Intelligent data processing method of intelligent needle management equipment
By constructing the maximum and minimum value envelopes of the needle vibration signal, the signal bias factor is obtained and the envelope weight of the EMD algorithm is adjusted, the problem of incomplete separation of asymmetric fluctuations in the needle vibration signal decomposition is solved, and a more accurate needle state analysis is achieved.
Patent Information
- Application Number
- CN202510622507.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-15
- Publication Date
- 2025-08-26
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing EMD empirical modal decomposition algorithm cannot effectively separate the asymmetric fluctuation characteristics in the abnormal state of the needle when processing the vibration signal of the needle, resulting in low decomposition accuracy and affecting the accuracy of the needle state analysis.
By constructing the maximum and minimum value envelopes of the needle vibration signal timing, the signal bias factor is obtained, the envelope weight in the EMD algorithm is adjusted, the vibration signal is adaptively decomposed to extract pure IMF components, and the decomposition process is optimized using technical means such as SDT algorithm and cosine similarity.
It improves the accuracy of needle vibration signal decomposition and the accuracy of needle state analysis, ensures that abnormal characteristics are clearly retained in the IMF component, and reduces misjudgment and false alarms.
Smart Images
Figure CN120541479A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing technology, and in particular to a data intelligent processing method for a needle intelligent management device. Background Art
[0002] The intelligent needle management device is an intelligent management tool for industrial sewing machines. Its main function is to monitor the condition, service life, and wear of sewing machine needles, thereby reducing production line downtime and maintenance costs, improving textile production efficiency, and reducing the incidence of failures. Needle anomalies manifest not only as wear, deformation, or breakage of the needle itself, but also as abnormalities during operation. For example, the needle may experience frequent jamming when piercing high-density fabrics or abnormally loose thread tension during operation, causing the needle to be repeatedly subjected to abnormal impacts, accelerating needle wear and deformation. When encountering such abnormal conditions, the needle's vibration signal will become abnormal.
[0003] The existing EMD empirical mode decomposition algorithm can decompose nonlinear and non-stationary needle vibration signals, thereby capturing subtle changes in the vibration signal, and is particularly suitable for extracting early abnormal characteristics of the needle. When the needle encounters an abnormal condition, it will cause the needle to be subjected to asymmetric reverse impact or inertia during the movement phase, and the extreme values of the vibration signal will be abnormally amplified, forming an asymmetric fluctuation characteristic of the signal; the EMD algorithm constructs upper and lower envelopes based on local extreme points, and calculates the equal-weighted mean as the local trend for separation. In this process, it is assumed that the signal fluctuation is symmetrical; but under abnormal needle working conditions, the vibration signal shows obvious bias, with more intense fluctuations on one side and relatively stable on the other side. If the signal is decomposed by equal-weighted mean, the abnormal peak biased to one side will be weakened, so that the remaining signal in the decomposition still retains the trend characteristics, the components are not effectively separated, and the abnormal behavior cannot be effectively reflected in the IMF, affecting the purity of the IMF components and reducing the accuracy of the analysis of the true abnormal state of the needle. Summary of the Invention
[0004] In order to solve the technical problem that abnormal needle working conditions may cause the vibration signal to present asymmetric fluctuation characteristics, resulting in low accuracy of EMD algorithm in decomposing the needle vibration signal and affecting the accuracy of needle status analysis, the purpose of the present invention is to provide a data intelligent processing method for needle intelligent management equipment. The technical solutions adopted are as follows: Obtain the vibration signal timing of the needle during sewing; Constructing envelopes of the maximum and minimum values of the vibration signal time series respectively; obtaining a signal deflection factor according to the fluctuation difference characteristics and position distribution characteristics of the two envelopes of the vibration signal time series; Decomposing the vibration signal time series using an EMD algorithm, obtaining IMF components and corresponding residual signals in sequence during the decomposition process; obtaining an adjusted envelope of the residual signal based on the signal deflection factor; obtaining a similarity based on difference characteristics between the IMF components and the corresponding residual signal; and obtaining asymmetry of the residual signal based on discrete characteristics of extreme values and data difference characteristics in the residual signal; The weight of the envelope to be adjusted is adjusted according to the similarity and the asymmetry to obtain an adaptive weight; and all IMF components of the vibration signal time series are obtained by further decomposing the envelope of the residual signal and the adaptive weight using an EMD algorithm.
