment is carried out according to the set accuracy index; for key parts with medium wear probability and that cannot be directly judged, further combined with historical operation records and maintenance records, considering the usage duration and maintenance interval of the parts, a more detailed assessment is carried out; this can comprehensively consider real-time monitoring data and historical experience, realize accurate judgment of the wear status of key parts, avoid unnecessary maintenance, and ensure that key parts can be promptly processed when wear occurs, effectively reducing the maintenance cost of the machine tool and improving production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 is a schematic diagram of the steps of a management method for parts of a five-axis CNC machine tool provided by the present invention;

[0052] Figure 2 is a schematic diagram of the structure of a management system for parts of a five-axis CNC machine tool provided by the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0053] Next, in combination with the accompanying drawings in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the protection scope of the present invention.

[0054] Reference Figure 1 , the present invention provides a management method for parts of a five-axis CNC machine tool, including:

[0055] Step 1: Obtain the key parts in the five-axis CNC machine tool; establish an associated data set for each key part and store the associated data corresponding to the key part; perform feature extraction operations on the associated data of each key part respectively, and establish a feature data set for each key part;

[0056] The first step includes the following steps:

[0057] Step 101: Obtain all key parts related to the work in the five-axis CNC machine tool. The key parts refer to the core components that directly affect the machining accuracy, efficiency, stability and function realization of the machine tool. The reliability of these parts determines whether the machine tool can complete complex machining tasks, specifically including: spindle bearings, feed axis lead screw nut pairs, tool magazine rotation drive gears, servo motor windings, cutting tools and swivel heads; among them:

[0058] The spindle bearing is used to support the high-speed rotation of the spindle, bear radial and axial loads, and ensure machining stability; during the operation of the five-axis CNC machine tool, it is necessary to monitor the vibration data, temperature data and sound data of the spindle bearing; among them, the vibration data is used to monitor the wear characteristic frequencies of the inner and outer rings and rolling elements of the bearing; the temperature data is used to prevent lubrication failure or material degradation caused by overheating; the sound data is used to capture high-frequency impact sounds (such as spalling, cracks) to assist vibration analysis; place the vibration sensor near the bearing on the spindle housing, place the temperature sensor on the outer ring of the bearing, and place the sound sensor on the surface of the spindle housing;

[0059] The feed axis lead screw nut pair is used to convert the rotational motion of the servo motor into a linear motion to achieve high-precision positioning; during the operation of the five-axis CNC machine tool, it is necessary to monitor the vibration data, temperature data and pressure data of the feed axis lead screw nut pair; among them, the vibration data is used to monitor the axial and radial vibrations of the lead screw nut pair to identify ball wear or insufficient preload; the temperature data is used to monitor the temperature rise of the nut and lead screw to prevent positioning errors caused by thermal deformation; the pressure data is used to monitor the ball preload to ensure the transmission stiffness; place the vibration sensor on the surface of the nut seat, place the temperature sensor on the surface of the nut, and place the pressure sensor at the preload adjustment device;

[0060] The tool magazine rotation drive gear is used to achieve fast and precise tool change; during the operation of a five-axis CNC machine tool, it is necessary to monitor the vibration data, sound data, and temperature data of the tool magazine rotation drive gear; among them, the vibration data is used to monitor gear meshing impact and identify tooth surface wear or pitting; the sound data is used to capture abnormal gear meshing noise; the temperature data is used to monitor the oil temperature of the gearbox and prevent lubrication failure or local overheating; place the vibration sensor on the housing of the tool magazine reducer, place the sound sensor on the surface of the tool magazine motor end cover, and place the temperature sensor on the surface of the gearbox oil sump wall;

[0061] The servo motor winding is used to drive the movement of each axis; during the operation of a five-axis CNC machine tool, it is necessary to monitor the vibration data, current data, and temperature data of the tool magazine rotation drive gear; among them, the vibration data is used to monitor motor rotor imbalance or bearing wear; the current data is used to monitor current fluctuations and identify inter-turn short circuits or load mutations; the temperature data is used to monitor the temperature rise of the winding and prevent insulation aging or overload burnout; place the vibration sensor on the surface of the motor end cover, connect the current sensor to the motor power supply line, and place the temperature sensor on the surface of the motor winding end;

