Numerical control machine tool spindle vibration real-time monitoring and automatic suppression method

Abstract

The invention provides a numerical control machine tool spindle vibration real-time monitoring and automatic restraining method which comprises the steps that spindle vibration parameters of a titanium alloy artificial hip joint femoral stem bionic structure in the numerical control machine tool machining process are obtained, the vibration parameters comprise the vibration frequency and the vibration amplitude, and meanwhile porosity distribution information of a machining area is collected; according to the porosity distribution information, the processing area is divided into a surface layer loose bone structure area and a deep layer compact bone structure area, vibration parameters of different areas in the processing process are extracted respectively, and the mapping relation between the vibration parameters and porosity distribution is established; and according to the vibration parameter change rule of the deep compact bone structure area, a collaborative change model of the vibration parameters and the cutting state is established, and the vibration trend of the main shaft during machining in the low-porosity area is predicted and obtained.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and particularly to a method for real-time monitoring and automatic suppression of spindle vibration in a numerically controlled machine tool. Background Art

[0002] When processing the bionic structure of the femoral stem of a titanium alloy artificial hip joint, from the surface loose bone structure to the deep dense bone structure, the porosity of the material shows a characteristic of gradient change. This difference in material structure results in different vibration responses of the spindle during machining in different regions. When the spindle cuts into the surface loose bone structure with high porosity and gradually penetrates into the dense bone structure with low porosity, the material removal rate, cutting force, and the contact state between the tool and the workpiece are all changing dynamically. This dynamic change in cutting conditions will cause severe fluctuations in the vibration frequency and amplitude of the spindle. Especially when continuously machining in the region of the dense bone structure with low porosity, the coupling effect between the vibration response and the cutting state is more significant. The non-linear evolution characteristics of spindle vibration lead to the instability of the machining process, seriously affecting the machining accuracy and surface quality of the bionic structure. Therefore, it is urgent to study the law of the coordinated change between spindle vibration and cutting state, and accordingly develop an automatic spindle vibration suppression technology based on real-time detection and feedback control of the machining state to achieve high-quality machining of the bionic structure of the femoral stem of a titanium alloy artificial hip joint. Summary of the Invention

[0003] The present invention provides a method for real-time monitoring and automatic suppression of spindle vibration in a numerically controlled machine tool, mainly including:

[0004] Obtain the spindle vibration parameters during the machining of the bionic structure of the femoral stem of a titanium alloy artificial hip joint in a numerically controlled machine tool. The vibration parameters include vibration frequency and vibration amplitude, and at the same time collect the porosity distribution information of the machining area;

[0005] According to the porosity distribution information, divide the machining area into a surface loose bone structure area and a deep dense bone structure area, respectively extract the vibration parameters during the machining process of different areas, and establish a mapping relationship between the vibration parameters and the porosity distribution;

[0006] Analyze the change trend of the vibration parameters in the surface loose bone structure area, and judge whether the vibration parameters show non-linear fluctuations with the change of porosity according to the change trend of the vibration parameters. If so, and the vibration amplitude exceeds a preset safety threshold, immediately trigger a preset vibration suppression mechanism;

[0007] Extract the vibration parameters during the continuous machining process in the deep dense bone structure area, and judge whether the vibration parameters show a coordinated change characteristic with the change of the cutting state. If the vibration frequency is positively correlated with the cutting force, record the change law of the vibration parameters under this cutting state;

[0008] According to the variation law of vibration parameters in the deep dense bone structure region, a co-variation model of vibration parameters and cutting state is established to predict the vibration trend of the spindle during machining in the low-porosity region;

[0009] Extract the characteristic features of the vibration parameters, including the vibration amplitude, vibration frequency, and the rate of change of vibration parameters over time. Combine with the co-variation model to determine whether the vibration amplitude continues to increase and the vibration frequency tends to be stable. If so, trigger the preset automatic suppression mechanism;

[0010] Through the preset automatic suppression mechanism, adjust the spindle speed and feed rate to reduce the vibration amplitude and stabilize the vibration frequency. At the same time, monitor the change of cutting state in real time, re-collect the vibration parameters, and determine whether the vibration suppression effect meets the preset requirements. If the vibration amplitude drops below the threshold and the vibration frequency tends to be stable, the automatic suppression process is completed.

[0011] Furthermore, obtain the spindle vibration parameters during the machining of the femoral stem bionic structure of the titanium alloy artificial hip joint on a numerical control machine tool. The vibration parameters include vibration frequency and vibration amplitude. At the same time, collect the porosity distribution information of the machining area, including:

[0012] Use a spindle speed sensor to obtain the spindle rotation frequency data in the machining area of the femoral stem of the titanium alloy artificial hip joint. Collect the spindle vibration frequency and vibration amplitude through a vibration sensor. Arrange a strain gauge at the machining point to measure the spindle torque. At the same time, read the spindle power parameters from the internal of the numerical control system; establish a parameter correlation matrix based on the collected spindle rotation frequency, vibration frequency, and vibration amplitude data to evaluate the spindle speed stability, thereby establish a vibration parameter distribution map in the machining area. Real-time monitor the surface roughness of the tool cutting point through an optical sensor; collect the acoustic emission signal during the machining process through an acoustic emission sensor, and establish a mapping relationship between the acoustic emission signal characteristic value and the tool feed speed and cutting depth, so as to on-line monitor the machining surface quality. Combine with the vibration parameter distribution map of the spindle to obtain the porosity distribution of the machining area; preset a threshold according to the material characteristics of the titanium alloy, compare the monitoring results with the preset threshold to judge the machining state. When it is detected that the acoustic emission signal characteristic value exceeds the preset range, adjust the tool feed speed and cutting depth until the acoustic emission signal characteristic value drops within the preset range.

[0013] Furthermore, according to the porosity distribution information, divide the machining area into the surface loose bone structure area and the deep dense bone structure area, respectively extract the vibration parameters during the machining process of different areas, and establish the mapping relationship between the vibration parameters and the porosity distribution, including:

[0014] Collect a point cloud data set from the processing area, process the point cloud data using a density clustering algorithm, mark the point sets of the loosely structured area and the densely structured area in the point cloud data set according to the porosity distribution density, and calibrate the boundaries of the point set distribution based on the regional boundary determination rules; calculate the width of the structural transition zone between the surface loose bone structure area and the deep dense bone structure area according to the boundary calibration result. For the point set within the surface area range, collect the vibration frequency parameter and the vibration amplitude parameter using a vibration sensor, and normalize the vibration parameters to obtain the surface vibration feature vector; measure the bone structure depth parameter in the regional point cloud data set through a depth detector, use support vector regression to establish the mapping relationship between the vibration feature vector and the porosity, collect the vibration frequency parameter and the vibration amplitude parameter from the deep area range, and normalize the vibration parameters to obtain the deep vibration feature vector; construct a mapping function based on the surface vibration feature vector and the deep vibration feature vector, extract the porosity threshold from the mapping function, and if the porosity of the detection point is higher than the threshold, divide it into the surface loose bone structure area, and if the porosity of the detection point is lower than the threshold, divide it into the deep dense bone structure area.