[0005] Furthermore, the step of obtaining a signal deflection factor according to the fluctuation difference characteristics and position distribution characteristics of the two envelopes of the vibration signal time series includes: The inflection point of the envelope is obtained according to the SDT algorithm; the difference in the number of inflection points between the envelope of the maximum value and the envelope of the minimum value in the vibration signal time series is calculated and mapped through the hyperbolic tangent function to obtain the quantity difference eigenvalue; the integral of the function corresponding to the envelope of the maximum value in the vibration signal time series is calculated to obtain a first value; the integral of the function corresponding to the absolute value of the envelope of the minimum value in the vibration signal time series is calculated to obtain a second value; the difference between the first value and the second value is calculated and mapped through the hyperbolic tangent function to obtain the area difference eigenvalue; the average value of the quantity difference eigenvalue and the area difference eigenvalue is calculated to obtain the signal deflection factor.
[0006] Furthermore, the step of obtaining the envelope to be adjusted of the residual signal according to the signal deflection factor includes: When the signal deflection factor exceeds the constant 0, the envelope corresponding to the maximum value in the remaining signal is used as the envelope to be adjusted; when the signal deflection factor does not exceed the constant 0, the envelope corresponding to the minimum value in the remaining signal is used as the envelope to be adjusted; when the signal deflection factor is the constant 0, there is no envelope to be adjusted.
[0007] Furthermore, the step of obtaining the similarity degree according to the difference characteristics between the IMF component and the corresponding residual signal includes: The cosine similarity between the IMF component and the corresponding residual signal is calculated to obtain the similarity between the IMF component and the corresponding residual signal.
[0008] Furthermore, the step of obtaining the asymmetry degree of the residual signal according to the discrete characteristics of the extreme values and the data difference characteristics in the residual signal includes: The variance of the extreme values in the residual signal distributed on the envelope to be adjusted is calculated and mapped through a hyperbolic tangent function to obtain a degree of discreteness; the absolute value of the sum of all extreme values in the residual signal is calculated and mapped through a hyperbolic tangent function to obtain a degree of difference; the average value of the degree of discreteness and the degree of difference is calculated to obtain a degree of asymmetry of the residual signal.
[0009] Furthermore, the step of adjusting the weight of the envelope to be adjusted according to the similarity and the asymmetry to obtain an adaptive weight includes: An average value of the similarity degree and the asymmetry degree is calculated to obtain an adjustment coefficient; and a sum of the adjustment coefficient and a constant 1 is calculated to obtain an adaptive weight of the envelope to be adjusted.
[0010] Furthermore, the step of further decomposing the envelope of the residual signal and the adaptive weight using an EMD algorithm to obtain all IMF components of the vibration signal time series includes: The product of the envelope to be adjusted of the residual signal and the adaptive weight is calculated to obtain a weighted envelope, the envelope to be adjusted in the residual signal is replaced with the weighted envelope, and the residual signal is further decomposed using an EMD algorithm to obtain all IMF components of the vibration signal time series.
