Method for monitoring the fatigue of a grooving machine

Abstract

The application discloses a slotting machine fatigue degree monitoring method, and relates to the technical field of slotting machine fatigue degree monitoring.The slotting machine fatigue degree monitoring method comprises the following steps: S1, working condition acquisition: the feeding speed and the cutting depth of the slotting machine are acquired in real time as current working condition parameters, parameter acquisition is completed through corresponding sensors, and the accuracy of working condition data is ensured; S2, reference construction and matching: a preset working condition-reference mapping library is constructed, the mapping library is constructed on the basis of the initial state of the health of the slotting machine, and fatigue characteristic reference data is collected under different working condition combinations of feeding speed and cutting depth; the slotting machine fatigue degree monitoring method realizes accurate matching of health reference values under different working conditions of feeding speed and cutting depth through the preset working condition-reference mapping library, multi-dimensional guarantee is provided for the accuracy of monitoring data and results, misjudgment and missed judgment caused by subjective judgment and instantaneous interference are avoided, and reliable support is provided for equipment fatigue monitoring.

Description

Technical Field

[0001] This invention relates to the field of fatigue monitoring technology for grooving machines, specifically to a fatigue monitoring method for grooving machines. Background Technology

[0002] Fatigue monitoring of grooving machines is an important task for monitoring and analyzing fatigue damage that may occur during long-term use. Because grooving machines operate under high loads, prolonged use can lead to gradual fatigue damage to the equipment's components, affecting its efficiency and lifespan.

[0003] Currently, the mainstream fatigue monitoring methods for grooving machines in the industry mainly rely on vibration monitoring, temperature monitoring, and current / power monitoring, supplemented by manual inspection and periodic disassembly. Some high-end equipment uses acoustic emission, laser, or strain monitoring. However, all of the above-mentioned existing technologies have obvious shortcomings and cannot meet the requirements of high-precision, high-reliability, and full-lifecycle intelligent monitoring. Among them, vibration monitoring, as the most common and mature method, can only reflect obvious faults in the middle and late stages and is not sensitive to early micro-cracks and initial fatigue damage, so it cannot achieve true early warning. At the same time, it is easily affected by environmental vibration, cutting impact, frame resonance, etc. Temperature monitoring is a lagging monitoring method and can only reflect the severe stage of fatigue wear through local heating. When the temperature rises significantly, the fatigue damage of the component is close to failure, and it is greatly affected by the ambient temperature, cutting cooling, and ventilation conditions, resulting in poor measurement stability. Current / power monitoring can only reflect the load of the equipment and has no direct correlation with the fatigue damage inside the component. It cannot distinguish whether the current fluctuation is caused by component fatigue or by changes in operating conditions such as tool wear, material changes, and feed rate fluctuations. Furthermore, it cannot monitor the fatigue state of structural components such as the frame, spindle, roller, and bearing housing, limiting its monitoring range. Manual inspection and periodic disassembly inspection rely heavily on the operator's experience, are highly subjective, and have inconsistent judgment standards, making them prone to missed or incorrect judgments. Therefore, this paper proposes an efficient, accurate, reliable, and industrially adaptable fatigue monitoring solution. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a fatigue monitoring method for grooving machines, which solves the problems mentioned in the background section.