[0015] Furthermore, analyze the change trend of the vibration parameters in the surface loose bone structure area, and judge whether the vibration parameters show non-linear fluctuations with the change of porosity according to the change trend of the vibration parameters. If so, and the vibration amplitude exceeds the preset safety threshold, immediately trigger the preset vibration suppression mechanism, including:

[0016] Vibration sensor is used to collect vibration frequency data and vibration amplitude data in the area of loose surface bone structure. Time-frequency analysis is performed on the vibration frequency data through Fourier transform to generate a vibration frequency spectrum, and the main frequency component and harmonic component are extracted from the vibration frequency spectrum; the vibration frequency change rate curve is calculated according to the main frequency component and harmonic component, and the Kalman filter is used to filter the vibration amplitude data to obtain the vibration amplitude curve. The fluctuation detection period is set as the data of consecutive sampling points; the maximum change rate and minimum change rate within the fluctuation period are extracted from the vibration frequency change rate curve, the change rate fluctuation range is calculated, and whether the vibration frequency shows non-linear fluctuation is judged according to the preset non-linear judgment threshold; the porosity data is collected through the vibration sensor, and a mapping relationship between the vibration frequency change rate and the porosity change is established by using a recurrent neural network. The input layer of the recurrent neural network is the vibration frequency change rate sequence, the hidden layer is a three-layer fully connected layer, and the output layer is the porosity change value; the maximum amplitude and minimum amplitude within the fluctuation detection period are extracted from the vibration amplitude curve, the vibration amplitude fluctuation range is calculated, and when the fluctuation range exceeds the preset safety threshold, the vibration suppression mechanism is triggered; a suppression parameter mapping table is established according to the vibration frequency change rate and the vibration amplitude fluctuation range, and the suppression parameters include the spindle speed adjustment amount and the feed speed adjustment amount. When the vibration suppression mechanism is triggered, the corresponding adjustment parameters are queried from the mapping table; a proportional-integral controller is used to realize the dynamic adjustment of the spindle speed and the feed speed. The input of the controller is the deviation value between the vibration amplitude and the safety threshold, and the output is the actual speed adjustment amount and feed speed adjustment amount until the vibration amplitude drops below the safety threshold.

[0017] Furthermore, the vibration parameters in the area of deep dense bone structure during continuous machining are extracted, and it is judged whether the vibration parameters show a co-variation characteristic with the change of the cutting state. If the vibration frequency is positively correlated with the cutting force, the change rule of the vibration parameters in this cutting state is recorded, including:

[0018] Vibration sensors and cutting force sensors are used to collect vibration frequency and cutting force data in the area of deep dense bone structure. Wavelet decomposition is performed on the vibration frequency data through wavelet transform to obtain vibration frequency components and vibration amplitude components. The mean value and standard deviation of the cutting force are calculated based on the cutting force data. When the standard deviation is less than 10% of the mean value, it is determined as a stable cutting state interval, and the vibration frequency components and vibration amplitude components within this interval are recorded. The Pearson correlation coefficient is used to calculate the correlation degree between the vibration frequency and the cutting force, and the vibration parameter fluctuation values of each density layer of the deep bone structure are extracted from the change trend of the vibration frequency. The depth data of each layer of the deep bone structure is obtained through a depth sensor. When the depth value exceeds the preset depth threshold, it is determined as the area of deep dense structure, and the bone structure density distribution curve is constructed based on the depth data. The tool wear degree data is obtained from the cutting tool wear sensor, and a long short-term memory network is used to establish a time series mapping relationship between the cutting force and the vibration frequency. The network input is the cutting force sequence and the vibration frequency sequence, and the output is the co-variation eigenvalue. When the co-variation eigenvalue is greater than zero and the ratio of the vibration frequency increment to the cutting force increment is greater than the preset threshold, it is determined as a positive correlation state, and the vibration frequency, vibration amplitude, cutting force data, and tool wear degree data in this state are recorded. According to the recorded data, a cutting state parameter mapping table is constructed. The mapping table includes four parameter dimensions: vibration frequency, vibration amplitude, cutting force, and tool wear degree, and the corresponding relationship between the parameters is established.

[0019] Furthermore, based on the variation law of the vibration parameters in the area of deep dense bone structure, a co-variation model between the vibration parameters and the cutting state is established to predict the vibration trend of the spindle during machining in the low-porosity area, including:

[0020] The original vibration parameter data is collected in the deep dense bone structure area by using a vibration sensor, the cutting state parameters are obtained through a cutting state sensor, the porosity value of the low-porosity area is read from an ultrasonic porosity detector, and the original vibration parameter data is normalized by using the range normalization method; the short-time Fourier transform is calculated according to the normalized vibration parameter data to obtain the time-frequency feature matrix of the vibration parameters, and the vibration frequency feature and the vibration amplitude feature are extracted from the time-frequency feature matrix; the spindle speed data is obtained by using a spindle speed sensor, the cutting force data collected by a cutting force sensor is recorded, and a multi-dimensional feature vector including vibration features, speed data, and cutting force data is constructed; a co-variation model of the vibration parameters and the cutting state is established through a recurrent neural network, the input layer of the model is the multi-dimensional feature vector, the hidden layer uses long short-term memory units, and the output layer is the predicted value of the vibration parameter change; the Bayesian regressor is trained by using the output value of the co-variation model and the measured porosity value, the input of the regressor is the predicted value of the vibration parameter change, and the output is the predicted curve of the vibration trend; when the ratio of the vibration frequency increment to the cutting force increment in the predicted curve is greater than a preset threshold, it is determined that the vibration parameters and the cutting state are positively correlated, and the predicted value of the vibration parameters in this state is recorded; the cubic spline interpolation method is used to fit the vibration trend predicted curve, and the slope of the fitted curve in the low-porosity area is calculated to obtain the change trend of the spindle vibration parameters.

[0021] Furthermore, for the feature extraction of the vibration parameters, the key feature quantities such as the vibration amplitude, the vibration frequency, and the change rate of the vibration parameters over time are extracted, and combined with the co-variation model, it is judged whether the vibration amplitude continuously increases and the vibration frequency tends to be stable. If so, a preset automatic suppression mechanism is triggered, including:

[0022] Vibration sensors are used to continuously collect vibration data sequences. The vibration data sequences are decomposed at multiple scales through wavelet transform. Vibration amplitude data, vibration frequency data, and the time change rate of parameters are extracted from the multi-scale decomposition results. The frequency variance within the time window is calculated based on the vibration frequency data. The Kalman filter is used to filter the vibration amplitude data, and the amplitude change rate within consecutive sampling periods is calculated from the filtered vibration amplitude. A vibration parameter predictor is constructed through a recurrent neural network. The inputs to the predictor include vibration frequency data, vibration amplitude data, and the time change rate of parameters, and the output is the trend of the vibration parameters over time. A co-variation discriminator is used to calculate the co-variation characteristics of the vibration frequency variance and the vibration amplitude change rate. When the vibration frequency variance is less than the stable determination threshold and the vibration amplitude change rate is greater than the growth determination threshold, the suppression controller is triggered. The suppression controller selects the corresponding spindle speed adjustment amount and feed rate adjustment amount according to the co-variation characteristics to form a suppression control command. The control command is executed to adjust the machining parameters, and a proportional-integral controller is used to achieve closed-loop control of the speed and feed rate. The output gain of the controller is calculated based on the vibration amplitude change rate. Real-time vibration data during the suppression process is collected from the vibration sensor. When the vibration amplitude change rate drops below the growth determination threshold, the current suppression control parameters are maintained unchanged.