[0011] The present invention has the following beneficial effects: In the present invention, obtaining the envelope of the vibration signal time series can be used to analyze the bias characteristics of the vibration signal time series. Obtaining the signal bias factor can characterize the bias and asymmetry of the vibration signal time series, thereby determining the envelope that needs to be adjusted. Obtaining the envelope to be adjusted can determine whether the envelope of the residual signal needs to be adjusted and the specific adjustment target, thereby initially improving the accuracy of vibration signal decomposition. Obtaining the degree of similarity between the IMF components and the corresponding residual signal can indicate whether the decomposition is complete and whether the component characteristics are thoroughly decomposed, thereby determining the degree of adjustment of the envelope to be adjusted. Obtaining the degree of asymmetry of the residual signal can characterize the bias characteristics of the residual signal and whether the needle is encountering an abnormal condition, thereby improving the adjustment accuracy of the envelope to be adjusted. Obtaining the adaptive weight can determine the adjustment weight of the envelope to be adjusted. Finally, based on the envelope of the residual signal and the adaptive weight, the EMD algorithm is used to further decompose and obtain all the IMF components of the vibration signal time series. By adjusting the envelope weights, the aliasing of the characteristics of the IMF components at different frequencies can be avoided, and the true trend of the vibration signal can be more accurately reflected, with abnormal characteristics clearly retained in each IMF component. BRIEF DESCRIPTION OF THE DRAWINGS
[0012] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0013] Figure 1 A flow chart of a data intelligent processing method for a needle intelligent management device provided in one embodiment of the present invention. DETAILED DESCRIPTION
[0014] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the data intelligent processing method of a machine needle intelligent management device proposed by the present invention, its specific implementation method, structure, characteristics and effects are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.
[0015] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
[0016] The following describes in detail the specific scheme of the data intelligent processing method of the needle intelligent management equipment provided by the present invention with reference to the accompanying drawings.
[0017] See also Figure 1 , which shows a flow chart of a data intelligent processing method for a needle intelligent management device provided by one embodiment of the present invention, the method comprising the following steps: Step S1, obtaining the vibration signal timing of the needle during the sewing process.
[0018] In an embodiment of the present invention, the implementation scenario is to analyze the working status of the needle, monitor and manage the status and service life of the needle of an industrial sewing machine, improve the efficiency of textile production, and reduce the failure rate. First, the vibration signal timing of the needle during the sewing process is obtained. The vibration sensor is installed on the needle fixing base or near the needle bar. The collected vibration signal is filtered through a low-pass filter to reduce the interference of random noise on the EMD decomposition. Finally, the vibration signal timing of the needle of the industrial sewing machine during the sewing process is obtained.
[0019] Step S2: construct the envelopes of the maximum and minimum values of the vibration signal time series respectively; and obtain the signal deflection factor according to the fluctuation difference characteristics and position distribution characteristics of the two envelopes of the vibration signal time series.
[0020] When the needle frequently jams while piercing high-density fabrics or experiences abnormally loose thread tension during operation, this indicates wear, bending, deformation, or a needle mechanism malfunction. The intelligent needle management device needs to provide an early warning of the needle's status. The needle must overcome significant fabric resistance when piercing high-density fabrics, resulting in significant force recoil during the needle's downward stroke, amplifying the positive local maximum of the vibration signal. However, during the needle's return stroke, the resistance is removed or significantly reduced, and the negative minimum does not change significantly. When the sewing thread becomes loose or breaks, the needle loses the cushioning effect provided by the thread tension during the return stroke or force rebound phase, resulting in a sudden downward inertial impact, amplifying the negative local minimum of the vibration signal. Therefore, when the needle encounters an abnormality during operation, it will cause asymmetric fluctuations in the vibration signal timing. The vibration signal timing under these conditions differs from that during normal operation, allowing the vibration signal to be used to analyze whether the needle is abnormal. The existing EMD empirical mode decomposition algorithm can be used to decompose the vibration signal time series, and the decomposed components can be used to analyze whether the needle is abnormal. EMD constructs upper and lower envelopes based on local extreme points, and calculates the equal-weighted mean as the local trend for separation. In this process, the signal fluctuation is assumed to be symmetrical. However, under abnormal needle working conditions, the vibration signal shows obvious bias, with more intense fluctuations on one side and relatively stable on the other side. If the signal is decomposed using the equal-weighted mean, the abnormal peak biased to one side will be weakened, so that the remaining signal in the decomposition still retains the trend characteristics, the components are not effectively separated, and the abnormal behavior cannot be effectively reflected in the IMF component, affecting the purity of the IMF component and the accuracy of the needle abnormality analysis. Therefore, it is necessary to adjust the envelope during the decomposition process to improve the decomposition accuracy.