[0005] To achieve the above objectives, the present invention provides a fatigue monitoring method for a grooving machine, comprising the following steps: S1. Working Condition Acquisition: The feed speed and cutting depth of the grooving machine are acquired in real time as current working condition parameters. The parameters are collected through corresponding sensors to ensure the accuracy of the working condition data. S2. Benchmark Construction and Matching: Preset Working Condition-Benchmark Mapping Library. The mapping library is constructed based on the initial healthy state of the grooving machine, by collecting fatigue characteristic benchmark data under working condition combinations of different feed speeds and cutting depths. According to the current working condition parameters, the corresponding healthy benchmark value is matched from the mapping library. S3. Condition Monitoring: Real-time acquisition of fatigue characteristic data of key components of the grooving machine. The fatigue characteristic data includes at least one of vibration signals, current signals, and temperature signals. Simultaneously, fatigue signals related to the saw blade can be selectively acquired. The acquired signals undergo anti-interference processing to extract true and effective features. The anti-interference processing employs an adaptive filtering algorithm, the expression of which is: in, This is the filtered output signal. For adaptive filtering weight vectors, The original acquired signal, The filtering error signal is used to iteratively update the weight vector through the minimum mean square error criterion, thereby achieving adaptive suppression of on-site clutter interference. S4. Data Verification: Verify the validity of the real-time collected fatigue characteristic data and current operating parameters, and remove abnormal collected data and abnormal fluctuation data of operating parameters to ensure the reliability of the data used for fatigue judgment. S5. Fatigue Judgment: The verified fatigue characteristic data is compared with the matched health benchmark value. When the data deviation exceeds a preset threshold or meets specific fatigue judgment conditions, the grooving machine is judged to be fatigued or malfunctioning. The data deviation is calculated using a relative deviation algorithm, and the expression is: Among them, This is a relative deviation. For fatigue characteristic data collected in real time, For the matched health baseline value, when When the threshold is exceeded, fatigue abnormality is determined.

[0006] S6. Data Archiving and Update: The operating condition parameters, effective fatigue characteristic data, and fatigue judgment results from each monitoring session are categorized and archived. The health baseline values ​​and preset thresholds of the operating condition-benchmark mapping library are periodically optimized based on the archived data to improve subsequent monitoring accuracy. The health baseline value optimization uses a moving average algorithm, expressed as: in, The optimized health baseline value, The number of samples in the archived data. The effective fatigue feature data for the i-th archive.

[0007] According to the above technical solution, the specific steps for obtaining the working condition are as follows: S1.1. The cutting depth is obtained by measuring the cutting depth of the grooving machine's cutting wheel using a displacement sensor or photoelectric sensor; S1.2 Measure the moving speed of the grooving machine relative to the workpiece using an encoder, Hall sensor or inertial measurement unit to obtain the feed speed, ensuring that the collected working condition parameters are accurately matched with the working condition combination in the subsequent working condition-reference mapping library.

[0008] According to the above technical solution, the vibration signal acquisition and processing method in the condition monitoring step includes: S2.1 Install the first vibration sensor on the bearing housing of the grooving machine and the second vibration sensor on the outer shell. After collecting the vibration data of both, use the transfer function or adaptive filtering algorithm to remove the resonance interference of the outer shell and extract the real vibration signal of the bearing body. S2.2 Simultaneously acquire the spindle speed signal of the grooving machine. Using this speed signal as a reference, perform time-domain synchronous averaging on the vibration signal to further suppress asynchronous random impact interference and improve the effectiveness of the vibration signal. The expression for the time-domain synchronous averaging algorithm is as follows: in, The averaged vibration signal, The average number of times, The vibration signal acquired in the kth sampling is... The spindle rotation period; According to the above technical solution, the key components include the bearings, saw blades, spindle and housing of the grooving machine, and each key component is equipped with a dedicated monitoring sensor.

[0009] According to the above technical solution, the condition monitoring step also includes saw blade fatigue monitoring: using an acoustic emission sensor to collect high-frequency stress wave signals generated by the saw blade, or using a strain gauge attached to the saw blade substrate to collect dynamic strain signals of the saw blade. When the amplitude or spectral characteristics of the signal exceed a preset threshold, it is determined that the saw blade has fatigue damage or breakage risk.

[0010] According to the above technical solution, in the condition monitoring step, the temperature signal is acquired and determined as follows: the temperature rise rate of the key component per unit time is obtained. In the fatigue judgment step, if the temperature rise rate exceeds a preset rate threshold, even if the current temperature has not reached the absolute temperature threshold, abnormal temperature rise fatigue is still determined. The temperature rise rate is calculated using a linear fitting algorithm, and the expression is: in, For the rate of temperature rise, , These are temperature values ​​collected at different times. , These represent the data collection times for the corresponding temperatures.