[0062] The tool is used for the cutting operation of the direct target object; during the operation of a five-axis CNC machine tool, it is necessary to monitor the vibration data, sound data, and temperature data of the tool; among them, the vibration data is used to monitor the change of cutting force and identify tool wear or chipping; the sound data is used to capture the sudden change of cutting noise and assist in judging the tool state; the temperature data is used to monitor the temperature of the cutting area and prevent thermal deformation or crater wear; place the vibration sensor at the spindle taper hole, place the sound sensor on the surface of the machine tool column, and place the temperature sensor near the cutting edge of the tool;

[0063] The swivel head is used to make the tool achieve multi-angle swing, cooperate with the rotary table and other moving axes to achieve the processing of complex curved surfaces; during the operation of a five-axis CNC machine tool, it is necessary to monitor the vibration data, angle data, and temperature data of the swivel head; among them, the vibration data is used to monitor the vibration situation of the swivel head during the swing process and check whether there are abnormalities in the transmission components, bearings, etc. of the swivel head; the angle data is used to monitor the swing angle of the swivel head and prevent deviation of the rotation angle due to wear of the swivel head parts; the temperature data is used to monitor the temperature of the swivel head drive motor and transmission components and prevent overheating problems; place the vibration sensor on the housing of the swivel head, place the swing angle sensor on the swing axis of the swivel head, and place the temperature sensor on the surface of the swivel head drive motor and reducer;

[0064] The vibration sensor uses a three-axis acceleration sensor, the temperature sensor uses an infrared temperature measurement probe, the sound sensor uses a MEMS microphone, the current sensor uses a Hall current sensor, and the swing angle sensor uses an angle encoder device;

[0065] Step 102: Set the monitoring period length and the monitoring period interval. After each complete monitoring period interval, monitor the data for a duration of one monitoring period length, that is, in each monitoring period, perform real-time monitoring of the associated data of each key component. The associated data represents the data monitored for each component. For example, the associated data of the spindle bearing is the vibration data, temperature data, and sound data detected;

[0066] Establish an associated data set for each key component. The associated data set of each key component stores the corresponding associated data subsets respectively. For example, the associated data set of the spindle bearing stores the vibration data subset, temperature data subset, and sound data subset, and each subset stores the corresponding data detected in real time for each monitoring period;

[0067] Step 103: Extract features from the real-time data monitored in each associated data subset to extract the feature data corresponding to each associated data subset;

[0068] For the vibration data monitored in each monitoring period in the associated data subset, perform double integration on the acceleration signals collected by the three-axis acceleration sensor to obtain the vibration amplitudes x, y, and z in the X, Y, and Z directions respectively ， Based on the formula Combine the vibration amplitudes in the three directions to obtain the real-time comprehensive vibration amplitude of each key component;

[0069] For the feature extraction of vibration data, it includes: performing mean operation and standard deviation operation on the real-time comprehensive vibration amplitude monitored in each monitoring period, and obtaining the amplitude peak value in each monitoring period; decomposing the vibration signal into 4-layer wavelet packets based on the short-time Fourier transform, and calculating the energy proportion of each frequency band; converting the time-domain signal into a frequency-domain signal based on the fast Fourier transform, and extracting frequency-domain features, including spectral peak value and spectral bandwidth;

[0070] For the feature extraction of temperature data, it includes: calculating the temperature mean value and temperature standard deviation in each monitoring period; calculating the temperature change rate based on the set sliding window;

[0071] For the feature extraction of pressure data, it includes: obtaining and calculating the pressure mean value, pressure peak value, and pressure standard deviation received by the corresponding key component (i.e., the feed shaft lead screw nut pair) in each monitoring period; similarly calculating the pressure change rate based on the set sliding window; converting the time-domain signal into a frequency-domain signal based on the fast Fourier transform, and extracting frequency-domain features, including spectral peak value;

[0072] For the feature extraction of sound data, it includes: calculating the mean sound pressure level and the standard deviation of the sound pressure level monitored in each monitoring period; performing Fourier transform on the segmented and windowed sound signal to convert the time-domain signal into a frequency-domain signal, obtaining a spectrogram, and extracting and analyzing the power spectral density, spectral bandwidth, and spectral amplitudes of each frequency component from the spectrogram;