[0021] Furthermore, the envelopes of the maximum and minimum values of the vibration signal time series are constructed respectively. It should be noted that the acquisition of the envelopes belongs to the existing technology, and the specific steps will not be repeated here. When there is a difference in the symmetry of the upper envelope of the maximum value and the lower envelope of the minimum value about the horizontal axis, it means that the envelopes have a bias feature, and the fluctuation of one envelope is more intense. Therefore, it is necessary to determine whether the vibration signal time series has a bias, so the signal bias factor is obtained based on the fluctuation difference characteristics and position distribution characteristics of the two envelopes of the vibration signal time series.
[0022] Preferably, in an embodiment of the present invention, the step of obtaining the signal deflection factor includes: obtaining the inflection point of the envelope according to the SDT algorithm; it should be noted that the SDT revolving door belongs to the existing technology, and the specific steps will not be repeated. When the machine needle encounters an abnormal situation, the vibration signal usually shows asymmetry, in which the extreme point on one side fluctuates more violently, the trend changes more frequently, and the number of inflection points is greater. The difference in the number of inflection points between the envelope of the maximum value and the envelope of the minimum value in the vibration signal time series is calculated and mapped through the hyperbolic tangent function to obtain the quantity difference eigenvalue; if the difference in the number of inflection points of the two envelopes is greater, the quantity difference eigenvalue is less close to 0, the envelope is more likely to have a deflection problem, and the machine needle is more likely to encounter high-frequency jamming and line loosening. The hyperbolic tangent function maps the inflection point quantity difference between -1 and 1. When the upper envelope has more inflection points, the quantity difference eigenvalue is closer to 1, and the volatility of the maximum value of the vibration signal time series is more obvious. Calculate the integral of the function corresponding to the envelope of the maximum value in the vibration signal time series to obtain a first value; the first value represents the area characteristics of the upper envelope and the horizontal axis. Calculate the integral of the function corresponding to the absolute value of the envelope of the minimum value in the vibration signal time series to obtain a second value; the second value represents the area characteristics of the lower envelope and the horizontal axis. Because the lower envelope is a negative number, it is necessary to calculate the integral of the function corresponding to the absolute value of the lower envelope. Calculate the difference between the first and second values and map it through the hyperbolic tangent function to obtain the area difference eigenvalue; the closer the area difference eigenvalue is to 0, the stronger the asymmetry of the two envelopes. The closer the area difference eigenvalue is to 1, the farther the upper envelope is from the horizontal axis, the more the vibration signal time series is biased toward the maximum value, and the stronger the bias and asymmetry. Calculate the average of the quantity difference eigenvalue and the area difference eigenvalue to obtain the signal deflection factor; when the signal deflection factor is larger, it means that the upper envelope of the vibration signal time series fluctuates more frequently, and when the signal deflection factor is smaller, it means that the lower envelope of the vibration signal time series fluctuates more frequently; when the signal deflection factor is closer to 0, it means that the bias and asymmetry of the vibration signal time series are more obvious, and the needle is more likely to encounter abnormal conditions, and therefore it is necessary to highlight the abnormal characteristics of the vibration signal during the decomposition process. The formula for obtaining the signal deflection factor includes:
[0023] Where P represents the signal bias factor, represents the hyperbolic tangent function, Indicates the number of inflection points in the upper envelope corresponding to the maximum value, Indicates the number of inflection points in the lower envelope corresponding to the minimum value, represents the quantitative difference eigenvalue, represents the first value, represents the second value, Represents the area difference eigenvalue.
