A method and system for detecting bearing wear in a skein processing shaft

Abstract

The application discloses a kind of hank processing bearing wear detection method and system, it is related to hank processing bearing wear detection field, for sensitive determination bearing health condition, including: the transmission system of hank machine injects with the detection torque fluctuation of specific frequency or broadband characteristics;Synchronous acquisition torque response signal of drive motor in transmission system, and the original instruction signal of detection torque fluctuation;Torque response signal is preprocessed, and the torque response signal after preprocessing is obtained;Preprocessing includes eliminating random spike pulse interference caused by uneven yarn thickness;The relationship between the torque response signal after preprocessing and original instruction signal is analyzed to identify the nonlinear characteristics caused by bearing wear;Based on the long-term trend change of nonlinear characteristics, the bearing wear state is judged.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of detection of bearing wear in the process of skein processing, and in particular to a method and system for detecting bearing wear in the process of skein processing. BACKGROUND

[0002] In the process of skein processing, the bearings inside the skein machine will inevitably wear out. When the raceway or rolling element of the bearing appears local wear, causing uneven load distribution, it will generate additional periodic frictional force in the rotation process, thereby causing periodic fluctuations in the output torque of the driving motor. In theory, by analyzing these torque fluctuations, early wear can be identified before the bearing fails seriously.

[0003] However, the yarn produced has uneven thickness, for example, there are some "yarn defects" of thick or thin sections on the yarn. When a thicker yarn defect passes through the guide and is wound, it will instantaneously increase the tension of the yarn, thereby forming a short and large-amplitude spike pulse on the torque signal.

[0004] In order to obtain better winding forming effect and prevent the yarn from repeatedly superimposing at the same position of the winding package to form an undesirable structure called "stacking", a "stacking prevention" technology is generally used. For example, the system can cause the rotating speed of the winding shaft to change back and forth within a small range at a specific frequency. This will also cause periodic fluctuations in the output torque of the driving motor.

[0005] Based on the above two kinds of interference, how to accurately extract the weak characteristic signal representing the early uneven wear of the bearing is a technical problem to be solved. SUMMARY

[0006] The present application provides a method for detecting bearing wear in the process of skein processing, which is used to sensitively determine the health status of the bearing.

[0007] In the first aspect, in order to solve the above technical problem, the present application provides a method for detecting bearing wear in the process of skein processing, comprising: injecting a detection torque fluctuation with a specific frequency or a wideband characteristic into the transmission system of the skein machine; synchronously collecting a torque response signal of the driving motor in the transmission system and an original command signal of the detection torque fluctuation; preprocessing the torque response signal to obtain a preprocessed torque response signal; the preprocessing includes eliminating random spike pulse interference caused by uneven yarn thickness; analyzing the relationship between the preprocessed torque response signal and the original command signal to identify nonlinear features caused by bearing wear; and judging the bearing wear state based on the long-term trend change of the nonlinear features.

[0008] Optionally, the relationship between the pre-processed torque response signal and the original command signal is analyzed to identify the non-linear feature caused by bearing wear, including: obtaining an internal execution signal of the drive system; determining a drive system execution deviation based on the original command signal for detecting torque fluctuation and the internal execution signal of the drive system; compensating the pre-processed torque response signal based on the drive system execution deviation to obtain a compensated torque response signal; and analyzing the relationship between the compensated torque response signal and the original command signal to identify the non-linear feature caused by bearing wear.

[0009] Optionally, the relationship between the pre-processed torque response signal and the original command signal is analyzed to identify the non-linear feature caused by bearing wear, including: obtaining an internal execution signal of the drive system; determining a drive system execution deviation based on the original command signal for detecting torque fluctuation and the internal execution signal of the drive system; compensating the pre-processed torque response signal based on the drive system execution deviation to obtain a compensated torque response signal; and analyzing the relationship between the compensated torque response signal and the original command signal to identify the non-linear feature caused by bearing wear.

[0010] Optionally, the bearing wear state is determined based on the long-term trend change of the non-linear feature, including: monitoring operating condition parameters of the hank winding machine; identifying a working condition change event based on the operating condition parameters; re-establishing a reference range of the non-linear feature when the working condition change event is identified; and determining the bearing wear state based on the long-term trend change of the non-linear feature according to the re-established reference range.

[0011] Optionally, the torque response signal is pre-processed to obtain a pre-processed torque response signal, including: monitoring operating condition parameters of the hank winding machine, the operating condition parameters including yarn type, processing speed, or environmental humidity; analyzing statistical characteristics of sharp pulse in the torque response signal based on the operating condition parameters; the statistical characteristics including pulse amplitude distribution, duration distribution, or frequency energy distribution; adaptively adjusting threshold parameters for identifying random sharp pulse based on the statistical characteristics of the sharp pulse; and identifying and reconstructing the sharp pulse in the torque response signal using the adjusted threshold parameters to eliminate random sharp pulse interference caused by uneven yarn thickness to obtain the pre-processed torque response signal.

[0012] Optionally, a detection torque fluctuation with specific frequency or broadband characteristics is injected into the transmission system of the hank winding machine, including: adjusting parameters of the anti-overlap control function of the hank winding machine; the adjusted parameters are used to make the anti-overlap control function generate torque fluctuation with specific frequency or broadband characteristics; and the torque fluctuation is used as the detection torque fluctuation, and the detection torque fluctuation is injected into the transmission system.

[0013] Optionally, the torque response signal of the driving motor in the transmission system is collected, including: monitoring the current signal and the voltage signal of the driving motor; and estimating the torque response signal of the driving motor according to the current signal and the voltage signal of the driving motor and parameters of the driving motor.

[0014] Optionally, the torque response signal of the driving motor in the transmission system is collected synchronously, and the original instruction signal of the torque fluctuation is detected, including: providing a unified clock source, the unified clock source being used for clock synchronization of a system for collecting the torque response signal of the driving motor in the transmission system and a system for generating the original instruction signal of the torque fluctuation; and collecting the torque response signal of the driving motor in the transmission system synchronously and detecting the original instruction signal of the torque fluctuation based on the unified clock source.

[0015] Optionally, the torque response signal is subjected to spike pulse identification and reconstruction, including: extracting time domain features and frequency domain features of candidate spike pulses from the torque response signal, the time domain features including pulse amplitude, duration or slope, and the frequency domain features including energy of a specific frequency band; determining a discrimination threshold or rule for distinguishing random spike pulses, bearing wear characteristic signals and machine normal operation transient fluctuations according to the time domain features, the frequency domain features and operation condition parameters of the twisting machine; identifying random spike pulses in the torque response signal according to the discrimination threshold or rule; and reconstructing the identified random spike pulses, the signal reconstruction including interpolation or local smoothing processing based on signals before and after the spike pulses, so that the torque response signal is subjected to spike pulse identification and reconstruction.