[0023] Further, through a preset automatic suppression mechanism, the spindle speed and feed rate are adjusted to reduce the vibration amplitude and stabilize the vibration frequency. Meanwhile, the change of the cutting state is monitored in real time, vibration parameters are re-collected, and it is judged whether the vibration suppression effect meets the preset requirements. If the vibration amplitude drops below the threshold and the vibration frequency tends to be stable, the automatic suppression process is completed, including:

[0024] The rotational speed adjustment amount and the feed adjustment amount in the preset parameter mapping table are obtained by using a suppression parameter selector, and parameter adjustment instructions are sent to the spindle controller and the feed controller through the numerical control communication interface; the actual rotational speed and feed speed are calculated based on the position feedback signals of the spindle encoder and the feed encoder, and a proportional-integral controller is used to achieve closed-loop control, and the proportional coefficient and integral coefficient of the controller are automatically adjusted according to the vibration amplitude; the vibration frequency and vibration amplitude data during the adjustment process are collected by a vibration sensor, and the collected data is filtered by a Kalman filter, and the state noise covariance and measurement noise covariance of the filter are automatically updated according to the sampling period; a vibration feature extractor is constructed by using a recurrent neural network, the network input layer includes a vibration frequency sequence and a vibration amplitude sequence, the hidden layer uses long short-term memory units, and the output layer generates the vibration frequency stability and the vibration amplitude change rate; the cutting state data is obtained from a cutting force sensor, a mapping relationship between the vibration feature and the cutting force is established, and the stable interval of the vibration parameters is calculated according to the mapping relationship; when the vibration amplitude is lower than the preset amplitude threshold and the vibration frequency stability is higher than the preset stability threshold, the current machining parameters are kept unchanged, and the vibration parameters and cutting parameters in this state are recorded; if the vibration parameters exceed the stable interval, new rotational speed adjustment amount and feed adjustment amount are calculated according to the vibration feature, and the parameter adjustment process is repeated until the vibration parameters enter the stable interval.

[0025] The technical solutions provided in the embodiments of the present invention may include the following beneficial effects:

[0026] The present invention discloses a method for real-time monitoring and automatic suppression of spindle vibration of a numerically controlled machine tool. This method obtains the spindle vibration parameters and porosity distribution information during the machining process, divides the machining area into a surface loose bone structure area and a deep dense bone structure area, and establishes a mapping relationship between the vibration parameters and the porosity distribution. The present invention analyzes the change trend of the vibration parameters in different areas, judges whether there are non-linear fluctuations or co-variation characteristics, and establishes a co-variation model between the vibration parameters and the cutting state. When abnormal vibration is detected, the present invention triggers a preset automatic suppression mechanism to reduce the vibration amplitude and stabilize the vibration frequency by adjusting the spindle speed and feed rate. This method can effectively monitor and suppress abnormal vibration during the machining process, and improve the machining accuracy and quality of the titanium alloy artificial hip femoral stem. Description of the Drawings

[0027] Figure 1 It is a flowchart of a method for real-time monitoring and automatic suppression of spindle vibration of a numerically controlled machine tool according to the present invention.

[0028] Figure 2 It is a schematic diagram of a method for real-time monitoring and automatic suppression of spindle vibration of a numerically controlled machine tool according to the present invention. Detailed Embodiments

[0029] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative work belong to the scope of protection of the present invention.

[0030] As Figure 1-2 , a real-time monitoring and automatic suppression method for the spindle vibration of a numerically controlled machine tool in this embodiment may specifically include:

[0031] S101. When machining the bionic structure of the femoral stem of a titanium alloy artificial hip joint on a numerically controlled machine tool, the sending end collects the spindle vibration parameters in real time, including the vibration frequency and vibration amplitude. At the same time, the porosity distribution data of the machining area is obtained, and the correlation relationship between the vibration characteristics and the machining state is established based on these data to support subsequent analysis and suppression operations.

[0032] In the embodiment of the present invention, the sending end realizes the comprehensive collection of data through the collaborative work of multiple sensors. The acquisition of the spindle vibration parameters depends on high-precision sensors, while the extraction of the porosity distribution data combines the physical property analysis of the machining area. The specific operation method can be flexibly adjusted according to the actual machining scenario and will not be overly limited. For example, when machining transitions from the surface loose bone structure to the deep dense bone structure, the spindle vibration characteristics will show different response laws due to the change in the material porosity. The sending end needs to ensure the real-time and accuracy of the data to provide a reliable basis for subsequent vibration suppression.

[0033] S1011. The sending end uses the spindle speed sensor to collect the spindle rotation frequency data. At the same time, the vibration frequency and vibration amplitude are obtained through the vibration sensor, and the spindle torque data measured by the strain gauge and the spindle power parameter read by the numerical control system are combined to form a multi-dimensional data set for analyzing the spindle operating state.

[0034] In the embodiment of the present invention, the spindle speed sensor adopts the Hall effect principle, captures the spindle rotation pulse signal at fixed time intervals and converts it into a frequency value. For example, when the spindle speed is set to 3000 revolutions per minute, the sensor collects data at a period of 20 milliseconds, and the calculated rotation frequency is about 50 Hz. The vibration sensor selects a piezoelectric accelerometer, whose measurement range covers 0 to 500 Hz, and the sensitivity reaches 100 mV / g. During the machining process, the detected vibration frequency is about 156 Hz, and the amplitude reaches 0.15 mm. The strain gauge measures the torque through the Wheatstone bridge circuit, with a sensitivity of 2 mV / Nm, and the measured average torque value is 120 Nm. These data together constitute a multi-dimensional characterization of the spindle operating state, laying a foundation for subsequent analysis.

[0035] S1012. The sending end constructs a parameter correlation matrix based on the collected spindle rotation frequency, vibration frequency, and vibration amplitude data, generates a vibration parameter distribution map based on this matrix, and simultaneously uses an acoustic emission sensor to collect the acoustic emission signals during the machining process, and establishes a mapping relationship through the signal characteristics and the porosity distribution data to judge the machining state.

[0036] In the embodiment of the present invention, the parameter correlation matrix is generated by calculating the correlation coefficients between the spindle rotation frequency, vibration frequency, and vibration amplitude. When the correlation coefficient is close to 1, it indicates a high degree of coupling between the parameters. The generated vibration parameter distribution map intuitively reflects the changes in the vibration characteristics of the machining area, such as the transition area from the surface loose bone structure to the deep dense bone structure. The acoustic emission sensor uses a piezoelectric ceramic element with a frequency response range of 100 kHz to 1 MHz. The amplitude and energy characteristics of the collected signals are closely related to the cutting state of the tool. For example, when the feed rate is 200 mm per minute and the cutting depth is 0.5 mm, the root mean square value of the acoustic emission signal reaches 0.5 V, and the event count is 800 times per second. By comparing with the porosity distribution data, the changes in the material characteristics of the machining area can be further verified.

[0037] S1013. The sending end presets an acoustic emission signal threshold according to the characteristics of the titanium alloy material, and monitors in real time whether the acoustic emission signal exceeds this threshold range. If it exceeds, the tool feed rate and cutting depth are adjusted to restore the signal value within the threshold range, thereby initially stabilizing the machining state.

[0038] In the embodiment of the present invention, based on the machining characteristics of the titanium alloy TC4 material, the root mean square value threshold of the acoustic emission signal is preset to 0.8 V. When it is monitored that the root mean square value of the signal exceeds this threshold, for example, reaches 0.9 V, it indicates that the cutting state is abnormal, which may be caused by a local increase in porosity or an increase in tool wear. At this time, the sending end automatically reduces the feed rate from 200 mm per minute to 150 mm per minute, and the cutting depth from 0.5 mm to 0.3 mm until the signal value drops below 0.5 V. After adjustment, the porosity distribution in the machining area tends to be uniform, and the average value remains at about 3%, initially realizing the stability of the machining state. In the embodiment of the present invention, through the above steps, the sending end can comprehensively collect the spindle vibration parameters and porosity distribution data, and establish a mapping relationship between the two. This method makes full use of the advantages of multi-sensor collaborative monitoring, providing data support for subsequent vibration trend prediction and automatic suppression. At the same time, the strategy of real-time adjusting the cutting parameters effectively reduces the vibration interference during the machining process, improving the quality and consistency of the machined surface. The subsequent steps can further optimize the suppression mechanism based on these data, which will not be elaborated here.

[0039] S102. The delivery end divides the processing area of the bionic structure of the titanium alloy artificial hip femoral stem into a surface cancellous bone structure area and a deep dense bone structure area according to the porosity distribution information of the processing area, extracts the main axis vibration parameters of different areas respectively, and establishes the mapping relationship between them and the porosity distribution. At the same time, through cluster analysis and regression modeling, the regional boundary is divided and the vibration characteristics are quantified, providing a data basis for subsequent vibration suppression.