[0073] For the feature extraction of current data, it includes: obtaining and calculating the maximum value, peak value, mean value, and standard deviation of the current monitored in each monitoring period; performing Fourier transform on the current signal to decompose the time-domain signal into sine and cosine components, generating a spectrogram, and extracting the spectral peak value and spectral bandwidth;

[0074] For the feature extraction of the swing angle, it includes: obtaining and calculating the maximum and minimum values of the swing angle monitored in each monitoring period, and the peak value of the swing angle, that is, the swing amplitude range; calculating the mean value and standard deviation of multiple peak values of the swing angle;

[0075] Step 104: Establish a feature data set for each key part. Based on the monitoring data of each key part, select the corresponding feature data from Step 103 and put it into the feature data set of the corresponding key part.

[0076] During use, combine the content in Steps 101 to 104:

[0077] For the characteristics of each key part, the types of data to be monitored, the placement positions of sensors, and the sensor selection are detailedly planned, and a comprehensive data monitoring system is established; set a reasonable monitoring period, collect and store the associated data of each key part in real time, providing a rich and accurate data basis for subsequent data analysis and wear assessment, helping to timely discover potential problems of parts, ensuring the stable operation of the machine tool, and improving the machining quality and efficiency.

[0078] Step Two: Perform fuzzy processing on each feature in the feature data set of each key part, and formulate an evaluation set based on three wear probabilities; obtain the historical operation records of the five-axis CNC machine tool, obtain the historical monitoring values of the associated data of each key part from them and perform feature extraction; perform fuzzy marking on each feature data and establish a corresponding quadratic parabolic membership function; combine the numerical intervals of the feature data in the normal and abnormal states to determine the wear interval, and based on the formulated fuzzy rules, determine the membership values of the feature data in each wear interval for each wear probability;

[0079] The steps in Step Two include the following steps:

[0080] Step 201: Perform fuzzy processing on each feature in the feature data set of each key part, and establish a factor set and an evaluation set for wear assessment;

[0081] Among them, the factor set is a general set composed of various factors that affect the evaluation object, that is, the extracted characteristic data; using V to represent the factor set, then V = {v1, v2,..., v n}, where n represents the total number of characteristics extracted from each key part;

[0082] The evaluation set is a set composed of various possible results of the evaluation object, that is, to judge whether wear occurs; using R to represent the evaluation set, the evaluation set R of this method is {low wear probability, medium wear probability, high wear probability};

[0083] Step 202: Obtain the historical operation records of the five-axis CNC machine tool. The historical operation records record the associated data monitoring values of each key part; obtain the historical monitoring values of the associated data of each key part from the historical operation records;

[0084] The historical operation records also record the operation status data of each key part, that is, to judge whether each key part is operating normally, and summarize the historical record data corresponding to "normal" and "abnormal" for the operation status data respectively, that is, at the end of each monitoring cycle, judge the operation status ("normal" or "abnormal") of the key part and the associated data monitoring values within the monitoring cycle;

[0085] Perform the feature data extraction operation in step 103 on the historical monitoring values of the associated data with the operation status data of "normal" and "abnormal" respectively, obtain the historical feature data of several historical monitoring values of the associated data, and correspond them one by one with the data in the feature data set of each key part;

[0086] Step 203: Use the linear normalization method to normalize the feature data sets of each key part and the historical feature data of each key part in the normal and abnormal states;

[0087] Denote the universe of discourse U as [0, 1], summarize the historical feature data of each key part in the normal and abnormal states, and determine the wear interval for each feature data of each key part respectively. Specifically, obtain the numerical interval [X 2ij , X 5ij of each feature data of each key part in the normal state, and the upper numerical interval [X 1ij , X 3ij and the lower interval [X 4ij , X 6ij of each feature data of each key part in the abnormal state, where X ij represents the historical value of the jth feature data of the ith key part;

[0088] Step 204: Fuzzily label each feature data, specifically as small, medium, and large; establish a quadratic parabolic membership function. Among them, the membership function of the feature labeled as small is:

[0089]

[0090] The membership function of the feature labeled as medium is:

[0091]

[0092] The membership function of the feature labeled as high is:

[0093]