[0016] In a second aspect, the present application provides a bearing wear detection system for a twisting machine, the system comprising:

[0017] a detection torque injection module configured to inject a detection torque fluctuation with a specific frequency or a wideband characteristic into a transmission system of the twisting machine;

[0018] a signal collection module configured to synchronously collect a torque response signal of a driving motor in the transmission system and an original instruction signal of the detection torque fluctuation;

[0019] a signal preprocessing module configured to preprocess the torque response signal to obtain a preprocessed torque response signal, the preprocessing including elimination of random spike pulse interference caused by uneven thickness of the yarn;

[0020] a feature identification module configured to analyze a relationship between the preprocessed torque response signal and the original instruction signal to identify a non-linear feature caused by bearing wear;

[0021] a state judgment module configured to judge a bearing wear state based on a long-term trend change of the non-linear feature.

[0022] Compared with the prior art, the present application has the following beneficial effects:

[0023] The present application provides a hank processing bearing wear detection method, aiming to effectively solve the complex interference problem in the bearing wear detection of the hank machine. By injecting a detection torque fluctuation with a specific frequency or a wideband characteristic into the transmission system of the hank machine, and synchronously collecting the torque response signal of the driving motor and the original instruction signal of the detection torque fluctuation, a reliable data basis is provided for subsequent signal analysis. Especially important is that this method can effectively eliminate the random spike pulse interference caused by uneven yarn thickness in the preprocessing stage, significantly reducing the false positive rate and overcoming the problem of inaccurate detection caused by yarn defect interference in the prior art. In addition, by analyzing the relationship between the preprocessed torque response signal and the original instruction signal, this method can identify the nonlinear characteristics caused by bearing wear. Unlike the linear characteristics that are easily confused with the periodic fluctuations introduced by the anti-piling technology, the nonlinear response caused by bearing wear has its unique signal fingerprint, enabling this method to effectively distinguish and accurately extract the weak signal representing the early uneven wear of the bearing, avoiding false judgments caused by anti-piling technology interference. Finally, based on the long-term trend change of this nonlinear characteristic, the bearing wear state is judged, further improving the accuracy and reliability of the detection, realizing early warning and predictive maintenance of the bearing state, thereby effectively prolonging the equipment life, reducing maintenance costs, and improving production efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 is a hank processing bearing wear detection method flowchart provided by an embodiment of the present application;

[0025] Figure 2 is another hank processing bearing wear detection method flowchart provided by an embodiment of the present application;

[0026] Figure 3 is a hank processing bearing wear detection system structure diagram provided by an embodiment of the present application. DETAILED DESCRIPTION

[0027] The technical solutions in the present application will be described in detail below with reference to the drawings in the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all embodiments. The components of the present application described and shown in the drawings herein can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed application, but only represents selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0028] It should be noted that similar reference numerals and letters refer to like items throughout the accompanying drawings, and once an item is defined in one drawing, it is not necessary to further define and explain it in subsequent drawings. Meanwhile, in the description of the present application, the terms "first", "second", and the like are only used to distinguish description, and cannot be understood as indicating or implying relative importance.

[0029] The hank processing bearing wear detection method provided by the embodiment of the present application will be described and explained in detail below through the following specific examples.

[0030] With reference to Figure 1 The present application provides a hank processing bearing wear detection method, comprising the following steps:

[0031] S1, injecting a detection torque fluctuation with specific frequency or wideband characteristics into the transmission system of the hank machine.

[0032] As a possible implementation manner, the system can generate the detection torque fluctuation by superimposing a preset periodic signal in the control loop of the driving motor.

[0033] Alternatively, a wideband signal containing multiple frequency components can also be injected, for example, by superimposing multiple sine waves of different frequencies, or using a pseudo-random sequence signal, to more comprehensively stimulate the dynamic response of the transmission system.

[0034] For example, a sine wave signal can be added to the speed command or current command of the driving motor, so that the torque output by the motor fluctuates accordingly. The frequency of the sine wave signal can be set to a single specific frequency, such as 50Hz or 100Hz, to produce a controllable excitation in the transmission system. These detection torque fluctuations are designed to be able to penetrate the transmission system and interact with the potential wear characteristics of the bearing.

[0035] As another possible implementation manner, the system can adjust the parameters of the anti-winding control function of the hank machine; the adjusted parameters are used to make the anti-winding control function produce torque fluctuations with specific frequency or wideband characteristics; and the torque fluctuations are taken as detection torque fluctuations, and the detection torque fluctuations are injected into the transmission system.

[0036] The anti-piling control function of the hank winding machine is generally used to ensure that the yarn is evenly wound on the bobbin, avoiding the accumulation of yarn in a certain area to form "piled yarn". This function achieves the uniform distribution of the yarn by periodically fine-tuning the winding speed or tension. The present application takes advantage of the feature that the anti-piling control function will introduce small and controllable torque fluctuations when it is working normally. By fine-tuning the parameters of the anti-piling control function, such as adjusting its oscillation frequency, amplitude or waveform, the torque fluctuations it produces can have a pre-set specific frequency or broadband characteristic. These adjusted and controlled torque fluctuations are then directly used as the detection torque fluctuations and naturally injected into the transmission system of the hank winding machine.

[0037] The specific frequency or broadband characteristic can be selected according to the detection needs of bearing wear, for example, a frequency range related to the bearing fault characteristic frequency can be selected, or a wider frequency range can be selected to cover multiple potential failure modes.

[0038] In some preferred embodiments, the following is described by a specific example. Assume that the anti-piling control function of the hank winding machine evenly distributes the yarn by periodically changing the winding motor speed. In order to inject detection torque fluctuations with a specific frequency, the oscillation frequency of the anti-piling control can be set to, for example, 50 Hz, and its amplitude can be adjusted so that it produces detectable 50 Hz torque fluctuations in the transmission system without affecting normal production. The 50 Hz torque fluctuations are injected into the transmission system as detection torque fluctuations. Alternatively, in order to inject broadband detection torque fluctuations, the anti-piling control parameters can be designed to change the oscillation frequency and amplitude in a random or pseudo-random manner within a certain period of time, thereby generating a wideband torque fluctuation covering a range of, for example, 10 Hz to 200 Hz. These adjusted parameters are configured through the control system of the hank winding machine, so that the anti-piling control function performs its normal work while also generating and injecting detection torque fluctuations.

[0039] S2, synchronously acquiring a torque response signal of the drive motor in the transmission system and an original instruction signal of the detection torque fluctuations.

[0040] As a possible implementation, the torque response signal of the drive motor can be directly measured by a torque sensor installed on the output shaft of the motor. At the same time, the original instruction signal for generating the detection torque fluctuations, i.e. the pre-set signal superimposed on the motor control instruction, can be directly obtained from the motor controller.

[0041] To ensure the time alignment of the signals, a separate, high-precision clock source can be used to synchronize the data acquisition system of the torque sensor and the instruction generation system of the motor controller. For example, a GPS time module or a high-precision crystal oscillator can be used to provide a unified timestamp, ensuring that the two signals correspond accurately on the time axis, thereby laying the foundation for subsequent signal analysis.