[0040] In the embodiment of the present invention, the delivery end first obtains the porosity distribution data of the processing area and uses these data to perform structural division on the processing area. The division process not only depends on the porosity itself but also combines the variation law of the vibration parameters to ensure the accuracy and practical processing significance of the regional division. The divided areas will be used to analyze the vibration characteristics of the surface and deep structures respectively, and provide a basis for the vibration suppression strategy through the mapping relationship. This step does not impose too many restrictions on the specific sensor type or algorithm details, and can be adjusted by technicians according to actual needs.

[0041] S1021. The delivery end processes the point cloud data by using the density clustering method through collecting the point cloud data set of the processing area, marks the point sets of the surface cancellous bone structure area and the deep dense bone structure area according to the density characteristics of the porosity distribution, and calculates the regional boundary based on the gradient method to complete the structural division. In the embodiment of the present invention, the delivery end uses a high-resolution scanning device to collect the point cloud data. For example, in a 10 mm × 10 mm processing area, 100,000 discrete points are generated at a sampling interval of 0.1 mm. The density clustering algorithm classifies the points with a porosity greater than 15% into the loose structure point set and the points with a porosity less than 5% into the dense structure point set by calculating the local density of each point and the distance between adjacent points. The boundary determination uses the gradient method to calculate the porosity change rate between adjacent points. When the change rate exceeds 0.1 per millimeter, it is determined as the regional boundary point. This method can effectively capture the transition characteristics of the porosity from high to low, and the measured transition zone width is about 2 mm, providing a spatial basis for subsequent parameter extraction.

[0042] S1022. For the surface cancellous bone structure area, the delivery end uses a vibration sensor to collect the vibration frequency and vibration amplitude data during the processing, normalizes these data to generate the surface vibration feature vector, and at the same time combines a depth detector to measure the bone structure depth, and constructs a mapping model between the vibration feature vector and the porosity distribution through the support vector regression method.

[0043] In the embodiments of the present invention, the vibration frequency range collected by the vibration sensor in the surface layer region is from 150 to 200 Hz, and the vibration amplitude is between 0.2 and 0.3 mm, reflecting the sensitivity of the low elastic modulus (about 5 GPa) of the loose structure to the cutting force fluctuation. Normalization maps the original data to the interval of 0 to 1, facilitating subsequent modeling. The depth detector uses ultrasonic technology with a transmission frequency of 20 MHz, and the measured depth change of the surface layer structure is relatively small, usually less than 5 mm. Support vector regression fits the relationship between the vibration frequency, amplitude, and porosity through a kernel function (such as the radial basis function). After model training, the correlation coefficient between the vibration amplitude and porosity is as high as 0.91. This mapping model can inversely deduce the porosity distribution based on the vibration data, providing auxiliary verification for regional division.

[0044] S1023. The sending end collects vibration parameters in the deep dense bone structure region, normalizes the vibration frequency and vibration amplitude to generate deep vibration feature vectors, and constructs a mapping function based on the surface layer and deep vibration feature vectors, and extracts the porosity threshold from it for regional division judgment.

[0045] In the embodiments of the present invention, the vibration frequency distribution in the deep layer region is from 80 to 120 Hz, and the vibration amplitude is between 0.05 and 0.1 mm, showing strong vibration suppression characteristics, which is closely related to the higher compressive strength (200 to 300 MPa) in this region. The normalized deep vibration feature vectors and surface layer vectors are jointly input into the mapping function, which can be constructed by polynomial fitting or neural network methods. The extracted porosity threshold usually ranges between 5% and 15%. For example, when the porosity of the detection point is higher than 10%, it is divided into the surface layer loose region; when it is lower than 10%, it is divided into the deep dense region. This threshold extraction method combines the physical meaning of the vibration parameters and the mechanical properties of the processing region, ensuring the reliability of the division result.

[0046] S1024. The sending end verifies the mapping relationship between the vibration parameters and the porosity distribution through periodic loading experiments, optimizes the regional division using the vibration gradient characteristics of the boundary region, and analyzes the differential characteristics of the vibration responses in different regions to support the effectiveness of the mapping model.

[0047] In the embodiment of the present invention, the delivery end applies a periodic load of 50 to 200 N to the machining area, records the vibration responses of the surface layer and the deep layer areas, and finds that the vibration amplitude of the surface layer area is 3 to 4 times that of the deep layer area, verifying the significance of the porosity on the vibration. The vibration frequency gradient in the boundary area is about 50 Hz per millimeter, and the vibration amplitude gradient is about 0.1 mm per millimeter. These transition characteristics are used to optimize the boundary determination rule. The effectiveness of the mapping model is further confirmed by experimental data. For example, when the vibration frequency exceeding 160 Hz and the amplitude greater than 0.25 mm are detected in the surface layer area, it is accurately identified as a loose structure, while when the frequency in the deep layer area is lower than 100 Hz and the amplitude is less than 0.08 mm, it is accurately classified as a dense structure. This differential analysis not only enhances the prediction ability of the model but also provides theoretical support for subsequent suppression strategies. In the embodiment of the present invention, the delivery end realizes the precise division of the machining area and the feature extraction of the vibration parameters through the above steps. The establishment of the mapping relationship makes full use of the advantages of density clustering and support vector regression, making the correlation between the porosity distribution and the vibration characteristics clearer. This method can adapt to the dynamic changes in the machining of titanium alloy materials, providing a solid data basis for real-time monitoring and suppression, and the application scenario of the model can be further expanded according to actual needs in the future.

[0048] S103. The delivery end analyzes the change trend of the spindle vibration parameters for the surface loose bone structure area, judges whether the vibration parameters show non-linear fluctuations due to the porosity change through time-frequency analysis and filtering processing, and immediately triggers the vibration suppression mechanism to stabilize the machining process when the vibration amplitude exceeds the preset safety threshold.

[0049] In the embodiment of the present invention, the delivery end identifies the dynamic change rule by collecting and processing the vibration data of the surface layer area in real time. The analysis process combines the characteristics of the frequency domain and the amplitude domain to ensure high accuracy in judging non-linear vibration. If abnormal fluctuations are detected, the suppression mechanism is immediately started to protect the machining quality. The implementation method of this step can be flexibly adjusted according to the actual situation of the machining equipment and will not be overly limited.

[0050] S1031. The sending end uses a vibration sensor to collect the vibration frequency and vibration amplitude data of the surface loose bone structure area, performs time-frequency analysis through Fourier transform to generate a vibration frequency spectrum, extracts the main frequency and harmonic components, and at the same time uses a Kalman filter to smooth the vibration amplitude data to generate a vibration amplitude curve. In the embodiment of the present invention, the vibration sensor operates at a sampling frequency of 10 kHz, continuously collects 10,000 data points in 1 second, covering the vibration characteristics of the surface area. Fourier transform converts the vibration frequency data into a frequency spectrum, and it is found that the main frequency is concentrated at 156 Hz, accompanied by harmonic components of 312 Hz and 468 Hz, reflecting the periodic characteristics of the vibration. The Kalman filter eliminates the influence of noise through iterative estimation, making the vibration amplitude curve smoother, and the measured amplitude fluctuates between 0.12 mm and 0.28 mm. This processing method effectively separates the main trend and noise in the vibration signal, providing a clear data basis for subsequent analysis.

[0051] S1032. The sending end calculates the vibration frequency change rate according to the vibration frequency spectrum, uses a recurrent neural network to construct a mapping model between the vibration frequency change rate and the porosity change, and combines the fluctuation range of the vibration amplitude curve to determine whether the vibration shows non-linear characteristics and exceeds the safety threshold.