[0094] Combine the numerical interval [X 2ij , X 5ij of each feature data of each key part in the normal state with the upper numerical interval [X 1ij , X 3ij and the lower interval [X 4ij , X 6ij of each feature data of each key part in the abnormal state to obtain 5 intervals, and combine them with the boundaries of the universe of discourse to obtain a total of 7 wear intervals, which are [0, X 1ij , (X 1ij , X 2ij , (X 2ij , X 3ij , (X 3ij , X 4ij , (X 4ij , X 5ij , (X 5ij , X 6ij , (X 6ij , 1];

[0095] Step 205: Determine the fuzzy rule that when the fuzzy label is small and large, the wear probability is high, and when the fuzzy label is medium, the wear probability is low. Specifically:

[0096] When the feature data X ij belongs to [0, X 1ij , the wear probability of the key part is high; when the feature data X ij belongs to (X 1ij , X 2ij ), the wear probability of the key part is high or medium; when the feature data X ij belongs to (X 2ij , X 3ij ), the wear probability of the key part is medium or low; when the feature data X ij belongs to (X 3ij , X 4ijWhen, the wear probability of the key parts is low; when the characteristic data X ij belongs to (X 4ij , X 5ij , the wear probability of the key parts is low or medium; when the characteristic data X ij belongs to (X 5ij , X 6ij , the wear probability of the key parts is medium or high; when the characteristic data X ij belongs to (X 6ij , 1], the wear probability of the key parts is high;

[0097] Step 206: Substitute each characteristic data value obtained from the analysis of each monitoring period into the membership function to calculate its membership degree to each fuzzy label; if only one of the calculated membership degrees of the fuzzy labels is not 0, that is, X ij belongs to [0, X 1ij , (X 3ij , X 4ij or (X 6ij , 1], respectively assign the corresponding membership degree to the membership value of the corresponding wear probability; if more than one of the calculated membership degrees of the fuzzy labels is not 0, that is, two membership degrees are not 0, then determine the membership value of each wear probability based on the magnitude relationship of the membership degrees of each fuzzy label. Specifically, assign the maximum value of the two membership degrees to the membership value of low wear probability or high wear probability.

[0098] When in use, combine the content in Steps 201 to 206:

[0099] Through fuzzy processing, establish the factor set and evaluation set for wear assessment, and use historical operation records and operation status data to extract historical characteristic data; perform normalization processing and fuzzy labeling on the characteristic data, and combine the quadratic parabolic membership function and fuzzy rules to realize the quantitative assessment of the wear probability of each characteristic data; this method can comprehensively consider the influence of various factors on part wear, avoid the limitations of single-index assessment, improve the accuracy and reliability of wear assessment, and provide a scientific basis for subsequent wear judgment and decision-making.

[0100] Step 3: At the end of each monitoring period, obtain the membership value of each wear probability corresponding to each characteristic data, combine all the characteristic data of the same key part to obtain the comprehensive evaluation value of each wear probability, and obtain the unit comprehensive evaluation value after normalization operation; combine the unit comprehensive evaluation value of each wear probability to judge whether to perform wear quality inspection operations on each key part. If it cannot be directly judged, further judgment is made based on the usage time of the key part after the previous maintenance.

[0101] The steps in Step 3 include the following steps:

[0102] Step 301. At the end of each monitoring cycle, obtain the membership values of each wear probability corresponding to each characteristic data respectively. Denote the membership values of each characteristic data for low wear probability, medium wear probability, and high wear probability as a 1ij 、a 2ij 、a 3ij ;

[0103] Combine all the characteristic data belonging to the same key part to obtain the comprehensive evaluation values b 1i 、b 2i 、b 3i of each wear probability. The expression is:

[0104] where M i represents the total number of characteristic data of the i-th key part, and k ij represents the weight coefficient of the j-th characteristic data in the i-th key part, which is specifically calculated by the entropy weight method. The entropy weight method is a method based on the information entropy theory that objectively calculates the weights of each index by quantifying the dispersion degree (entropy value) of the index data. The smaller the entropy value, the stronger the data volatility, the greater the amount of information, and the higher the corresponding weight. And