[0011] According to the above technical solution, the construction process of the working condition-reference mapping library is as follows: In the initial health stage of the grooving machine, it is controlled to run under multiple working condition combinations with different feed speeds and cutting depths. Fatigue characteristic data corresponding to each working condition combination are collected, and the health reference values ​​of each working condition combination are calculated and stored to complete the construction of the mapping library. At the same time, a reference calibration mechanism is established to periodically calibrate the health reference values ​​in the mapping library to avoid reference distortion and ensure the accuracy of subsequent fatigue judgment. The reference calibration adopts a linear correction algorithm, the expression of which is: in, For the calibrated health baseline value, The health baseline value before calibration. The calibration coefficient is calculated based on real-time data collected from the comparative health equipment.

[0012] According to the above technical solution, the fatigue judgment step also includes early warning and control: when it is determined that the grooving machine is fatigued or malfunctioning, corresponding early warning information is generated.

[0013] According to the above technical solution, the fatigue characteristic data collected in the condition monitoring step simultaneously includes vibration signals, current signals, and temperature signals. The fatigue judgment step employs a multi-feature fusion algorithm to weightedly fuse the features of the three signals to obtain a comprehensive fatigue index. When the comprehensive fatigue index exceeds a preset threshold, the grooving machine is determined to be fatigued or malfunctioning. This method complements single-signal judgment, improving the accuracy of fatigue judgment. The multi-feature fusion algorithm uses a weighted summation algorithm, expressed as: in, The comprehensive fatigue index, , , These are the relative deviations of the vibration signal, current signal, and temperature signal, respectively. , , These are the weighting coefficients, and The weighting is set based on the fatigue failure weight of key components.

[0014] According to the above technical solution, the preset threshold is an adaptive threshold, which is adjusted in segments based on the grooving machine's running time and the component aging stage. It is also determined by combining the duration of data deviation. Only when the deviation exceeds the threshold and persists for a preset duration is fatigue or a fault determined, thus avoiding misjudgments caused by momentary interference. This, combined with the anti-interference processing of the condition monitoring steps, further improves monitoring reliability. The adaptive threshold uses a piecewise linear adjustment algorithm, expressed as: in, The adaptive threshold at time t, As the initial threshold, This is the threshold adjustment coefficient. The cumulative running time of the grooving machine is set according to the fatigue aging characteristics of the components. The value of .

[0015] According to the above technical solution, the data verification step specifically involves: setting a fluctuation threshold for working parameters and a data acquisition threshold for fatigue characteristics. When the fluctuations in the acquired feed rate and cutting depth exceed the fluctuation threshold for working parameters, or when the fatigue characteristic data exceeds the acquisition threshold range, it is determined to be abnormal data and removed. Simultaneously, three consecutive sets of acquired data are compared. If the data deviation exceeds 10%, a second data acquisition verification is performed on that set to ensure data validity. The continuous data deviation is determined using a variance algorithm, expressed as: in, The variance of three consecutive sets of data. Collect data for the i-th group. The variance is the average of the three sets of data. When the variance exceeds the preset variance threshold, the data is determined to be abnormal, and a second data collection and verification is performed.

[0016] This invention provides a method for monitoring the fatigue of a grooving machine. It has the following beneficial effects: (1) The fatigue monitoring method of the grooving machine achieves accurate matching of health benchmark values ​​under different feed speeds and cutting depths by using a preset working condition-benchmark mapping library, thus solving the problem of poor adaptability of a single benchmark. At the same time, it adopts algorithms such as adaptive filtering and time-domain synchronous averaging to effectively remove noise such as vibration and electromagnetic interference in the field environment and extract the real fatigue characteristic signals of key components. In the data verification process, abnormal data is removed by means of variance algorithm, and fatigue judgment is made by quantitative judgment of relative deviation. The accuracy of monitoring data and results is guaranteed in multiple dimensions, avoiding misjudgment and omission caused by subjective judgment and instantaneous interference, and providing reliable support for equipment fatigue monitoring.