[0052] In the embodiment of the present invention, the vibration frequency change rate is calculated through the dynamic changes of the main frequency and its harmonic components. In a 0.2-second detection window, the maximum change rate reaches 12 Hz per second, the minimum is -8 Hz per second, and the change range is 20 Hz per second. The recurrent neural network adopts a three-layer fully connected hidden layer structure. The input layer receives the frequency change rate data of 20 time steps, and the number of hidden layer neurons is 32, 16, and 8 in sequence. The output layer predicts the porosity change value. The trained model shows that the correlation between the frequency change rate and the porosity is as high as 0.85, and the porosity fluctuation range is from 12% to 18%. If the preset non-linear determination threshold is 15 Hz per second, the current change range exceeds the threshold, indicating that the vibration has non-linear characteristics. At the same time, the vibration amplitude fluctuation range is 0.16 mm, exceeding the preset safety threshold of 0.15 mm, and further processing is required.

[0053] S1033. When the sending end detects abnormal vibration, it dynamically adjusts the spindle speed and feed speed according to the suppression parameter mapping table through a proportional-integral controller until the vibration amplitude drops below the safety threshold, and at the same time monitors the porosity fluctuation in real time to verify the suppression effect.

[0054] In an embodiment of the present invention, the suppression parameter mapping table is pre-constructed based on the vibration frequency change rate and the amplitude fluctuation range. For example, when the frequency change rate is between 15 and 20 Hertz per second and the amplitude fluctuation is between 0.15 and 0.20 millimeters, it is recommended to reduce the spindle speed by 200 revolutions per minute and the feed rate by 50 millimeters per minute. The proportional-integral controller takes the amplitude deviation as the input, with a proportional coefficient of 2 and an integral coefficient of 0.5, and outputs an adjustment amount to reduce the rotational speed from 3000 revolutions per minute to 2800 revolutions per minute and adjust the feed rate from 200 millimeters per minute to 150 millimeters per minute. After adjustment, the frequency change rate drops below 10 Hertz per second, the amplitude fluctuation is reduced to 0.08 millimeters, and the change range of the porosity shrinks from 6% to 3%, indicating a significant improvement in machining stability. The entire adjustment process takes about 0.5 seconds, ensuring a rapid response.

[0055] S1034. The sending end verifies the non-linear characteristics of the vibration parameter trend through multi-period data, predicts the potential impact of porosity changes on vibration using the mapping model, and optimizes the triggering conditions of the suppression mechanism to adapt to different machining scenarios.

[0056] In an embodiment of the present invention, the sending end repeatedly collects data within multiple 0.2-second detection cycles and finds that the non-linear characteristics of the vibration frequency change rate are highly correlated with the porosity fluctuation, especially when the cutting force in the surface layer area is unstable. The mapping model predicts that when the porosity suddenly increases by more than 5%, the vibration amplitude may increase by more than 0.1 millimeter. Based on this, the safety threshold is optimized to 0.14 millimeters to trigger the suppression mechanism in advance. This predictive ability enhances the adaptability of the system to the dynamic machining environment and ensures the effectiveness of the suppression mechanism under different material characteristics.

[0057] In an embodiment of the present invention, the sending end achieves accurate judgment and timely suppression of the vibration trend in the surface layer area through the above analysis and adjustment steps. The identification of non-linear characteristics benefits from time-frequency analysis and neural network modeling, while the suppression process quickly stabilizes the vibration parameters through an accurate control strategy. The advantage of this method lies in taking into account both machining efficiency and quality, and the model parameters or control logic can be further adjusted according to actual needs to improve performance in the future.

[0058] S104. The delivery end extracts the vibration parameters of the deep dense bone structure area during continuous processing, determines whether there is a positive correlation between the vibration frequency and the cutting force by analyzing the co-variation characteristics with the cutting state, and records the parameter change law to construct a mapping table after confirming the correlation, providing a basis for subsequent vibration suppression. In the embodiment of the present invention, the delivery end collects the vibration and cutting data of the deep area through multi-sensor collaboration, and uses signal decomposition and time series modeling methods to reveal the dynamic association between the parameters. The analysis process aims to capture the stability of the processing state and its specific impact on vibration, ensuring that the recorded data can truly reflect the processing characteristics of the deep structure. The specific implementation of this step can be adjusted according to the equipment performance and processing requirements.

[0059] S1041. The delivery end uses a vibration sensor and a cutting force sensor to collect the vibration frequency and cutting force data of the deep dense bone structure area in real time, decomposes the vibration frequency signal through wavelet transform to extract the frequency component and amplitude component, and calculates the statistical characteristics of the cutting force to determine the stability of the processing state.

[0060] In the embodiment of the present invention, the vibration frequency range measured by the vibration sensor in the deep area is 80 to 120 Hz, and the cutting force recorded by the cutting force sensor fluctuates between 200 and 300 N. The wavelet transform adopts a 4-layer decomposition method. The high-frequency component highlights the instantaneous vibration characteristics, such as the peak signal when the tool cuts in, and the low-frequency component reflects the smooth change of the processing trend. The mean value of the cutting force data is 250 N, the standard deviation is 20 N, and the proportion is 8%, which is lower than the preset threshold of 10%, indicating that the processing is in a stable range. At this time, the mean value of the recorded vibration frequency is 100 Hz, and the amplitude is 0.08 mm, providing reliable data for subsequent analysis.

[0061] S1042. The delivery end analyzes the correlation between the vibration frequency and the cutting force using the Pearson correlation coefficient, combines the depth sensor to measure the density distribution of the deep bone structure, and constructs a time series mapping model of the cutting force and the vibration frequency through a long short-term memory network to extract the co-variation characteristic values.

[0062] In the embodiment of the present invention, the calculation result of the Pearson correlation coefficient is 0.82, indicating a strong positive correlation between the vibration frequency and the cutting force. When the depth sensor measures the depth from 10 to 15 mm through ultrasonic technology, the density increases from 88% to 96%, forming a density distribution curve, which reflects the gradual change characteristics of the deep structure.

[0063] The long short-term memory network takes the time series of cutting force and vibration frequency as inputs. The network contains two hidden layers, with 50 units in each layer, and outputs co-varying eigenvalue. For example, when the increment of cutting force is 50 Newtons and the increment of vibration frequency is 15 Hertz, the eigenvalue reaches 0.75, indicating significant co-variation between the two. This model can memorize long-term dependencies in the machining process and improve the accuracy of feature extraction.

[0064] S1043. The sending end determines the positive correlation state based on the tool wear data and the co-varying eigenvalue. If the ratio of the increment of vibration frequency to the increment of cutting force exceeds the preset threshold, relevant parameters are recorded and a cutting state parameter mapping table is constructed to quantify the dynamic law of the machining process.

[0065] In the embodiment of the present invention, the tool wear sensor detects that the wear amount increases from 0.1 mm to 0.3 mm. When the wear amount is 0.2 mm, the cutting force increases to 280 Newtons and the vibration frequency rises to 110 Hertz. The increment ratio is 0.35 Hertz per Newton, which is higher than the preset threshold of 0.3 Hertz per Newton, and it is determined as a strong positive correlation state. The parameters recorded at this time include a vibration frequency of 110 Hertz, an amplitude of 0.09 mm, a cutting force of 280 Newtons, and a wear amount of 0.25 mm. The mapping table further records the parameter combinations at different stages. For example, when the wear amount is 0.15 mm, the vibration frequency is 95 Hertz, the amplitude is 0.06 mm, and the cutting force is 220 Newtons; when the wear amount increases to 0.28 mm, the parameters become 115 Hertz, 0.1 mm, and 300 Newtons. This mapping relationship clearly shows the evolution path of the machining parameters.

[0066] S1044. The sending end verifies the cause of the stability of the vibration parameters through physical property tests in the deep region, analyzes the potential influence of cutting temperature and elastic modulus on vibration using multi-dimensional data, and optimizes the application range of the mapping table to adapt to complex machining conditions.