[0042] As yet another possible implementation, the system can monitor a current signal and a voltage signal of the drive motor; and estimate a torque response signal of the drive motor according to the current signal and the voltage signal of the drive motor and parameters of the drive motor.

[0043] For example, the system can calculate the electromagnetic torque generated by the drive motor based on its equivalent circuit model or flux linkage model, in combination with the real-time acquired current signal and voltage signal, and pre-calibrated or known drive motor parameters (e.g., stator resistance, inductance, rotor resistance, flux linkage constant, number of pole pairs, etc.).

[0044] As yet another possible implementation, the system can provide a unified clock source, and synchronize the acquisition of the torque response signal of the drive motor in the transmission system and the acquisition of the original instruction signal for detecting torque fluctuations based on the unified clock source.

[0045] The unified clock source is used to synchronize the clock of the system that acquires the torque response signal of the drive motor in the transmission system and the system that generates the original instruction signal for detecting torque fluctuations.

[0046] S3, pre-processing the torque response signal to obtain a pre-processed torque response signal.

[0047] The pre-processing includes eliminating random spike pulse interference caused by uneven yarn thickness.

[0048] As a possible implementation, the system can use signal filtering techniques. For example, an adaptive median filter or a threshold filter based on statistical characteristics can be designed. The filter identifies spikes that have an amplitude far beyond the normal range and a very short duration by analyzing the local statistical characteristics of the signal, such as the mean and variance. Once a spike is identified, it can be reconstructed by interpolating the values of the normal signal points before and after it, or using a local smoothing algorithm, to remove its impact and obtain the pre-processed torque response signal.

[0049] For example, when a sampling point is detected to have a value that deviates significantly from the average value of its neighboring points, it can be judged as a spike, and the average value of the previous and next sampling points can be used to replace the value of the point to achieve signal smoothing.

[0050] A median filter is a non-linear digital filter widely used in image processing and signal processing to remove salt and pepper noise (i.e., sharp spikes). Its specific algorithmic principles are as follows:

[0051] 1. First, determine a sliding window of odd size (for example, window size 3, 5, 7, etc.). The choice of window size will affect the smoothing effect, the larger the window, the more obvious the smoothing effect, but it may lead to loss of signal details.

[0052] 2. Slide the window along the torque response signal sequence from beginning to end.

[0053] 3. At each window position, arrange all data points contained in the window in ascending or descending order. Then, select the middle value (i.e., median) of the arranged data sequence as the new value of the center point of the current window.

[0054] 4. Replace the value of the corresponding window center point in the original signal with the calculated median.

[0055] For example, suppose we have a torque response signal sequence that contains a random spike caused by uneven yarn thickness. The original signal sequence is: [10.2, 10.5, 10.3, 98.7, 10.8, 10.6, 10.4].

[0056] Among them, 98.7 is a clear spike.

[0057] We choose a median filter with a window size of 3 to process. To handle the edges, we usually perform padding or special processing, here we simplify the processing and only focus on the area completely covered by the window.

[0058] Process the first completely covered window: [10.2, 10.5, 10.3]. Replace 10.5 (the second point) in the original sequence with 10.3.

[0059] Process the second window: [10.5, 10.3, 98.7] (here 10.5 is the original value, to demonstrate, we assume it has not been replaced by the previous step, or we start sliding from the second point of the original sequence). Replace 10.3 (the third point) in the original sequence with 10.5.

[0060] Process the window containing the spike: [10.3, 98.7, 10.8]` (here 10.3 is the original value). Replace 98.7 (the spike point) in the original sequence with 10.8.

[0061] Process the next window: [98.7, 10.8, 10.6] (here 98.7 is the original value). The 10.8 (the fifth point) in the original sequence is replaced by 10.8.

[0062] After the median filtering process, the spike value 98.7 in the original sequence is replaced by its adjacent normal value 10.8, effectively eliminating the spike pulse interference, and the surrounding normal signal value is also properly smoothed.

[0063] S4, analyze the relationship between the pretreated torque response signal and the original instruction signal to identify the nonlinear characteristics caused by bearing wear.

[0064] As a possible implementation, the system can identify the nonlinear characteristics caused by bearing wear by comparing the spectral characteristics between the pretreated torque response signal and the original instruction signal.

[0065] For example, the system can perform Fourier transform on the pretreated torque response signal to obtain its frequency domain representation. At the same time, the frequency spectrum of the original instruction signal usually only contains the fundamental frequency component. When the bearing is worn, due to its internal nonlinear dynamic characteristics, new frequency components will be generated in the frequency spectrum of the torque response signal, such as the multiple of the fundamental frequency (harmonic) or the fractional multiple of the fundamental frequency (subharmonic), or even the modulation frequency. By observing whether there are these frequency components in the frequency spectrum of the pretreated torque response signal that do not exist in the frequency spectrum of the original instruction signal or have abnormally increased energy, it can be preliminarily judged whether there are nonlinear characteristics caused by bearing wear. For example, if the original instruction signal is a single frequency.

[0066] The frequency spectrum of the original instruction signal usually only contains the fundamental frequency component refers to the ideal state, the instruction signal without interference.

[0067] In an example, the frequency corresponding to the frequency spectrum of the original instruction signal is the original frequency, and the energy is the original energy. In the case that there is a frequency greater than the preset multiple of the original frequency in the frequency spectrum of the pretreated torque response signal, it can be preliminarily judged whether there are nonlinear characteristics caused by bearing wear. Or, in the case that there is a frequency with energy greater than the preset multiple of the original energy in the frequency spectrum of the pretreated torque response signal, it can be preliminarily judged whether there are nonlinear characteristics caused by bearing wear.

[0068] The preset multiple can be 1.2, 1.3, 1.5, etc.

[0069] As yet another possible implementation, the system can identify the nonlinear characteristics caused by bearing wear according to the following steps:

[0070] S41, obtain an internal execution signal of the drive system.

[0071] The drive system internal execution signal can include actual current, voltage, rotation speed, position feedback signal of the drive motor, or torque instruction, current instruction inside the drive controller. These signals can represent the actual performance of the internal execution link of the drive system after receiving the original instruction signal.

[0072] S42, determining the drive system execution deviation according to the original instruction signal for detecting the torque fluctuation and the drive system internal execution signal.

[0073] The drive system execution deviation determined according to the original instruction signal for detecting the torque fluctuation and the drive system internal execution signal can be understood as quantifying the inconsistency or error of the drive system in executing the instruction.

[0074] As a possible implementation manner, the system can compare the original instruction signal with the drive system internal execution signal, calculate the difference value, ratio, or residual error analysis by establishing a drive system model, and take the difference value, ratio, or residual error analysis result by establishing a drive system model as the drive system execution deviation.

[0075] S43, compensating the pretreated torque response signal based on the drive system execution deviation to obtain a compensated torque response signal.

[0076] As a possible implementation manner, the system can subtract the execution deviation signal from the torque response signal, or eliminate or weaken the influence of the execution deviation on the torque response signal through adaptive filtering, model predictive control, etc., to obtain the compensated torque response signal.