[0017] (2) The fatigue monitoring method of the grooving machine can detect abnormal temperature rise fatigue in advance by calculating the temperature rise rate of key components through linear fitting algorithm; the saw blade fatigue monitoring can predict the fatigue damage and breakage risk of saw blade by collecting high-frequency stress wave or dynamic strain signal; at the same time, the adaptive threshold is dynamically adjusted according to the equipment running time and component aging stage, and combined with the deviation duration judgment, it can give timely warning when the equipment has slight fatigue, remind the staff to take maintenance measures, avoid further aggravation of fatigue damage, and effectively extend the service life of key components and the whole machine of the grooving machine.

[0018] (3) The fatigue monitoring method of this grooving machine optimizes the benchmark value and threshold value periodically through moving average and linear correction algorithms, adapting to scenarios such as equipment aging and changes in working conditions, without the need for frequent adjustment of monitoring parameters. The monitoring process is highly automated, forming a closed loop from working condition acquisition and signal processing to fatigue judgment and early warning control, reducing the difficulty of operation for staff and improving monitoring efficiency.

[0019] (4) The fatigue monitoring method of this grooving machine covers key components such as bearings, saw blades, and spindles. It also collects multi-dimensional fatigue characteristic data such as vibration, current, and temperature. The multi-feature fusion is achieved through a weighted summation algorithm to avoid the limitations of single signal monitoring. The data archiving and updating steps realize the dynamic optimization of the benchmark value and threshold, continuously improving the monitoring accuracy. The early warning control function can issue early warnings and take corresponding control strategies according to the severity of fatigue, reducing the probability of equipment failure and downtime. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the detection method of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0022] Please see Figure 1 One embodiment of the present invention is a fatigue monitoring method for a grooving machine, comprising the following steps: S1. Working Condition Acquisition: The feed speed and cutting depth of the grooving machine are acquired in real time as current working condition parameters. The parameters are collected through corresponding sensors to ensure the accuracy of the working condition data. S2. Benchmark Construction and Matching: Preset Working Condition - Benchmark Mapping Library. The mapping library is constructed based on the initial healthy state of the grooving machine, using fatigue characteristic benchmark data collected under working condition combinations of different feed speeds and cutting depths. According to the current working condition parameters, the corresponding healthy benchmark value is matched from the mapping library. S3. Condition Monitoring: Real-time acquisition of fatigue characteristic data of key components of the grooving machine. The fatigue characteristic data includes at least one of vibration, current, and temperature signals. Simultaneously, fatigue signals related to the saw blade can be selectively acquired. The acquired signals undergo anti-interference processing to extract true and effective features. The anti-interference processing employs an adaptive filtering algorithm, the expression of which is: in, This is the filtered output signal. For adaptive filtering weight vectors, The original acquired signal, The filtering error signal is used to iteratively update the weight vector through the minimum mean square error criterion, thereby achieving adaptive suppression of on-site clutter interference. S4. Data Verification: Verify the validity of the real-time collected fatigue characteristic data and current operating parameters, and remove abnormal collected data and abnormal fluctuation data of operating parameters to ensure the reliability of the data used for fatigue judgment. S5. Fatigue Judgment: The verified fatigue characteristic data is compared with the matched health benchmark value. When the data deviation exceeds a preset threshold or meets specific fatigue judgment conditions, the grooving machine is judged to be fatigued or malfunctioning. The data deviation is calculated using a relative deviation algorithm, expressed as: Among them, This is a relative deviation. For fatigue characteristic data collected in real time, For the matched health baseline value, when When the threshold is exceeded, fatigue abnormality is determined.