[0105] Step 302. Perform a normalization operation on the comprehensive evaluation values b 1i 、b 2i 、b 3i of each wear probability to obtain the unit comprehensive evaluation values c 1i 、c 2i 、c 3i ;

[0106] After the end of each monitoring cycle, combine the unit comprehensive evaluation values c 1i 、c 2i 、c 3i of each wear probability to determine whether the corresponding key part is worn. Specifically:

[0107] If c 1i ≥μ*(c 2i +c 3i ), it means that the wear probability of the i-th key part is low, that is, it means that the i-th key part does not need to be inspected for wear at present; where μ represents the precision index, and the specific value is set by the five-axis CNC machine tool management personnel themselves, and μ>1;

[0108] If c 3i >c 1i and c 3i >c 2i , it means that the wear probability of the i-th key part is high, that is, it means that the i-th key part needs to be inspected for wear at present;

[0109] Otherwise, in the wear probability of the i-th key component, it needs to be further determined, specifically as follows:

[0110] Obtain the historical operation record and historical maintenance record of this key component, obtain the usage duration and usage time data of this key component from the historical operation record, and obtain the historical maintenance time data of this key component from the historical maintenance record; combine the historical operation record and historical maintenance record to obtain the usage time data of this key component after each historical maintenance, that is, the usage duration data of this key component in two adjacent maintenance records; perform a mean operation on all the usage duration data of this key component in two adjacent maintenance records to obtain the unit usage duration data T of this key component i ;

[0111] Combine the usage time t of this key component after the last maintenance currently i , if t i ≥λ*T i , it indicates that the time interval of this key component since the last maintenance is too long, and wear quality inspection needs to be carried out on this key component, where λ represents the time correction index, and the specific value is set by the five-axis CNC machine tool management personnel themselves, and λ > 0.7.

[0112] During use, combine the content in steps 301 to 302:

[0113] At the end of each monitoring cycle, according to the wear probability membership value of each characteristic data, combined with the weight coefficient calculated by the entropy weight method, comprehensively evaluate the wear probability of each key component; obtain the unit comprehensive evaluation value through normalization operation, and perform wear judgment according to the set accuracy index; for the key components with wear probability that cannot be directly judged, further combine the historical operation record and maintenance record, and consider the usage duration and maintenance interval of the components for more detailed evaluation; this can comprehensively consider real-time monitoring data and historical experience, realize the accurate judgment of the wear state of key components, avoid unnecessary maintenance, ensure that key components can be processed in time when wear occurs, effectively reduce the maintenance cost of the machine tool, and improve production efficiency.

[0114] Refer to Figure 2 , the present invention also provides a management system for five-axis CNC machine tool parts, including:

[0115] A data acquisition and processing module, used to acquire key components in a five-axis CNC machine tool; establish an associated data set for each key component, store the associated data of the corresponding key component; perform feature extraction operations on the associated data of each key component respectively, and establish a feature data set for each key component;

[0116] The fuzzy algorithm processing module is used to perform fuzzy processing on each feature in each key part feature dataset, and formulate an evaluation set based on three wear probabilities; obtain the historical operation records of the five-axis CNC machine tool, and obtain the historical monitoring values of the associated data of each key part from them and perform feature extraction; perform fuzzy marking on each feature data, and establish a corresponding quadratic parabolic membership function; combine the numerical intervals of the feature data in the normal and abnormal states to determine the wear interval, and based on the formulated fuzzy rules, determine the membership values of the feature data in each wear interval for each wear probability.

[0117] The wear judgment module is used to, at the end of each monitoring cycle, obtain the membership values of each wear probability corresponding to each feature data, combine all the feature data of the same key part to obtain the comprehensive evaluation value of each wear probability, and obtain the unit comprehensive evaluation value after normalization; combine the unit comprehensive evaluation values of each wear probability to judge whether to perform wear quality inspection operations on each key part. If it cannot be directly judged, further judgment is made based on the usage time of the key part after the last maintenance.

[0118] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer storage medium or transmitted through a computer storage medium.

[0119] The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired manner (such as coaxial cable, optical fiber, digital subscriber line (DSL)) or wirelessly (such as infrared, wireless, microwave, etc.). The computer storage medium can be any available medium that the computer can access or a data storage device such as a server or data center that includes one or more integrated available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid state disk (SSD)).