[0077] In this way, the nonlinear or dynamic response of the drive system itself can be stripped from the torque response signal, so that the remaining signal more purely reflects the real state of the bearing in the transmission chain.

[0078] S44, analyzing the relationship between the compensated torque response signal and the original instruction signal to identify the nonlinear characteristics caused by bearing wear.

[0079] In some preferred embodiments, the following is described by a specific example. It is assumed that the transmission system of the hank winding machine is driven by a servo motor, and the servo motor is connected to the hank winding shaft through a speed reducer. In order to identify the nonlinear characteristics caused by bearing wear, a detection torque fluctuation of a specific frequency (for example, 50 Hz) is injected into the transmission system. The torque response signal of the drive motor and the original instruction signal for detecting the torque fluctuation are collected synchronously. In the pretreatment step, random spike pulse interference caused by uneven yarn thickness is eliminated.

[0080] To further improve the recognition accuracy, the execution deviation of the drive system itself needs to be considered. Specifically, the actual current signal of the drive motor can be acquired as the internal execution signal of the drive system. Since the torque of the motor is approximately proportional to the current, the actual current signal can indirectly reflect the actual torque output of the drive system. Then, the original command signal (e.g., an ideal sinusoidal torque command) to detect the torque fluctuation is compared with the actual torque of the drive motor estimated according to the actual current signal, and the instantaneous deviation between the two is calculated. For example, the difference between the original command torque and the actual estimated torque can be calculated as the execution deviation of the drive system.

[0081] Next, based on the execution deviation of the drive system, the pre-processed torque response signal is compensated. For example, if the execution deviation of the drive system shows that the actual torque output of the motor at a certain moment is lower than the command torque, the deviation amount can be added back to the pre-processed torque response signal to correct the "underdrive" effect of the drive system itself. Conversely, the same applies. In this way, the compensated torque response signal will more accurately reflect the real torque fluctuation of the bearing in the transmission chain, excluding the influence of the nonlinear or dynamic response of the drive system itself.

[0082] Finally, the relationship between the compensated torque response signal and the original command signal is analyzed, such as through frequency domain analysis, to identify the harmonic frequency components caused by bearing wear. Since the execution deviation of the drive system has been effectively compensated, these harmonic components will more clearly and accurately indicate the wear state of the bearing.

[0083] The indication of the wear state of the bearing includes:

[0084] 1. First, the torque response signal after compensation processing is converted into the frequency domain, usually using methods such as Fast Fourier Transform (FFT), to obtain the frequency spectrum of the signal. The compensation processing has effectively eliminated the execution deviation of the drive system itself before this, making the nonlinear characteristics caused by bearing wear more prominent in the frequency spectrum.

[0085] In the frequency spectrum, harmonic frequency components caused by bearing wear need to be identified. These harmonic frequency components are multiple frequency components of the fundamental frequency component in the original command signal to detect the torque fluctuation. For example, if the injected fundamental frequency of the detection torque fluctuation is f, then the energy or amplitude at the frequency points of 2f, 3f, 4f, etc. in the frequency spectrum needs to be concerned. In an ideal linear system, the energy of these multiple frequency components should be very low or non-existent. The nonlinearity caused by bearing wear will significantly increase the energy of these harmonic components.

[0086] 2、After identifying the harmonic frequency components, their intensities need to be quantified. The specific method is to calculate the ratio of the energy (or amplitude) of each harmonic frequency component to that of the corresponding fundamental frequency component in the original command signal. This ratio can be used as an indicator of the degree of nonlinearity of the system. For example, the ratio of the energy of the second harmonic (2f) to that of the fundamental frequency (f), the ratio of the energy of the third harmonic (3f) to that of the fundamental frequency (f), etc. can be calculated. Through such ratio calculation, the influence of factors such as fluctuation of the injected torque amplitude on the absolute energy value can be eliminated, making the wear indication more stable and comparable.

[0087] 3、When the bearing is in a healthy state, a baseline range of these harmonic energy ratios can be established. When the bearing begins to wear, its nonlinearity will gradually increase, causing the harmonic energy ratios to begin to rise and possibly exceed the healthy baseline range.

[0088] Bearing wear is a gradual process, so relying solely on instantaneous ratios may not be sufficient. The key is to monitor the long-term trend of these harmonic energy ratios. If these ratios continue to rise and show a clear upward trend, it indicates that bearing wear is occurring and may intensify. By setting different thresholds, the wear state can be divided into early wear, moderate wear, and severe wear stages.

[0089] In one example, assume that a probe torque fluctuation with a fundamental frequency of 20 Hz is injected into the drive system of a spinning machine.

[0090] When the bearing is healthy, the frequency spectrum of the compensated torque response signal mainly has a significant energy peak at 20 Hz. At harmonic frequencies such as 40 Hz (second harmonic), 60 Hz (third harmonic), etc., the energy is very low, for example, the ratio of 40 Hz harmonic energy to 20 Hz fundamental energy may be less than 0.01.

[0091] Early wear: When the bearing begins to show slight wear, the drive system will produce slight nonlinearity. At this time, in the frequency spectrum of the compensated torque response signal, the energy at harmonic frequencies such as 40 Hz and 60 Hz will increase slightly, and the ratio of their energy to the fundamental energy may rise to 0.02-0.05. Although these values may still be within an acceptable range, the upward trend indicates the beginning of wear.

[0092] Moderate wear: As wear intensifies, nonlinearity becomes more pronounced. The energy of the 40 Hz and 60 Hz harmonics will increase significantly, and their ratio may reach 0.05-0.15. At this point, the system may issue a warning, suggesting further inspection or planned maintenance.

[0093] Severe wear: When the bearing is severely worn, the nonlinear effect is very strong. Not only the 40 Hz and 60 Hz harmonic energy is very high in the frequency spectrum, but even higher order harmonics (e.g. 80 Hz) or more complex modulation frequencies can appear. The harmonic energy ratio can exceed 0.15 or even higher. At this time, the system will issue an urgent alarm indicating that the bearing is on the verge of severe failure and needs to be shut down for immediate maintenance to avoid equipment damage or production accidents.

[0094] By the above technical solutions, the application can effectively eliminate or significantly reduce the interference of the driving system's own execution deviation on the bearing wear characteristic identification. This makes the nonlinear features extracted from the torque response signal more accurately reflect the real wear state of the bearing, avoiding misjudgment caused by the inherent characteristics of the driving system. Thus, the detection method of the application has higher robustness and reliability in complex industrial environments, especially in the hank winding machine where the dynamic characteristics of the driving system are significant, so as to more accurately evaluate the bearing wear degree, provide a more reliable basis for predictive maintenance of the equipment, effectively prolong the service life of the equipment and reduce the maintenance cost.

[0095] S5, judging the bearing wear state based on the long-term trend change of the nonlinear feature.