[0023] S6. Data Archiving and Updates: The operating condition parameters, effective fatigue characteristic data, and fatigue judgment results from each monitoring session are categorized and archived. The health baseline values ​​and preset thresholds of the operating condition-benchmark mapping library are periodically optimized based on the archived data to improve the accuracy of subsequent monitoring. The health baseline value optimization uses a moving average algorithm, expressed as: in, The optimized health baseline value, The number of samples in the archived data. The effective fatigue feature data for the i-th archive.

[0024] To achieve dynamic optimization of benchmark values ​​and thresholds, adapt to scenarios such as equipment aging and changes in operating conditions, continuously improve monitoring accuracy, and extend the adaptation cycle of monitoring methods.

[0025] The specific steps for obtaining operating conditions are as follows: S1.1. The cutting depth is obtained by measuring the cutting depth of the grooving machine's cutting wheel using a displacement sensor or photoelectric sensor; S1.2 Measure the moving speed of the grooving machine relative to the workpiece using an encoder, Hall sensor or inertial measurement unit to obtain the feed speed, ensuring that the collected working condition parameters are accurately matched with the working condition combination in the subsequent working condition-reference mapping library.

[0026] In the condition monitoring process, the methods for acquiring and processing vibration signals include: S2.1 Install the first vibration sensor on the bearing housing of the grooving machine and the second vibration sensor on the outer shell. After collecting the vibration data of both, use the transfer function or adaptive filtering algorithm to remove the resonance interference of the outer shell and extract the real vibration signal of the bearing body. S2.2 Simultaneously acquire the spindle speed signal of the grooving machine. Using this speed signal as a reference, perform time-domain synchronous averaging on the vibration signal to further suppress asynchronous random impact interference and improve the effectiveness of the vibration signal. The expression for the time-domain synchronous averaging algorithm is as follows: in, The averaged vibration signal, The average number of times, The vibration signal acquired in the kth sampling is... The spindle rotation period; Key components include the bearings, saw blades, spindle, and housing of the grooving machine. Each key component is equipped with a dedicated monitoring sensor. By accurately collecting vibration signals from key components such as bearings, various interferences can be effectively eliminated, improving the effectiveness of vibration signals. At the same time, comprehensive monitoring of key components can be achieved to avoid overlooking potential fatigue risks.

[0027] The condition monitoring process also includes specialized monitoring of saw blade fatigue: using acoustic emission sensors to collect high-frequency stress wave signals generated by the saw blade, or using strain gauges attached to the saw blade substrate to collect dynamic strain signals of the saw blade. When the amplitude or spectral characteristics of the signal exceed a preset threshold, it is determined that the saw blade has fatigue damage or breakage risk. By providing early warning of saw blade fatigue damage and breakage risk, equipment failure and safety accidents caused by saw blade failure can be avoided, thus achieving specialized and accurate monitoring of saw blade fatigue.

[0028] In the condition monitoring step, the temperature signal is acquired and determined as follows: the temperature rise rate of key components per unit time is obtained. In the fatigue judgment step, if the temperature rise rate exceeds a preset rate threshold, even if the current temperature has not reached the absolute temperature threshold, abnormal temperature rise fatigue is still determined. The temperature rise rate is calculated using a linear fitting algorithm, and the expression is: in, For the rate of temperature rise, , These are temperature values ​​collected at different times. , These data are collected at corresponding temperature times. By detecting abnormal temperature rise and fatigue in key components in advance, further damage can be prevented, achieving early warning of temperature fatigue and solving the problem of lag in traditional absolute temperature determination.