[0024] Step S3, decompose the vibration signal time series by the EMD algorithm, and obtain the IMF components and the corresponding residual signals in sequence during the decomposition process; obtain the adjusted envelope of the residual signal according to the signal deflection factor; obtain the similarity degree according to the difference characteristics of the IMF components and the corresponding residual signals; obtain the asymmetry degree of the residual signal according to the discrete characteristics of the extreme values in the residual signal and the data difference characteristics.
[0025] After determining the deviation of the vibration signal timing, the weight of the envelope can be adjusted according to the deviation; first, the vibration signal timing is decomposed by the EMD algorithm, and the IMF components and the corresponding residual signals are obtained in sequence during the decomposition process; it should be noted that the EMD algorithm belongs to the existing technology, and the specific steps are not repeated here. Then, the envelope to be adjusted of the residual signal is obtained according to the signal deviation factor; preferably, in an embodiment of the present invention, the step of obtaining the envelope to be adjusted includes: when the signal deviation factor exceeds the constant 0, it means that the vibration signal timing is biased to the maximum value, so the envelope corresponding to the maximum value in the residual signal is used as the envelope to be adjusted. When the signal deviation factor does not exceed the constant 0, it means that the vibration signal timing is biased to the minimum value, and the envelope corresponding to the minimum value in the residual signal is used as the envelope to be adjusted. When the signal deviation factor is a constant 0, it means that the symmetry of the vibration signal timing about the horizontal axis is relatively strong, and the possibility of the needle encountering an abnormal state is small, so there is no envelope to be adjusted, and the vibration signal timing can be decomposed normally.
[0026] Furthermore, since EMD decomposition is performed step by step, the asymmetric characteristics of the residual signal after each decomposition to obtain an IMF component will change. Therefore, it is necessary to analyze the residual signal after each decomposition and adaptively adjust the weight of each residual signal's envelope. In EMD decomposition, each order IMF component should extract a certain layer of frequency components from the original signal as much as possible, making the residual signal gradually smooth and monotonic. If the IMF component is not fully extracted, some features will remain in the residual signal. Therefore, the similarity is obtained according to the difference characteristics of the IMF component and the corresponding residual signal; preferably, in an embodiment of the present invention, the step of obtaining the similarity includes: calculating the cosine similarity of the IMF component and the corresponding residual signal, and obtaining the similarity between the IMF component and the corresponding residual signal; it should be noted that cosine similarity belongs to the prior art, and the specific calculation steps are not repeated here. The more similar the two vectors are, the greater the cosine similarity is. The IMF component and the corresponding residual signal are regarded as one-dimensional vectors. The greater the similarity between the IMF component and the corresponding residual signal, it means that the fluctuation trend in the corresponding IMF component is retained in the residual signal, the component is not completely extracted, the decomposition process is not thorough, and the subsequent adjustment degree of the adjusted envelope is greater.
[0027] If the distribution difference between the maximum and minimum values corresponding to the residual signal is large, it means that the asymmetric feature is more obvious, so the asymmetric degree of the residual signal is obtained according to the discrete features and data difference features of the extreme values in the residual signal; preferably, in an embodiment of the present invention, the step of obtaining the asymmetric degree includes: calculating the variance of the extreme values distributed in the residual signal on the envelope to be adjusted and mapping it through the hyperbolic tangent function to obtain the discrete degree; the envelope to be adjusted is the envelope in the deflection direction. When the variance of the extreme values of the residual signal on the envelope to be adjusted is larger, it means that the fluctuation of the envelope to be adjusted is more obvious, and the needle is more likely to encounter an abnormal state, such as the jamming of the needle when passing through high-density fabrics, which causes the local maximum value of the vibration signal to be amplified, increasing the discrete features between the extreme values. Therefore, when the discrete degree is larger, it means that the asymmetric feature of the residual signal is more obvious, and the envelope needs to be adjusted more. Calculate the absolute value of the sum of all extreme values in the residual signal and map it through the hyperbolic tangent function to obtain the degree of difference. The more obvious the symmetry of the residual signal about the horizontal axis, the closer the sum of the maximum and minimum values will be to 0. A greater degree of difference means that the residual signal is more biased towards the extreme value on one side, the more likely the needle will encounter an abnormal state, and the more obvious the asymmetry. Calculate the average of the degree of dispersion and the degree of difference to obtain the degree of asymmetry of the residual signal. The greater the degree of asymmetry, the more biased the residual signal is towards the side of the envelope to be adjusted, and the more necessary it is to adjust the weight of the envelope to be adjusted.