[0096] It should be noted that the identified nonlinear features, such as the energy or amplitude of specific harmonic components, do not reach a high level at the initial stage of bearing wear. On the contrary, they gradually increase as the bearing wear intensifies.

[0097] As a possible implementation, the system can establish a historical database to record the values of these nonlinear features at different time points. By performing trend analysis on these values, such as calculating their rate of change or cumulative change over time, the bearing wear state can be judged.

[0098] For example, an initial reference range can be set, and when the value of the nonlinear feature continuously exceeds this range and shows an upward trend, it can be judged that the bearing is being worn, and the severity of the wear can be evaluated according to the growth rate. For example, if the energy of a certain harmonic component continuously rises for several weeks, it can be considered that the bearing wear is deteriorating.

[0099] It can be understood that this identification method based on nonlinear features can effectively distinguish between bearing wear and linear periodic fluctuations introduced by the anti-winding technology, as the latter usually does not cause significant nonlinear response. Finally, by judging the long-term trend change of the nonlinear feature, this method avoids misjudgment of the instantaneous signal, improving the robustness and reliability of the detection.

[0100] In a possible design, as Figure 2As shown, in order to analyze the relationship between the compensated torque response signal and the original instruction signal to identify the nonlinear characteristics caused by bearing wear, the application can further include the following steps:

[0101] S101, frequency domain conversion is performed on the compensated torque response signal to obtain the frequency spectrum of the compensated torque response signal.

[0102] As a possible implementation, the system can use digital signal processing techniques such as Fast Fourier Transform (FFT) to perform frequency domain conversion on the compensated torque response signal to obtain the frequency spectrum of the compensated torque response signal.

[0103] In this way, the time domain signal is converted into a frequency domain representation, thereby revealing various frequency components contained in the signal and their corresponding energy or amplitude information. This conversion makes it possible to analyze the periodicity, harmonic characteristics, etc. of the signal.

[0104] S102, according to the frequency spectrum of the compensated torque response signal, identify the harmonic frequency components caused by bearing wear in the frequency spectrum.

[0105] Among them, the harmonic frequency component is a multiple of the fundamental frequency component in the original instruction signal.

[0106] As a possible implementation, the system can look for frequency components in the frequency domain that have a multiple relationship with the fundamental frequency component of the original instruction signal, and take the frequency components with the multiple relationship as the harmonic frequency components caused by bearing wear.

[0107] For example, if the fundamental frequency of the original instruction signal is f, the energy changes at frequencies 2f, 3f, 4f, etc. need to be focused on. The appearance and energy changes of these harmonic frequency components are key indicators for judging the degree of nonlinearity of the system.

[0108] It should be noted that bearing wear will usually cause the system to produce a nonlinear response, and this nonlinear response is manifested in the frequency domain as an integer multiple of the input signal fundamental frequency harmonic component.

[0109] S103, according to the energy of the harmonic frequency component and the energy of the corresponding fundamental frequency component in the original instruction signal, calculate the ratio of the energy of the harmonic frequency component to the energy of the corresponding fundamental frequency component in the original instruction signal.

[0110] Among them, the energy of the harmonic frequency component and the energy of the corresponding fundamental frequency component in the original instruction signal are used to quantify the strength of the nonlinear characteristics.

[0111] S104, according to the ratio, judge the nonlinear characteristics caused by bearing wear.

[0112] As a possible implementation, the system can consider that there is a non-linear feature caused by bearing wear when the ratio is greater than or equal to a preset threshold. It is considered that there is no non-linear feature caused by bearing wear when the ratio is less than the preset threshold.

[0113] In some preferred embodiments, the following is described by a specific example. Assume that the probe torque fluctuation injected into the transmission system of the hank machine has a fundamental frequency of 10 Hz. In the preprocessed and compensated torque response signal, the frequency spectrum can be obtained by frequency domain conversion (for example, using FFT). If the bearing is worn, the energy of the 20 Hz, 30 Hz, 40 Hz, etc. multiple frequency components in the frequency spectrum may be observed to increase significantly. For example, the ratio of the energy of the 20 Hz harmonic component to the energy of the 10 Hz fundamental frequency component can be calculated. If the ratio is usually below a certain threshold (for example, 0.05) under normal operating conditions, and when the bearing starts to wear, the ratio gradually rises and exceeds the threshold (for example, reaches 0.1 or higher), it can be judged that there is a non-linear feature caused by bearing wear. By continuously monitoring the trend of these harmonic energy ratios, the severity and development speed of wear can be further evaluated.

[0114] Through the above technical solutions, the present application can provide a more accurate and quantitative method to identify the non-linear feature caused by the wear of the hank processing bearing. Compared with only general relationship analysis, the introduction of frequency domain conversion, harmonic identification and energy ratio calculation enables the non-linear feature of bearing wear to be accurately quantified and distinguished, significantly improving the sensitivity and accuracy of detection. This quantitative indicator provides a reliable basis for subsequent wear state judgment, helps to avoid misjudgment, and can discover potential wear problems earlier, thereby realizing more effective predictive maintenance.

[0115] In a possible design, in order to judge the bearing wear state based on the long-term trend change of the non-linear feature, the present application further includes the following steps:

[0116] S201, monitoring the operating condition parameters of the hank machine.

[0117] The operating condition parameters can be yarn type, processing speed, ambient temperature, ambient humidity, batch information, etc.

[0118] As a possible implementation, the system can monitor the operating condition parameters of the hank machine through the sensors of the hank machine itself, or through manual input or external environment monitoring equipment.

[0119] S202, identifying a working condition change event according to the operating condition parameters.

[0120] As a possible implementation, the system can determine whether the yarn type changes, whether the environmental humidity fluctuation value is greater than or equal to a fluctuation threshold, whether the processing speed gear changes, and determine that a working condition change event is identified in the case that the yarn type changes, or the environmental humidity fluctuation value is greater than or equal to the fluctuation threshold, or the processing speed gear changes (for example, which can include low speed, high speed).

[0121] In the case that the yarn type does not change, and the environmental humidity fluctuation value is less than the fluctuation threshold, and the processing speed gear does not change, it is determined that no working condition change event is identified.

[0122] In an example, the system can identify a working condition change event when the yarn type changes from A to B, or the processing speed switches from low speed to high speed, or the environmental humidity is greater than or equal to the fluctuation threshold.

[0123] S203, when a working condition change event is identified, re-establishing the reference range of the non-linear feature.

[0124] Wherein, when a working condition change event is identified, re-establishing the reference range of the non-linear feature means that under the new working condition, the system will re-collect the torque response signal for a period of time, and re-calculate and determine the normal fluctuation range or reference value of the non-linear feature (such as the harmonic energy ratio) when the bearing is in a normal state under the new working condition according to the above method (for example, by frequency domain conversion, identifying harmonic frequency components, calculating energy ratio, etc.). The re-establishment of this reference range is dynamic and can be based on statistical methods (such as mean, standard deviation, percentile, etc.) or machine learning models. The purpose is to ensure that there is an accurate reference point for judging bearing wear under different working conditions.