[0029] The construction process of the working condition-reference mapping library is as follows: In the initial health stage of the grooving machine, it is controlled to run under multiple working condition combinations with different feed speeds and cutting depths. Fatigue characteristic data corresponding to each working condition combination is collected, and the health reference value of each working condition combination is calculated and stored to complete the construction of the mapping library. At the same time, a reference calibration mechanism is established to periodically calibrate the health reference values ​​in the mapping library to avoid reference distortion and ensure the accuracy of subsequent fatigue judgment. The reference calibration adopts a linear correction algorithm, the expression of which is: in, For the calibrated health baseline value, The health baseline value before calibration. The calibration coefficient is calculated based on real-time data collected from comparative health devices to ensure the accuracy and timeliness of the working condition-benchmark mapping library, avoid misjudgment of fatigue due to benchmark distortion, and further improve monitoring reliability.

[0030] The fatigue assessment process also includes early warning and control: when the grooving machine is determined to be fatigued or malfunctioning, corresponding early warning information is generated.

[0031] The fatigue characteristic data collected in the condition monitoring step includes vibration signals, current signals, and temperature signals. The fatigue judgment step uses a multi-feature fusion algorithm to weight and fuse the features of the three signals to obtain a comprehensive fatigue index. When the comprehensive fatigue index exceeds a preset threshold, the grooving machine is judged to have fatigue or a fault. This method complements the single-signal judgment, improving the accuracy of fatigue judgment. The multi-feature fusion algorithm uses a weighted summation algorithm, and the expression is: in, The comprehensive fatigue index, , , These are the relative deviations of the vibration signal, current signal, and temperature signal, respectively. , , These are the weighting coefficients, and Based on the fatigue failure weight setting of key components, and by comprehensively considering multi-dimensional fatigue characteristics, the limitations of single signal monitoring are avoided, the comprehensiveness and accuracy of fatigue judgment are improved, and the probability of misjudgment and missed judgment is reduced.

[0032] The preset threshold is an adaptive threshold, adjusted in segments based on the grooving machine's running time and component aging stage. It also considers the duration of data deviation for judgment; fatigue or malfunction is only determined when the deviation exceeds the threshold and persists for the preset duration, avoiding misjudgments caused by momentary interference. This, combined with the anti-interference processing in the condition monitoring steps, further improves monitoring reliability. The adaptive threshold uses a piecewise linear adjustment algorithm, expressed as: in, The adaptive threshold at time t, As the initial threshold, This is the threshold adjustment coefficient. The cumulative running time of the grooving machine is set according to the fatigue aging characteristics of the components. The value of the threshold is adapted to the fatigue characteristics of the equipment at different operating stages, avoiding the limitations of fixed thresholds, and eliminating misjudgments caused by instantaneous interference, thereby further improving the reliability and adaptability of monitoring.

[0033] The data verification steps are as follows: Set thresholds for fluctuations in operating parameters and for collecting fatigue characteristic data. When the fluctuations in the collected feed rate or cutting depth exceed the threshold for fluctuations in operating parameters, or when the fatigue characteristic data exceeds the range of the collection threshold, it is judged as abnormal data and removed. Simultaneously, compare three consecutive sets of collected data. If the data deviation exceeds 10%, a second data collection and verification is performed on that set to ensure data validity. The variance algorithm is used to assist in judging the deviation of continuous data; the expression is: in, The variance of three consecutive sets of data. Collect data for the i-th group. The variance is the average of the three sets of data. When the variance exceeds the preset variance threshold, the data is considered abnormal, and a second data collection and verification is performed to further filter out various abnormal data. This ensures that the data used for fatigue assessment is accurate and reliable, reduces the risk of misjudgment in fatigue assessment from the source, and guarantees the accuracy of monitoring results.