[0067] In the embodiment of the present invention, the test result of the elastic modulus in the deep region is 18 GPa, which is much higher than 5 GPa in the surface region, making the stress distribution more uniform and enhancing the vibration stability. Infrared temperature measurement shows that the cutting temperature is about 200 degrees Celsius, which is 150 degrees Celsius lower than that in the surface layer, reducing the interference of thermal deformation on vibration. After the mapping table is optimized in combination with these physical properties, it can still maintain the accuracy of parameter recording when the tool wear intensifies or the cutting depth changes. For example, when the depth increases to 15 mm and the density approaches 96%, the vibration amplitude is only 0.05 mm, reflecting the inhibitory effect of high density on vibration. This multi-dimensional analysis provides a wider adaptability for the application of the mapping table.

[0068] In the embodiment of the present invention, the delivery end comprehensively extracts and analyzes the co-variation characteristics of the vibration parameters and the cutting state in the deep region through the above steps. The construction of the mapping table not only quantifies the dynamic law of the machining process, but also lays a foundation for the subsequent prediction of the vibration trend and the formulation of the suppression strategy. The method makes full use of the signal processing capabilities of wavelet transform and long short-term memory network to ensure the depth and accuracy of data analysis, and the parameter dimension in the table can be further enriched according to the machining requirements in the future.

[0069] S105. The delivery end establishes a co-variation model based on the variation law of the vibration parameters in the deep dense bone structure region and combines the cutting state data, and derives the vibration trend of the spindle in the low-porosity region through multi-dimensional feature extraction and time series prediction methods, providing a scientific basis for the optimization of the machining process.

[0070] In the embodiment of the present invention, the delivery end collects the vibration and cutting parameters in the deep region through multiple sensors, and constructs a prediction model by using signal processing and machine learning techniques. The model aims to reveal the internal relationship between vibration and cutting state and predict the vibration behavior under low-porosity conditions. The implementation method of this step can be flexibly adjusted according to the machining equipment and material characteristics.

[0071] S1051. The delivery end uses a vibration sensor to collect the original vibration parameter data in the deep dense bone structure region, converts it to the standard range through range normalization processing, and at the same time combines the cutting state sensor and the ultrasonic porosity detector to obtain the cutting force and porosity data, laying a foundation for subsequent feature extraction.

[0072] In the embodiment of the present invention, the vibration sensor operates at a sampling frequency of 10 kHz, the collected vibration frequency range is between 80 and 120 Hz, and the amplitude is between 0.05 and 0.15 mm. Range normalization maps these data to the interval of 0 to 1, avoiding the interference of dimension differences on the analysis. The force values recorded by the cutting force sensor fluctuate between 220 and 280 N, and the porosity in the low-porosity region measured by the ultrasonic detector is stable at 3% to 5%. These multi-source data together constitute a comprehensive characterization of the machining state, ensuring the richness and reliability of the model input.

[0073] S1052. The delivery end performs short-time Fourier transform on the normalized vibration parameter data to generate a time-frequency feature matrix, extracts the vibration frequency and amplitude features, and constructs a multi-dimensional feature vector by fusing the spindle speed and cutting force data to capture the dynamic characteristics of the machining process.

[0074] In the embodiment of the present invention, the short-time Fourier transform uses a 256-point window and a 50% overlap rate. The generated time-frequency feature matrix shows that the main frequency component is concentrated around 100 Hz, and the amplitude spectrum ranges from 0.08 to 0.12 mm. The stable rotational speed recorded by the main shaft speed sensor is 3000 revolutions per minute, and the cutting force data reflects the real-time change of the machining load. The multi-dimensional feature vector integrates 8 frequency features, 8 amplitude features, as well as the rotational speed and cutting force values, totaling 18 dimensions. This high-dimensional representation method can comprehensively reflect the interactive influence between vibration and cutting state, providing solid data support for model training.

[0075] S1053. The transmitting end constructs a co-variation model of vibration parameters and cutting state through a recurrent neural network, and uses a Bayesian regressor to optimize the predicted output in combination with the measured porosity data to generate a vibration trend prediction curve to judge the correlation and change trend between vibration and cutting state.

[0076] In the embodiment of the present invention, the recurrent neural network is designed with a three-layer structure. The input layer with 18 neurons receives the multi-dimensional feature vector. The hidden layer contains 32 long short-term memory units to capture the time series dependence. The output layer with 4 neurons predicts the changes in vibration frequency and amplitude. The model training is based on 1000 sets of time series samples, with 20 time steps in each set. The Bayesian regressor optimizes the relationship between the predicted value and the measured porosity with a Gaussian prior distribution. The output curve shows that the vibration frequency fluctuation in the low porosity region is less than 5 Hz. For example, when the frequency increment is 10 Hz and the cutting force increment is 35 N, the ratio of 0.286 Hz / N exceeds the threshold of 0.25 Hz / N, indicating a significant positive correlation. At this time, the predicted frequency is 105 Hz and the amplitude is 0.09 mm, reflecting the prediction ability of the model.

[0077] S1054. The transmitting end uses the cubic spline interpolation method to smooth the vibration trend prediction curve, calculates the trend slope in the low porosity region, and evaluates the model accuracy through multiple rounds of verification to ensure the credibility of the prediction results in engineering applications.

[0078] In the embodiment of the present invention, the cubic spline interpolation uses 20 control points to fit the prediction curve, and the calculated slope in the low porosity region is 0.15 Hz / s, indicating that the vibration change is relatively gentle. The model verification shows that the average prediction error is controlled within 5%. When the porosity is lower than 4%, the frequency prediction error is less than 3%, and the amplitude prediction accuracy reaches 0.01 mm. This high accuracy stems from the deep fusion of multi-source data by the model and the accurate capture of machining physical characteristics, enabling the prediction results to effectively guide subsequent parameter adjustment.

[0079] In the embodiment of the present invention, the delivery end successfully establishes a co-variation model of deep-region vibration and cutting state through the above steps, and realizes accurate prediction of vibration trends. The advantage of the method lies in combining the powerful processing capabilities of time-frequency analysis and neural networks, ensuring the controllability of the machining process in low-porosity regions. Subsequently, the model can be further extended to adapt to more complex machining scenarios.

[0080] S106. The delivery end extracts features from the vibration parameters, obtains the key feature quantities of vibration amplitude, frequency, and change rate, and analyzes the vibration trend in combination with the co-variation model. If it is found that the amplitude continues to increase while the frequency tends to be stable, an automatic suppression mechanism is triggered to optimize the machining stability.

[0081] In the embodiment of the present invention, the delivery end comprehensively extracts vibration features and judges their dynamic behavior through multi-scale signal processing and prediction models. The triggering of the suppression mechanism is based on the co-variation of the feature quantities, ensuring the timely control of vibration during the machining process. The implementation details of this step can be flexibly optimized according to the actual machining conditions.

[0082] S1061. The delivery end uses a vibration sensor to collect a vibration data sequence, performs multi-scale decomposition using wavelet transform to extract vibration amplitude and frequency data, and calculates its change rate after smoothing the amplitude data through a Kalman filter, providing accurate input for subsequent trend analysis.

[0083] In the embodiment of the present invention, the vibration sensor operates at a sampling frequency of 10 kHz, and 1000 data points are collected per cycle. The wavelet transform selects the db4 basis function and decomposes to 3 layers. The high-frequency coefficients reflect the amplitude change, ranging from 0.05 to 0.15 mm, and the low-frequency coefficients extract the frequency features, fluctuating between 80 and 120 Hz. The Kalman filter sets the state noise covariance to 0.01 and the measurement noise to 0.1. The filtered amplitude curve is smoother, and the change rate within a 200-ms window calculated is 0.002 mm / s. This processing method effectively reduces noise interference and improves the accuracy of feature extraction.