[0125] S204, judging the long-term trend change of the non-linear feature based on the re-established reference range to determine the bearing wear state.

[0126] As a possible implementation, after the new reference range is determined, the subsequent collected non-linear feature values will be compared with the new reference range to determine the bearing wear state.

[0127] By continuously monitoring the degree of deviation and change trend of the non-linear feature value relative to the new reference range, the bearing wear can be more accurately judged whether it is happening or getting worse. For example, if the non-linear feature value continuously exceeds the new reference range and shows an upward trend, it indicates that the bearing wear may be deteriorating.

[0128] As a specific embodiment, the following is described through a specific example. Assume that a hank machine processes cotton yarn in the morning and switches to polyester yarn in the afternoon. During the processing of cotton yarn, the system monitors that the yarn type is "cotton yarn" and establishes a reference range A of the nonlinear feature according to the torque response signal at this time. When switching to polyester yarn in the afternoon, the system identifies that the operating condition parameter "yarn type" has changed, that is, identifies the operating condition change event. At this time, the system will not continue to use the reference range A, but will immediately re-collect data and establish a new reference range B of the nonlinear feature in the initial stage of the processing of polyester yarn. Thereafter, the judgment of the bearing wear state is based on the reference range B. For example, if the normal range of the harmonic energy ratio under the reference range A is 0.01-0.05, and due to the different material properties, the newly established reference range B under the polyester yarn working condition may be 0.02-0.08. In this way, even under different working conditions, the system can accurately distinguish between normal fluctuations and changes in nonlinear features caused by actual wear, thereby providing more reliable wear diagnosis.

[0129] Through the above technical solution, the robustness and accuracy of the hank processing bearing wear detection method can be significantly improved. Compared with the judgment under the fixed reference, the present application can adaptively cope with the change of the operating condition of the hank machine, effectively avoids false positives or false negatives caused by the change of the operating condition, and makes the judgment of the bearing wear state more accurate and reliable. This helps to timely discover and handle the bearing wear problem, prolongs the service life of the equipment, reduces the maintenance cost, and improves the production efficiency.

[0130] In a possible design, in order to pre-process the torque response signal to obtain a pre-processed torque response signal, the present application further includes the following steps:

[0131] S301, monitoring an operating condition parameter of the hank machine.

[0132] The operating condition parameter includes the yarn type, the processing speed, or the environmental humidity.

[0133] S302, analyzing the statistical characteristics of the sharp pulse in the torque response signal according to the operating condition parameter.

[0134] The statistical characteristics include the pulse amplitude distribution, the duration distribution, or the frequency energy distribution.

[0135] As a possible implementation manner, the system can calculate the average amplitude, the maximum amplitude, the mean and variance of the pulse duration, and the energy concentration in a specific frequency range of the sharp pulse under a specific operating condition as the statistical characteristics of the sharp pulse in the torque response signal through statistical analysis on the historical data or the torque response signal collected in real time.

[0136] S303, according to the statistical characteristics of the spike pulse, the threshold parameter for identifying the random spike pulse is adaptively adjusted.

[0137] As a possible implementation, the system can adjust the amplitude threshold, duration threshold or frequency energy threshold to ensure that the threshold value matches the actual characteristics of the spike pulse under the current working condition, so as to adaptively adjust the threshold parameter for identifying the random spike pulse.

[0138] It should be noted that this adjustment can be based on a pre-set lookup table, a machine learning model or a real-time optimization algorithm.

[0139] The present application can use a real-time optimization algorithm, including:

[0140] First, initialize the threshold parameter. When the twisting machine starts or the working condition is stable, the system will collect a segment of torque response signal which is considered to be free of random spike pulse interference as a reference. Based on this reference signal, its statistical characteristics such as mean (μ) and standard deviation (σ) are calculated. The initial spike pulse identification threshold parameter can be set to μ plus a pre-set sensitivity factor k multiplied by σ (i.e. threshold = μ + k * σ). This sensitivity factor k determines the sensitivity of the identification, the larger the k value, the higher the threshold, the fewer the identified spikes, and vice versa.

[0141] Second, real-time monitoring and statistical characteristics updating. During the normal operation of the twisting machine, the system will continuously monitor the incoming torque response signal. For each new signal sampling point, the algorithm will first determine whether it exceeds the current spike identification threshold.

[0142] If the sampling point does not exceed the threshold (i.e. it is considered to be a normal signal), it is included in the data set used to update the background noise statistical characteristics. This data set can be a fixed size sliding window, or μ and σ can be updated smoothly through Exponential Moving Average (EMA). In this way, μ and σ can reflect the average level and fluctuation range of the current background noise in real time, without being contaminated by the identified spike pulses.

[0143] If the sampling point exceeds the threshold (i.e. it is identified as a spike pulse), the point will not be used to update μ and σ to avoid the interference of the spike itself on the background noise statistics.

[0144] Next, the threshold parameter is adaptively adjusted. After the real-time update of μ and σ, the spike identification threshold parameter is recalculated according to the new μ and σ values. For example, the new threshold = new μ + k * new σ. In this way, the threshold parameter can be automatically adjusted with the slow changes of the background noise.

[0145] For example, assume that the twisting machine needs to identify the torque spikes caused by uneven yarn thickness during operation. The process includes:

[0146] 1. The twisting machine is stably running with standard cotton yarn. The system collects a segment of torque signals and calculates the average torque μ of the background noise = 10 Nm, the standard deviation σ = 0.5 Nm. Set the initial sensitivity factor k = 3. Therefore, the initial spike identification threshold is 10 + 3 * 0.5 = 11.5 Nm.

[0147] 2. When the subsequent torque signals fluctuate slightly around 10 Nm (for example, 10.2 Nm), which does not exceed the threshold of 11.5 Nm, these data points are used to update μ and σ. Assume that after a period of update, the background noise increases slightly, μ becomes 10.1 Nm, and σ becomes 0.52 Nm. At this time, the threshold is automatically adjusted to 10.1 + 3 * 0.52 = 11.66 Nm.

[0148] 3. If the torque signal suddenly reaches 15 Nm at some time, far exceeding the current threshold 11.66 Nm, the point is identified as a random spike pulse. The value of 15 Nm will not be used to update μ and σ to maintain the purity of the background noise statistics. Subsequently, the spike point is reconstructed (for example, replaced by the average value of its normal points before and after).

[0149] 4. The yarn type is switched from "standard cotton yarn" to "thick polyester yarn" by the twisting machine operator. The system monitors this working condition change event. Since the inherent unevenness of thick polyester yarn can cause greater normal torque fluctuations, the system adjusts the sensitivity factor k from 3 to 3.5 according to the preset rules. At the same time, the system will clear the previous μ and σ statistics, re-collect a segment of signals under the thick polyester yarn working condition, and re-calculate μ and σ. Assume that under the new working condition, μ becomes 11 Nm, and σ becomes 0.8 Nm. At this time, the new spike identification threshold will be adjusted to 11 + 3.5 * 0.8 = 13.8 Nm.