[0028] Step S4: Adjust the weight of the envelope to be adjusted according to the degree of similarity and asymmetry to obtain an adaptive weight; continue to decompose the envelope of the remaining signal and the adaptive weight using the EMD algorithm to obtain all IMF components of the vibration signal time series.
[0029] After obtaining the degree of similarity and asymmetry, the weight of the envelope to be adjusted can be adjusted based on the degree of similarity and asymmetry to obtain an adaptive weight. Preferably, in an embodiment of the present invention, the step of obtaining the adaptive weight includes: calculating the average of the degree of similarity and the degree of asymmetry to obtain an adjustment coefficient. A larger adjustment coefficient means that a greater weight is given to the envelope to be adjusted when calculating the envelope mean curve during the residual signal decomposition process, thereby highlighting abnormal features in the vibration signal, allowing the envelope mean curve to more accurately reflect the central trend of the true signal, and preventing important abnormal vibration signals from being weakened by the mean curve. The adaptive weight of the envelope to be adjusted is obtained by calculating the sum of the adjustment coefficient and a constant 1.
[0030] Furthermore, the EMD algorithm can be used to further decompose the residual signal based on its envelope and adaptive weights to obtain all IMF components of the vibration signal time series, specifically including: calculating the product of the envelope to be adjusted of the residual signal and the adaptive weight to obtain a weighted envelope. Compared with the envelope to be adjusted, the weighted envelope can better highlight the bias characteristics and asymmetric characteristics of the vibration signal, and highlight the abnormal vibration characteristics. The envelope to be adjusted in the residual signal is replaced with the weighted envelope, and the residual signal is further decomposed using the EMD algorithm to obtain all IMF components of the vibration signal time series. By adjusting the weights of the envelope, the aliasing of the characteristics of the IMF components at different frequencies can be avoided, and the true trend of the vibration signal can be more accurately reflected, and the abnormal characteristics are clearly retained in each IMF component. Finally, the vibration signal can be analyzed based on the decomposed IMF components to determine whether it is abnormal, thereby determining whether the needle state is abnormal and issuing an early warning, thereby improving the accuracy of the needle state analysis.
[0031] In summary, an embodiment of the present invention provides a method for intelligent data processing of a machine needle intelligent management device; a signal deflection factor is obtained based on the fluctuation difference characteristics and position distribution characteristics of the two envelopes of the vibration signal time series; the vibration signal time series is decomposed by the EMD algorithm; the envelope to be adjusted of the remaining signal is obtained according to the signal deflection factor; the degree of similarity is obtained based on the difference characteristics of the IMF component and the corresponding remaining signal; the degree of asymmetry of the remaining signal is obtained based on the discrete characteristics of the extreme values in the remaining signal and the data difference characteristics; the weight of the envelope to be adjusted is adjusted based on the degree of similarity and the degree of asymmetry. The present invention continues to decompose the envelope of the remaining signal and the adaptive weight through the EMD algorithm to obtain all the IMF components of the vibration signal time series, thereby improving the accuracy of signal decomposition and needle state analysis.
[0032] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0033] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.