[0120] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A management method for parts of a five-axis CNC machine tool, characterized in that: The method includes the following steps: Obtain key components in a five-axis CNC machine tool; set the monitoring cycle length, establish an associated data set for each key component, and store the associated data corresponding to each key component; perform feature extraction operations on the associated data of each key component respectively, and establish a feature data set for each key component. Perform fuzzy processing on each feature in the feature data set of each key component, and formulate an evaluation set based on three wear probabilities; obtain the historical operation records of the five-axis CNC machine tool, and obtain the historical monitoring values of the associated data of each key component from them and perform feature extraction. Perform fuzzy marking on each feature data, and establish a corresponding quadratic parabolic membership function; determine the wear interval by combining the numerical intervals of feature data in normal and abnormal states, and based on the formulated fuzzy rules, determine the membership values of the feature data for each wear probability in each wear interval. At the end of each monitoring cycle, obtain the membership values of each wear probability corresponding to each feature data, combine all the feature data of the same key component to obtain a comprehensive evaluation value for each wear probability, and obtain a unit comprehensive evaluation value after normalization; combine the unit comprehensive evaluation values of each wear probability to determine whether to perform a wear quality inspection operation on each key component. If it cannot be directly judged, further judgment is made based on the usage time of the key component after the last maintenance.

2. A management method for parts of a five-axis CNC machine tool according to claim 1, wherein: The key components include spindle bearings, feed shaft screw-nut pairs, tool magazine rotation drive gears, servo motor windings, cutting tools, and swiveling heads, among which: The associated data of the spindle bearing includes vibration data, temperature data, and sound data; the associated data of the feed shaft screw-nut pair includes vibration data, temperature data, and pressure data; the associated data of the tool magazine rotation drive gear includes vibration data, sound data, and temperature data; the associated data of the servo motor winding includes vibration data, current data, and temperature data; the associated data of the cutting tool includes vibration data, sound data, and temperature data; the associated data of the swiveling head includes vibration data, swing angle data, and temperature data.

3. A management method for parts of a five-axis CNC machine tool according to claim 2, wherein: The characteristic data of the vibration data include the mean and standard deviation of the real-time comprehensive vibration amplitude, the amplitude peak value, the energy proportion of each frequency band, the spectrum peak value, and the spectrum bandwidth; the real-time comprehensive vibration amplitude is calculated through the vibration amplitudes x, y, and z in each of the X, Y, and Z directions obtained; The feature data of the temperature data includes temperature mean, temperature standard deviation, and the change rate of temperature within the sliding window; The feature data of the pressure data includes pressure mean, pressure peak value, pressure standard deviation, pressure change rate, and spectral peak value; The feature data of the sound data includes sound pressure level mean, sound pressure level standard deviation, power spectral density, spectral bandwidth, and spectral amplitude of each frequency component; The feature data of the current data includes current maximum and minimum values, current peak value, current mean, current standard deviation, spectral peak value, and spectral bandwidth; The feature data of the swing angle data includes swing angle maximum and minimum values, swing angle peak value, and the mean and standard deviation of multiple swing angle peak values within the monitoring cycle.

4. A management method for parts of a five-axis CNC machine tool according to claim 3, wherein: The evaluation set is {low wear probability, medium wear probability, high wear probability}; Obtain the historical operation records of the five-axis CNC machine tool, and obtain the associated data historical monitoring values of each key component from the historical operation records; Summarize the historical record data corresponding to the operation status data of "normal" and "abnormal" respectively, that is, at the end of each monitoring cycle, judge the associated data monitoring values within the monitoring cycle corresponding to the operation status of the key component; Perform feature data extraction operations on the associated data historical monitoring values of the operation status data of "normal" and "abnormal" respectively, and obtain the historical feature data of a number of associated data historical monitoring values.

5. A management method for parts of a five-axis CNC machine tool according to claim 4, characterized in that: Use the linear normalization method to normalize the feature data sets of each key component, as well as the historical feature data of each key component in the normal and abnormal states; Obtain the numerical range [X 2ij , X 5ij of each characteristic data of each key component in the normal state, and the upper numerical range [X 1ij , X 3ij and the lower numerical range [X 4ij , X 6ij of each characteristic data of each key component in the abnormal state, where X ij represents the historical value of the jth characteristic data of the ith key component; Perform fuzzy marking on each feature data, specifically marked as small, medium, and large; establish a quadratic parabolic membership function.