[0084] S1062. The delivery end constructs a vibration parameter predictor through a recurrent neural network, inputs the vibration frequency, amplitude, and change rate, predicts the future trend, and judges whether to trigger the suppression controller in combination with the frequency variance and amplitude change rate, and generates a suppression instruction according to the prediction result.

[0085] In an embodiment of the present invention, the recurrent neural network includes three layers. The input layer receives three features: frequency, amplitude, and rate of change. The hidden layer with 32 recurrent units captures the temporal relationship, and the output layer predicts the vibration trend at the next 5 time points. The frequency variance within a 200-ms window is 25 Hz², which is lower than the stable threshold of 30 Hz², and the amplitude rate of change is 0.0015 mm / s, which is higher than the growth threshold of 0.001 mm / s, indicating that the amplitude is increasing and the frequency is stable. The prediction result shows that the amplitude continues to rise, triggering the suppression controller. The controller looks up the table to determine that the rotational speed adjustment amount is -200 rpm and the feed rate adjustment amount is -50 mm / min, and generates and sends commands through the numerical control system.

[0086] S1063. The sending end executes the suppression instruction, uses a proportional-integral controller to achieve closed-loop adjustment of the rotational speed and the feed rate, and monitors the vibration data in real time to verify the suppression effect, and dynamically adjusts the control parameters to adapt to the changes in the machining conditions.

[0087] In an embodiment of the present invention, the proportional coefficient of the proportional-integral controller is set to 2.5, the integral coefficient is 0.5, and the output gain is linearly adjusted according to the rate of change. For example, when the rate of change is 0.0015 mm / s, the gain is 1.5. The actuator reduces the rotational speed from 3000 rpm to 2800 rpm and the feed rate from 200 mm / min to 150 mm / min. The real-time data shows that after suppression, the rate of change drops below 0.0008 mm / s, taking about 500 ms, and the frequency variance remains within 20 Hz². Tests show that when the workpiece hardness increases from 35 HRC to 45 HRC or the tool wear reaches 0.3 mm, the controller can still adaptively adjust the parameters to ensure stable suppression effect.

[0088] S1064. The sending end verifies and optimizes the suppression mechanism through multiple working conditions, analyzes the coupling relationship between vibration and machining state using co-varying features, improves the adaptability of the control strategy to complex environments, and ensures machining efficiency.

[0089] In an embodiment of the present invention, the sending end repeatedly tests under different hardness and tool wear conditions and finds that the co-varying features of the amplitude rate of change and the frequency variance can accurately reflect the changes in the machining state. For example, when the hardness increases, the amplitude rate of change briefly rises to 0.002 mm / s, and the controller quickly adjusts the rotational speed and the feed rate to bring them back below the threshold. This adaptive ability stems from the comprehensiveness of feature extraction and the robustness of the prediction model, ensuring the dual optimization of machining efficiency and vibration.

[0090] In the embodiment of the present invention, the delivery end realizes the accurate judgment of the vibration trend through feature extraction and prediction analysis, and effectively suppresses abnormal vibration through closed-loop control. The advantage of the method lies in combining the high efficiency of wavelet transform and neural network, being able to quickly respond in a dynamic processing environment, and the control parameters can be further refined according to actual needs in the future to expand the applicable range.

[0091] S107. The delivery end adjusts the spindle speed and feed rate through a preset automatic suppression mechanism to reduce the vibration amplitude and stabilize the frequency. Meanwhile, it monitors the cutting state in real time and re-collects vibration parameters, and judges whether the suppression effect meets the expectation through feature extraction and mapping analysis. If the amplitude drops below the threshold and the frequency is stable, the suppression process is completed.

[0092] In the embodiment of the present invention, the delivery end realizes the dynamic suppression of vibration by using multi-sensor data and neural network technology. The suppression process not only focuses on the real-time changes of vibration parameters, but also combines the feedback of the cutting state to ensure the continuous optimization of machining quality. The implementation method of this step can be flexibly adjusted according to the actual working conditions.

[0093] S1071. The delivery end uses a suppression parameter selector to obtain the speed and feed adjustment amounts from a preset mapping table, sends instructions to the spindle and feed controllers through a numerical control communication interface, and uses a proportional-integral controller to achieve closed-loop adjustment according to the feedback signal.

[0094] In the embodiment of the present invention, the mapping table is constructed based on experimental data. When the vibration amplitude is between 0.1 and 0.15 mm, the recommended speed adjustment amount is -200 revolutions per minute, and the feed adjustment amount is -50 mm per minute. The instructions are sent through an industrial bus with a cycle of 100 milliseconds. The spindle encoder resolution is 2500 pulses per revolution, and the feed encoder is 0.1 micron per pulse, providing high-precision feedback. The initial proportional coefficient of the proportional-integral controller is 2.5, and the integral coefficient is 0.5. When the amplitude exceeds 0.12 mm, the proportional coefficient increases to 3.0, and the integral coefficient increases to 0.8 to ensure the rapidity and stability of the adjustment.

[0095] S1072. The delivery end collects vibration data during the adjustment process through a vibration sensor, smooths it using a Kalman filter and then inputs it into a recurrent neural network to extract features, and establishes a mapping relationship between vibration and cutting state in combination with cutting force data to determine the stable interval.

[0096] In the embodiment of the present invention, the vibration sensor collects data at a sampling frequency of 10 kHz. The initial value of the state noise covariance of the Kalman filter is 0.01 and is adjusted to a maximum of 0.05 with data fluctuations. The measurement noise covariance varies from 0.1 to 0.2. The input layer of the recurrent neural network contains 20 neurons, receiving the vibration sequence in a 200-ms window. The hidden layer has 32 long short-term memory units, and the forgetting gate threshold is 0.6. The output is the frequency stability and the amplitude change rate. The cutting force sensor measures the force value between 200 and 300 N. The mapping relationship is generated by cubic spline interpolation with an interpolation interval of 50 N. The stable interval is 90 to 110 Hz in frequency and 0.05 to 0.08 mm in amplitude.

[0097] S1073. The sending end determines the suppression effect based on the vibration characteristics and the stable interval. If the amplitude is lower than 0.1 mm and the frequency stability exceeds 0.8, the parameters remain unchanged; otherwise, the adjustment amount is recalculated and the suppression process is repeated. At the same time, the adaptability of the mechanism to hardness and tool wear is verified.

[0098] In the embodiment of the present invention, when it is detected that the amplitude drops to 0.08 mm and the stability reaches 0.85, the rotational speed is maintained at 2800 revolutions per minute, the feed rate is 150 mm per minute, and the cutting force is stable at 250 N. If the amplitude suddenly increases to 0.13 mm, the system calculates a new adjustment amount, the rotational speed is reduced by another 250 revolutions per minute, and the feed is reduced by 60 mm per minute. It returns to stability after 3 to 4 cycles. Tests show that when the hardness increases from 35 HRC to 45 HRC or the tool wear reaches 0.3 mm, the controller adaptively adjusts the coefficient to maintain the suppression effect, demonstrating strong robustness.

[0099] S1074. The sending end optimizes the suppression mechanism through real-time monitoring and repeated adjustment, and uses multi-dimensional data analysis to analyze the potential impact of cutting state changes on vibration, ensuring the continuous stability and high efficiency of machining parameters in a dynamic environment.

[0100] In the embodiment of the present invention, the sending end continuously collects vibration and cutting data and finds that when the hardness increases, the amplitude briefly rises. The controller quickly responds to adjust the parameters to bring it back below the threshold. When the tool wear intensifies, the mapping relationship shows that the frequency stable interval shifts slightly. The system adapts to the change by increasing the adjustment amount. The whole process converges quickly, and the response time is about 300 to 500 ms. This dynamic optimization ability ensures the dual goals of machining efficiency and vibration control, and the content of the mapping table can be further improved according to the working conditions later.