Claims
1. A data intelligent processing method for needle intelligent management equipment, characterized in that: The method comprises the following steps: Obtain the vibration signal timing of the needle during sewing; Constructing envelopes of the maximum and minimum values of the vibration signal time series respectively; obtaining a signal deflection factor according to the fluctuation difference characteristics and position distribution characteristics of the two envelopes of the vibration signal time series; Decomposing the vibration signal time series using an EMD algorithm, obtaining IMF components and corresponding residual signals in sequence during the decomposition process; obtaining an adjusted envelope of the residual signal based on the signal deflection factor; obtaining a similarity based on difference characteristics between the IMF components and the corresponding residual signal; and obtaining asymmetry of the residual signal based on discrete characteristics of extreme values and data difference characteristics in the residual signal; The weight of the envelope to be adjusted is adjusted according to the similarity and the asymmetry to obtain an adaptive weight; and all IMF components of the vibration signal time series are obtained by further decomposing the envelope of the residual signal and the adaptive weight using an EMD algorithm.
2. The data intelligent processing method of the needle intelligent management device according to claim 1 is characterized in that: The step of obtaining the signal deflection factor according to the fluctuation difference characteristics and position distribution characteristics of the two envelopes of the vibration signal time series includes: The inflection point of the envelope is obtained according to the SDT algorithm; the difference in the number of inflection points between the envelope of the maximum value and the envelope of the minimum value in the vibration signal time series is calculated and mapped through the hyperbolic tangent function to obtain the quantity difference eigenvalue; the integral of the function corresponding to the envelope of the maximum value in the vibration signal time series is calculated to obtain a first value; the integral of the function corresponding to the absolute value of the envelope of the minimum value in the vibration signal time series is calculated to obtain a second value; the difference between the first value and the second value is calculated and mapped through the hyperbolic tangent function to obtain the area difference eigenvalue; the average value of the quantity difference eigenvalue and the area difference eigenvalue is calculated to obtain the signal deflection factor.
3. The data intelligent processing method of the needle intelligent management device according to claim 1 is characterized in that: The step of obtaining the to-be-adjusted envelope of the residual signal according to the signal deflection factor comprises: When the signal deflection factor exceeds the constant 0, the envelope corresponding to the maximum value in the remaining signal is used as the envelope to be adjusted; when the signal deflection factor does not exceed the constant 0, the envelope corresponding to the minimum value in the remaining signal is used as the envelope to be adjusted; when the signal deflection factor is the constant 0, there is no envelope to be adjusted.
4. The data intelligent processing method of the needle intelligent management device according to claim 1 is characterized in that: The step of obtaining the similarity degree according to the difference characteristics between the IMF component and the corresponding residual signal comprises: The cosine similarity between the IMF component and the corresponding residual signal is calculated to obtain the similarity between the IMF component and the corresponding residual signal.
5. The data intelligent processing method of the needle intelligent management device according to claim 3 is characterized in that: The step of obtaining the asymmetry degree of the residual signal according to the discrete characteristics of the extreme values and the data difference characteristics in the residual signal comprises: The variance of the extreme values in the residual signal distributed on the envelope to be adjusted is calculated and mapped through a hyperbolic tangent function to obtain a degree of discreteness; the absolute value of the sum of all extreme values in the residual signal is calculated and mapped through a hyperbolic tangent function to obtain a degree of difference; the average value of the degree of discreteness and the degree of difference is calculated to obtain a degree of asymmetry of the residual signal.
6. The data intelligent processing method of the needle intelligent management device according to claim 1 is characterized in that: The step of adjusting the weight of the envelope to be adjusted according to the similarity and the asymmetry to obtain an adaptive weight comprises: An average value of the similarity degree and the asymmetry degree is calculated to obtain an adjustment coefficient; and a sum of the adjustment coefficient and a constant 1 is calculated to obtain an adaptive weight of the envelope to be adjusted.
7. The data intelligent processing method of the needle intelligent management device according to claim 1 is characterized in that: The step of further decomposing the residual signal envelope and the adaptive weight using the EMD algorithm to obtain all IMF components of the vibration signal time series includes: The product of the envelope to be adjusted of the residual signal and the adaptive weight is calculated to obtain a weighted envelope, the envelope to be adjusted in the residual signal is replaced with the weighted envelope, and the residual signal is further decomposed using an EMD algorithm to obtain all IMF components of the vibration signal time series.