6. A management method for parts of a five-axis CNC machine tool according to claim 5, characterized in that: Substitute each feature data value obtained by analyzing each monitoring cycle into the membership function to calculate its membership degree to each fuzzy mark; If only one of the calculated membership degrees of the fuzzy marks is not 0, assign the corresponding membership degree to the membership value of the corresponding wear probability; If there is more than one calculated membership degree of the fuzzy mark, that is, two membership degrees are not 0, then determine the membership value of each wear probability based on the magnitude relationship of the membership degrees of each fuzzy mark. Specifically, assign the maximum value of the two membership degrees to the membership value of the low wear probability or the high wear probability.

7. A management method for parts of a five-axis CNC machine tool according to claim 6, characterized in that: Let the membership values of each characteristic data for low wear probability, medium wear probability, and high wear probability be a 1ij 、a 2ij 、a 3ij ; Combine all the feature data belonging to the same key part to obtain the comprehensive evaluation value b of each wear probability 1i 、b 2i 、b 3i , and the expression is: Among them, M i represents the total number of characteristic data of the i-th key part, k ij represents the weight coefficient of the j-th characteristic data in the i-th key part, and 8. A management method for parts of a five-axis CNC machine tool according to claim 7, characterized in that: The comprehensive evaluation value b for each wear probability 1i 、b 2i 、b 3i Perform a normalization operation to obtain the unit comprehensive evaluation value c for each wear probability 1i 、c 2i 、c 3i ; After the end of each monitoring cycle, combining the unit comprehensive evaluation value c of each wear probability 1i 、c 2i 、c 3i , if c 1i ≥μ*(c 2i +c 3i ), it means that the i-th key part does not need to be inspected for wear quality; where μ represents the precision index and μ>1; If c 3i > c 1i and c 3i > c 2i , it means that the i-th key part needs to be inspected for wear and tear; Otherwise, it means that the i-th key component needs to be further determined whether wear quality inspection is required.

9. The management method of a five-axis CNC machine tool part according to claim 8, characterized in that: The further determination is specifically: Obtain the historical operation record and historical maintenance record of the key component, obtain the usage duration and usage time data of the key component from the historical operation record, and obtain the historical maintenance time data of the key component from the historical maintenance record; Obtain the usage duration data of this key part in each pair of adjacent historical maintenance records in combination with the historical operation records and historical maintenance records; perform a mean operation on all the usage duration data of this key part in two adjacent maintenance records to obtain the unit usage duration data T of this key part i ; Combined with the usage time t of the current key part after the last maintenance i , if t i ≥λ*T i , it means that wear quality inspection needs to be carried out on this key part, where λ represents the time correction index and λ > 0.

7.

10. A management system for parts of a five-axis CNC machine tool, characterized in that, Including: A data acquisition and processing module for acquiring key components in a five-axis CNC machine tool; Set the monitoring cycle length, establish an associated data set for each key component, store the associated data of the corresponding key component; perform feature extraction operations on the associated data of each key component respectively, and establish a feature data set for each key component; A fuzzy algorithm processing module for performing fuzzy processing on each feature in the feature data set of each key component, formulating an evaluation set based on three wear probabilities; obtaining the historical operation records of the five-axis CNC machine tool, obtaining the associated data historical monitoring values of each key component from them and performing feature extraction; performing fuzzy marking on each feature data and establishing a corresponding quadratic parabolic membership function; determining the wear interval in combination with the numerical intervals of the feature data in the normal and abnormal states, and determining the membership value of the feature data to each wear probability in each wear interval based on the formulated fuzzy rules; The wear judgment module is used to, at the end of each monitoring cycle, obtain the membership values of each wear probability corresponding to each feature data, combine all the feature data of the same key part to obtain the comprehensive evaluation value of each wear probability, and obtain the unit comprehensive evaluation value after normalization; combine the unit comprehensive evaluation values of each wear probability to judge whether to perform wear quality inspection operations on each key part. If it cannot be directly judged, further judgment is made based on the usage time of the key part after the last overhaul.

Images