[0034] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A fatigue monitoring method for a grooving machine, characterized in that: Includes the following steps: S1. Working Condition Acquisition: The feed speed and cutting depth of the grooving machine are acquired in real time as current working condition parameters. The parameters are collected through corresponding sensors to ensure the accuracy of the working condition data. S2. Benchmark Construction and Matching: Preset Working Condition-Benchmark Mapping Library. The mapping library is constructed based on the initial healthy state of the grooving machine, by collecting fatigue characteristic benchmark data under working condition combinations of different feed speeds and cutting depths. According to the current working condition parameters, the corresponding healthy benchmark value is matched from the mapping library. S3. Condition Monitoring: Real-time acquisition of fatigue characteristic data of key components of the grooving machine. The fatigue characteristic data includes at least one of vibration signals, current signals, and temperature signals. Simultaneously, fatigue signals related to the saw blade can be selectively acquired. The acquired signals undergo anti-interference processing to extract true and effective features. The anti-interference processing employs an adaptive filtering algorithm, the expression of which is: in, This is the filtered output signal. For adaptive filtering weight vectors, The original acquired signal, The filtering error signal is used to iteratively update the weight vector through the minimum mean square error criterion, thereby achieving adaptive suppression of on-site clutter interference. S4. Data Verification: Verify the validity of the real-time collected fatigue characteristic data and current operating parameters, and remove abnormal collected data and abnormal fluctuation data of operating parameters to ensure the reliability of the data used for fatigue judgment. S5. Fatigue Judgment: The verified fatigue characteristic data is compared with the matched health benchmark value. When the data deviation exceeds a preset threshold or meets specific fatigue judgment conditions, the grooving machine is judged to be fatigued or malfunctioning. The data deviation is calculated using a relative deviation algorithm, and the expression is: Among them, This is a relative deviation. For fatigue characteristic data collected in real time, For the matched health baseline value, when When the threshold is exceeded, fatigue abnormality is determined. S6. Data Archiving and Update: The operating condition parameters, effective fatigue characteristic data, and fatigue judgment results from each monitoring session are categorized and archived. The health baseline values ​​and preset thresholds of the operating condition-benchmark mapping library are periodically optimized based on the archived data to improve subsequent monitoring accuracy. The health baseline value optimization uses a moving average algorithm, expressed as: in, The optimized health baseline values, The number of samples in the archived data. The effective fatigue feature data for the i-th archive.

2. The fatigue monitoring method for a grooving machine according to claim 1, characterized in that: The specific steps for obtaining the operating conditions are as follows: S1.

1. The cutting depth is obtained by measuring the cutting depth of the grooving machine's cutting wheel using a displacement sensor or photoelectric sensor; S1.2 Measure the moving speed of the grooving machine relative to the workpiece using an encoder, Hall sensor or inertial measurement unit to obtain the feed speed, ensuring that the collected working condition parameters are accurately matched with the working condition combination in the subsequent working condition-reference mapping library.

3. The fatigue monitoring method for a grooving machine according to claim 2, characterized in that: The vibration signal acquisition and processing method in the aforementioned condition monitoring step includes: S2.1 Install the first vibration sensor on the bearing housing of the grooving machine and the second vibration sensor on the outer shell. After collecting the vibration data of both, use the transfer function or adaptive filtering algorithm to remove the resonance interference of the outer shell and extract the real vibration signal of the bearing body. S2.2 Simultaneously acquire the spindle speed signal of the grooving machine. Using this speed signal as a reference, perform time-domain synchronous averaging on the vibration signal to further suppress asynchronous random impact interference and improve the effectiveness of the vibration signal. The expression for the time-domain synchronous averaging algorithm is as follows: in, The averaged vibration signal, The average number of times, The vibration signal acquired in the kth sampling is... The spindle rotation period; The key components include the bearings, saw blades, spindle, and housing of the grooving machine, with dedicated monitoring sensors installed for each key component.

4. The fatigue monitoring method for a grooving machine according to claim 3, characterized in that: The condition monitoring step also includes a special monitoring of saw blade fatigue: using an acoustic emission sensor to collect high-frequency stress wave signals generated by the saw blade, or using a strain gauge attached to the saw blade substrate to collect dynamic strain signals of the saw blade. When the amplitude or spectral characteristics of the signal exceed a preset threshold, it is determined that the saw blade has fatigue damage or breakage risk.