[0150] S304, using the adjusted threshold parameter, identifying and reconstructing the spike pulse of the torque response signal to eliminate the random spike pulse interference caused by uneven yarn thickness, and obtaining the preprocessed torque response signal.

[0151] As a possible implementation, the system can minimize the impact of the spike pulse based on interpolation (such as linear interpolation, spline interpolation) or local smoothing processing (such as moving average, Gaussian smoothing) of the signal before and after the spike pulse, so as to obtain a pretreated torque response signal that removes the random spike pulse interference caused by uneven yarn thickness.

[0152] In some preferred embodiments, the following is described by a specific example. Assume that the twisting machine is processing different yarn types, for example, switching from processing fine yarn to processing thick yarn. When processing fine yarn, random spike pulses caused by uneven yarn thickness may exhibit lower amplitude and shorter duration; while when processing thick yarn, due to the increase in yarn diameter and unevenness, the amplitude of the spike pulse may be higher and the duration may also be longer. The scheme of the present application will monitor the operating condition parameter of the change in yarn type, and analyze the statistical characteristics of the spike pulse in the torque response signal under the current operating condition, such as pulse amplitude distribution and duration distribution, based on historical data or a preset model. Based on these analysis results, the system will automatically adjust the threshold parameters for identifying random spike pulses. For example, when switching to thick yarn processing, the amplitude threshold and duration threshold for identifying spike pulses will be correspondingly increased to avoid misidentifying normal operation fluctuations as random spike pulses, while ensuring that random spike pulses specific to thick yarn processing can be accurately captured. Subsequently, the torque response signal is subjected to spike pulse identification and reconstruction using these adjusted threshold parameters, for example, by local interpolation or smoothing processing to fill in the identified spike pulse area, so as to obtain a pretreated torque response signal that removes the interference of uneven yarn thickness, providing more accurate data for subsequent bearing wear feature identification.

[0153] Through the above technical solutions, the present application can significantly improve the adaptability and accuracy of the twisting processing bearing wear detection method under complex and variable operating conditions. Especially when the operating condition parameters such as yarn type, processing speed or environmental humidity change, the scheme can ensure that random spike pulse interference is effectively and accurately eliminated, avoiding misjudgment or omission due to improper preprocessing, thereby improving the reliability and accuracy of bearing wear state judgment, helping to timely discover and handle bearing wear problems, prolonging the service life of the equipment and reducing maintenance costs.

[0154] In a possible design, in order to identify and reconstruct the torque response signal, the present application further comprises:

[0155] S401、extracting time domain features and frequency domain features of candidate spike pulses from the torque response signal.

[0156] Wherein, the time domain features include pulse amplitude, duration or slope, and the frequency domain features include energy of a specific frequency band.

[0157] S402, determine the discrimination threshold or rule for distinguishing random spike pulse, bearing wear characteristic signal and machine normal operation transient fluctuation according to the time domain feature, the frequency domain feature and the operation condition parameter of the twisting machine.

[0158] Among them, the discrimination threshold or rule can be a preset fixed value, or dynamically adjusted according to historical data or machine learning model.

[0159] For example, a magnitude threshold can be set, and signals higher than the threshold are initially considered to be spike pulses; At the same time, combined with the duration threshold, the normal transient fluctuation with too long duration is excluded. In addition, frequency domain features can be used, such as by analyzing the energy ratio of specific frequency components, to distinguish random interference and harmonic signals caused by bearing wear. The operation condition parameters of the twisting machine, such as yarn type, processing speed or environmental humidity, have a significant impact on signal characteristics, so they need to be considered when determining the discrimination threshold or rule to improve the accuracy of identification.

[0160] S403, identify the random spike pulse in the torque response signal according to the discrimination threshold or rule.

[0161] As a possible implementation, the system can determine the pulses meeting the discrimination threshold or rule as random spike pulses in the torque response signal.

[0162] S404, signal reconstruction is performed on the identified random spike pulse; Signal reconstruction includes interpolation or local smoothing processing based on the signals before and after the spike pulse, so as to identify and reconstruct the spike pulse in the torque response signal.

[0163] Among them, the purpose of signal reconstruction is to eliminate the influence of these interference signals on subsequent analysis, while retaining as much effective information as possible in the non-interference part of the original signal.

[0164] As a possible implementation, the system can perform signal reconstruction on the identified random spike pulse based on interpolation or local smoothing processing of the signals before and after the spike pulse. Interpolation methods such as linear interpolation, spline interpolation, etc. estimate and fill the data in the spike pulse area by using the effective data points before and after the spike pulse. Local smoothing processing such as moving average, Gaussian smoothing, etc. reduces the amplitude of the spike pulse by weighted average of the data in the spike pulse area and its vicinity, so as to make it smooth.

[0165] By the technical solution, the precision and robustness of signal preprocessing in the hank processing bearing wear detection method can be improved. Specifically, by performing multi-dimensional feature extraction on the torque response signal and adaptive discrimination based on working condition parameters, random sharp pulse caused by uneven yarn thickness, bearing wear characteristic signal and machine normal operation transient fluctuation can be more accurately identified and distinguished. This avoids the misjudgment or missed judgment problem that may occur in the traditional method, ensuring that the preprocessed torque response signal can more truly reflect the actual state of the transmission system. Thus, high-quality input data is provided for subsequent nonlinear feature analysis, thereby improving the accuracy and reliability of bearing wear state judgment and reducing production loss and maintenance cost caused by false positives or false negatives.

[0166] As shown in Figure 3 The hank processing bearing wear detection system provided by the embodiment of the application also includes:

[0167] The detection torque injection module is configured to inject a detection torque fluctuation with a specific frequency or broadband characteristic into the transmission system of the hank machine.

[0168] The signal acquisition module is configured to synchronously acquire a torque response signal of the driving motor in the transmission system and an original instruction signal of the detection torque fluctuation.

[0169] The signal preprocessing module is configured to preprocess the torque response signal to obtain a preprocessed torque response signal. The preprocessing includes eliminating random sharp pulse interference caused by uneven yarn thickness.

[0170] The feature recognition module is configured to analyze the relationship between the preprocessed torque response signal and the original instruction signal to identify a nonlinear feature caused by bearing wear.

[0171] The state judgment module is configured to judge the bearing wear state based on long-term trend changes of the nonlinear feature.

[0172] The computer program can be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the application. One or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the terminal device.

[0173] The terminal device can be a desktop computer, a notebook computer, a palm computer, a smart tablet, and the like. The terminal device can include, but is not limited to, a processor, a memory. Those skilled in the art can understand that the above components are only examples of the terminal device, and do not constitute a limitation on the terminal device, and can include more or fewer components than the above, or combine certain components, or different components, for example, the terminal device can also include an input / output device, a network access device, a bus, and the like.

[0174] The processor can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, and the like. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, and the like. The processor is a control center of the terminal device, and connects all parts of the terminal device through various interfaces and lines.

[0175] The memory can be used to store computer programs and / or modules, and the processor realizes various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory, and calling data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application program required by a function (such as a sound playing function, an image playing function, and the like), and the like; and the data storage area can store data created according to use of the terminal device (such as audio data, a phone book, and the like), and the like. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, for example, a hard disk, a memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state memory device.

[0176] The modules / units integrated in the terminal device, if realized in the form of software function units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on such understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of each method embodiment when executed by a processor. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or some intermediate forms, etc. The computer-readable medium can include any entity or system that can carry computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0177] It should be noted that the above-described system embodiments are only schematic, and the units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, that is, they can be located in one place, or distributed on multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment scheme according to actual needs. In addition, the connection relationship between the modules in the system embodiment provided by the present application indicates that there is a communication connection between them, which can be realized as one or more communication buses or signal lines. Those skilled in the art can understand and implement it without creative labor.

[0178] The above specific embodiments further illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above is only a specific embodiment of the present application and is not used to limit the protection scope of the present application. It is particularly pointed out that any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for detecting bearing wear in skein yarn processing, characterized in that, Includes the following steps: Injecting detection torque fluctuations with specific frequency or wide bandwidth characteristics into the transmission system of the yarn skein machine; The torque response signal of the drive motor in the transmission system and the original command signal for detecting torque fluctuations are collected simultaneously. The torque response signal is preprocessed to obtain a preprocessed torque response signal; the preprocessing includes eliminating random spike pulse interference caused by uneven yarn thickness; The relationship between the preprocessed torque response signal and the original command signal is analyzed to identify nonlinear characteristics caused by bearing wear; Based on the long-term trend changes of the aforementioned nonlinear characteristics, the bearing wear condition is determined; The analysis of the relationship between the preprocessed torque response signal and the original command signal to identify nonlinear characteristics caused by bearing wear includes: Acquire internal execution signals of the drive system; Based on the original command signal for detecting torque fluctuations and the internal execution signal of the drive system, the execution deviation of the drive system is determined; Based on the execution deviation of the drive system, the preprocessed torque response signal is compensated to obtain a compensated torque response signal. The relationship between the compensated torque response signal and the original command signal is analyzed to identify nonlinear characteristics caused by bearing wear; The analysis of the relationship between the compensated torque response signal and the original command signal, to identify nonlinear characteristics caused by bearing wear, includes: The frequency domain transformation of the compensated torque response signal is performed to obtain the spectrum of the compensated torque response signal; Based on the spectrum of the compensated torque response signal, identify the harmonic frequency components in the spectrum caused by bearing wear; the harmonic frequency components are harmonics of the fundamental frequency component in the original command signal. Calculate the ratio of the energy of the harmonic frequency component to the energy of the corresponding fundamental frequency component in the original command signal based on the energy of the harmonic frequency component and the energy of the corresponding fundamental frequency component in the original command signal. Based on the ratio, determine the nonlinear characteristics caused by bearing wear; The preprocessing of the torque response signal to obtain a preprocessed torque response signal includes: Monitor the operating parameters of the yarn skein machine, including yarn type, processing speed, or ambient humidity; Based on the operating condition parameters, analyze the statistical characteristics of the spike pulses in the torque response signal; the statistical characteristics include pulse amplitude distribution, duration distribution, or frequency energy distribution. Based on the statistical characteristics of the spike pulses, the threshold parameters used to identify the random spike pulses are adaptively adjusted. Using the adjusted threshold parameters, the torque response signal is subjected to spike pulse identification and reconstruction to eliminate random spike pulse interference caused by uneven yarn thickness, thereby obtaining a preprocessed torque response signal. The process of identifying and reconstructing spike pulses in the torque response signal includes: Extract the time-domain and frequency-domain features of candidate spike pulses from the torque response signal. The time-domain features include pulse amplitude, duration, or slope, and the frequency-domain features include energy in a specific frequency band. Based on the time-domain features, the frequency-domain features, and the operating parameters of the yarn twisting machine, a discrimination threshold or rule is determined to distinguish between the random spike pulses, bearing wear characteristic signals, and transient fluctuations during normal machine operation. Based on the discrimination threshold or rule, identify random spike pulses in the torque response signal; The identified random spike pulses are reconstructed; the signal reconstruction includes interpolation or local smoothing based on the signals before and after the spike pulse, thereby identifying and reconstructing the spike pulses in the torque response signal.

2. The method for detecting bearing wear in skein yarn processing according to claim 1, characterized in that, The determination of bearing wear status based on the long-term trend change of the nonlinear characteristics includes: Monitor the operating parameters of the yarn twisting machine; Based on the operating condition parameters, identify operating condition change events; When the operating condition change event is detected, the baseline range of the nonlinear characteristic is re-established; Based on the re-established baseline range, the long-term trend changes of the nonlinear characteristics are judged to determine the bearing wear condition.

3. The method for detecting bearing wear in skein yarn processing according to claim 1, characterized in that, The injection of a probed torque fluctuation with specific frequency or wideband characteristics into the transmission system of the yarn winch includes: Adjust the parameters of the anti-overlap control function of the yarn twisting machine; the adjusted parameters are used to make the anti-overlap control function generate torque fluctuations with specific frequency or wide bandwidth characteristics; The torque fluctuation is used as the detected torque fluctuation, and the detected torque fluctuation is injected into the transmission system.

4. The method for detecting bearing wear in skein yarn processing according to claim 1, characterized in that, The acquisition of the torque response signal of the drive motor in the transmission system includes: Monitor the current and voltage signals of the drive motor; The torque response signal of the drive motor is estimated based on the current and voltage signals of the drive motor and the parameters of the drive motor.

5. The method for detecting bearing wear in skein yarn processing according to claim 1, characterized in that, The synchronous acquisition of the torque response signal of the drive motor in the transmission system, and the original command signal for detecting torque fluctuations, includes: A unified clock source is provided, which is used to synchronize the clocks of the system that collects the torque response signal of the drive motor in the transmission system and the system that generates the original command signal for detecting torque fluctuations. Based on the unified clock source, the torque response signal of the drive motor in the transmission system and the original command signal for detecting torque fluctuations are synchronously acquired.

6. A bearing wear detection system for yarn twisting, characterized in that, For performing the method as described in claim 1, the system comprises: The detection torque injection module is used to inject detection torque fluctuations with specific frequency or wide bandwidth characteristics into the transmission system of the yarn winch. The signal acquisition module is used to synchronously acquire the torque response signal of the drive motor in the transmission system, as well as the original command signal for detecting torque fluctuations. The signal preprocessing module is used to preprocess the torque response signal to obtain a preprocessed torque response signal; the preprocessing includes eliminating random spike pulse interference caused by uneven yarn thickness. The feature recognition module is used to analyze the relationship between the preprocessed torque response signal and the original command signal in order to identify nonlinear features caused by bearing wear; The status determination module is used to determine the bearing wear status based on the long-term trend changes of the nonlinear characteristics.

Images