5. The fatigue monitoring method for a grooving machine according to claim 4, characterized in that: In the condition monitoring step, the temperature signal is acquired and determined as follows: the temperature rise rate of key components per unit time is obtained. In the fatigue judgment step, if the temperature rise rate exceeds a preset rate threshold, even if the current temperature has not reached the absolute temperature threshold, abnormal temperature rise fatigue is still determined. The temperature rise rate is calculated using a linear fitting algorithm, and the expression is: in, For the rate of temperature rise, , These are temperature values ​​collected at different times. , These represent the data collection times for the corresponding temperatures.

6. The fatigue monitoring method for a grooving machine according to claim 5, characterized in that: The construction process of the working condition-reference mapping library is as follows: In the initial health stage of the grooving machine, it is controlled to run under multiple working condition combinations with different feed speeds and cutting depths. Fatigue characteristic data corresponding to each working condition combination is collected, and the health reference value of each working condition combination is calculated and stored to complete the construction of the mapping library. At the same time, a reference calibration mechanism is established to periodically calibrate the health reference values ​​in the mapping library to avoid reference distortion and ensure the accuracy of subsequent fatigue judgment. The reference calibration adopts a linear correction algorithm, the expression of which is: in, For the calibrated health baseline value, The health baseline value before calibration. The calibration coefficient is calculated based on real-time data collected from the comparative health equipment.

7. The fatigue monitoring method for a grooving machine according to claim 6, characterized in that: The fatigue judgment step also includes early warning and control: when it is determined that the grooving machine is fatigued or malfunctioning, corresponding early warning information is generated.

8. The fatigue monitoring method for a grooving machine according to claim 7, characterized in that: The fatigue characteristic data collected in the condition monitoring step includes vibration signals, current signals, and temperature signals. The fatigue judgment step employs a multi-feature fusion algorithm to weightedly fuse the features of the three signals to obtain a comprehensive fatigue index. When the comprehensive fatigue index exceeds a preset threshold, the grooving machine is determined to be fatigued or malfunctioning. This method complements single-signal judgment, improving the accuracy of fatigue judgment. The multi-feature fusion algorithm uses a weighted summation algorithm, expressed as: in, The comprehensive fatigue index, , , These are the relative deviations of the vibration signal, current signal, and temperature signal, respectively. , , These are the weighting coefficients, and The weighting is set based on the fatigue failure weight of key components.

9. The fatigue monitoring method for a grooving machine according to claim 8, characterized in that: The preset threshold is an adaptive threshold, adjusted in segments based on the grooving machine's running time and component aging stage. It is also determined by the duration of data deviation; fatigue or a fault is only identified when the deviation exceeds the threshold and persists for a preset duration. This avoids misjudgments caused by momentary interference and, in conjunction with the anti-interference processing in the condition monitoring steps, further improves monitoring reliability. The adaptive threshold uses a piecewise linear adjustment algorithm, expressed as: in, The adaptive threshold at time t. As the initial threshold, This is the threshold adjustment coefficient. The cumulative running time of the grooving machine is set according to the fatigue aging characteristics of the components. The value of .

10. The fatigue monitoring method for a grooving machine according to claim 9, characterized in that: The data verification step specifically involves: setting a threshold for fluctuations in operating parameters and a threshold for collecting fatigue characteristic data. When the fluctuations in the collected feed rate or cutting depth exceed the threshold for fluctuations in operating parameters, or when the fatigue characteristic data exceeds the range of the collection threshold, it is determined to be abnormal data and removed. Simultaneously, three consecutive sets of collected data are compared. If the data deviation exceeds 10%, a second data collection and verification is performed on that set to ensure data validity. The continuous data deviation is determined using a variance algorithm, expressed as: in, The variance of three consecutive sets of data. Collect data for the i-th group. The variance is the average of the three sets of data. When the variance exceeds the preset variance threshold, the data is determined to be abnormal, and a second data collection and verification is performed.

Images