[0101] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and supplements can be made, and these improvements and supplements should also be regarded as the protection scope of the present invention.

Claims

1. A method for real-time monitoring and automatic suppression of spindle vibration of a CNC machine tool, characterized in that: The method comprises: Obtain the spindle vibration parameters of the titanium alloy artificial hip joint femoral stem bionic structure during the CNC machine tool processing process. The vibration parameters include vibration frequency and vibration amplitude, and the porosity distribution information of the processing area is collected at the same time; According to the porosity distribution information, the processing area is divided into the surface loose bone structure area and the deep dense bone structure area. The vibration parameters of different areas during processing are extracted respectively, and the mapping relationship between vibration parameters and porosity distribution is established. Analyze the vibration parameter change trend of the surface porous bone structure area, and determine whether the vibration parameter shows nonlinear fluctuation with the change of porosity according to the vibration parameter change trend. If so, and the vibration amplitude exceeds the preset safety threshold, the preset vibration suppression mechanism is immediately triggered; Extract the vibration parameters of the deep dense bone structure area during continuous machining, and determine whether the vibration parameters show coordinated change characteristics with the cutting state. If the vibration frequency is positively correlated with the cutting force, then record the change pattern of the vibration parameters under this cutting state. According to the variation law of vibration parameters in the deep dense bone structure area, a coordinated variation model of vibration parameters and cutting state is established to predict the vibration trend of the spindle when machining in the low porosity area. Extract the characteristics of vibration parameters, extract the key characteristics of vibration amplitude, vibration frequency and the change rate of vibration parameters over time, and combine with the coordinated change model to determine whether the vibration amplitude continues to increase and the vibration frequency tends to be stable. If so, trigger the preset automatic suppression mechanism; Through the preset automatic suppression mechanism, the spindle speed and feed rate are adjusted to reduce the vibration amplitude and stabilize the vibration frequency. At the same time, the cutting state changes are monitored in real time, and the vibration parameters are re-collected to determine whether the vibration suppression effect meets the preset requirements. If the vibration amplitude drops below the threshold and the vibration frequency tends to be stable, the automatic suppression process is completed.

2. The method according to claim 1, characterized in that The method of obtaining the spindle vibration parameters of the titanium alloy artificial hip joint femoral stem bionic structure during the CNC machine tool processing, the vibration parameters including the vibration frequency and the vibration amplitude, and collecting the porosity distribution information of the processing area at the same time, includes: Collecting spindle rotation frequency data obtained by a spindle speed sensor and vibration frequency data obtained by a vibration sensor, and establishing a parameter association matrix according to the spindle rotation frequency data and the vibration frequency data; Establishing a vibration parameter distribution map according to the parameter association matrix, wherein the vibration parameter distribution map is used to obtain porosity distribution data of the processing area; Acquiring an acoustic emission signal through an acoustic emission sensor, and establishing a mapping relationship between the acoustic emission signal and the porosity distribution data of the processing area; It is determined whether the acoustic emission signal exceeds a preset threshold range. If the acoustic emission signal exceeds the preset threshold range, the tool feed speed and cutting depth are adjusted until the acoustic emission signal falls within the preset threshold range.

3. The method according to claim 1, characterized in that According to the porosity distribution information, the processing area is divided into a surface loose bone structure area and a deep dense bone structure area, vibration parameters in the processing of different areas are extracted respectively, and a mapping relationship between the vibration parameters and the porosity distribution is established, including: Density clustering is performed on the point cloud data set. Point sets in loose structure areas and dense structure areas are marked according to the porosity distribution density. The boundary calibration result is obtained by using the regional boundary determination rule. The width of the structural transition zone between the surface loose bone structure area and the deep dense bone structure area is calculated according to the boundary calibration result, the vibration parameters within the surface area are collected by a vibration sensor, and the vibration parameters are normalized to obtain the surface vibration feature vector; The depth parameters of the bone structure in the point cloud data set are measured by a depth detector, a mapping relationship between the vibration characteristic vector and the porosity is established by using support vector regression, and the vibration parameters in the deep region are normalized to obtain a deep vibration characteristic vector; A mapping function is constructed based on the surface vibration eigenvector and the deep vibration eigenvector, and a porosity threshold is extracted from the mapping function. If the porosity of the detection point is higher than the porosity threshold, it is divided into the surface loose bone structure area; if the porosity of the detection point is lower than the porosity threshold, it is divided into the deep dense bone structure area.

4. The method according to claim 1, characterized in that: The vibration parameter variation trend of the surface porous bone structure area is analyzed, and whether the vibration parameter presents nonlinear fluctuation with the change of porosity is judged according to the vibration parameter variation trend. If so, and the vibration amplitude exceeds a preset safety threshold, a preset vibration suppression mechanism is immediately triggered, including: A vibration sensor is used to obtain vibration frequency data and vibration amplitude data of the surface porous bone structure area, and a time-frequency analysis is performed on the vibration frequency data through Fourier transform to obtain a vibration frequency spectrum; Calculating the vibration frequency change rate according to the main frequency component and the harmonic component in the vibration frequency spectrum, and filtering the vibration amplitude data using a Kalman filter to obtain a vibration amplitude curve; A recursive neural network mapping model is established for the vibration frequency change rate sequence, wherein the recursive neural network comprises an input layer, three fully connected hidden layers and an output layer, and a porosity change value is obtained from the mapping model; If the vibration frequency change rate and the vibration amplitude fluctuation range exceed the preset safety threshold, the spindle speed and feed speed are dynamically adjusted according to the proportional-integral controller until the vibration amplitude drops below the safety threshold.

5. The method according to claim 1, characterized in that The method extracts vibration parameters of the deep dense bone structure area during continuous processing, determines whether the vibration parameters show a coordinated change characteristic with the cutting state, and if the vibration frequency is positively correlated with the cutting force, records the change law of the vibration parameters under the cutting state, including: A vibration sensor and a cutting force sensor are used to obtain vibration frequency data and cutting force data, and the vibration frequency data are decomposed by wavelet transform to obtain vibration frequency components and vibration amplitude components; Calculating the mean and standard deviation of the cutting force according to the cutting force data, and if the standard deviation is less than a preset ratio of the mean, determining that the cutting state is in a stable interval, and recording the vibration frequency component and the vibration amplitude component; Calculating a Pearson correlation coefficient based on the vibration frequency component and the cutting force data, and obtaining a vibration parameter fluctuation value of a deep bone structure from the correlation coefficient; A long short-term memory network is used to process the cutting force data and the vibration frequency data to obtain a coordinated change characteristic value. If the coordinated change characteristic value is greater than a preset threshold, the vibration frequency component and the vibration amplitude component and the cutting force data are recorded to construct a cutting state parameter mapping table.

6. The method according to claim 1, characterized in that According to the vibration parameter variation law of the deep dense bone structure area, a coordinated variation model of vibration parameters and cutting state is established to predict the vibration trend of the spindle when processing in the low porosity area, including: A vibration sensor is used to obtain original vibration parameter data, and the original vibration parameter data is used to obtain normalized vibration parameter data through a range normalization method; Calculating a short-time Fourier transform according to the normalized vibration parameter data to obtain a time-frequency feature matrix, and extracting vibration frequency features and vibration amplitude features from the time-frequency feature matrix; Obtaining speed data and cutting force data through a spindle speed sensor and a cutting force sensor, and constructing a multidimensional feature vector according to the vibration frequency characteristics, vibration amplitude characteristics, speed data and cutting force data; A synergistic variation model is established by using a recurrent neural network. The input layer of the synergistic variation model is the multidimensional feature vector. The output value of the synergistic variation model and the measured porosity value are used to train a Bayesian regressor, and the Bayesian regressor outputs a vibration trend prediction curve.

7. The method according to claim 1, characterized in that The feature extraction of the vibration parameters is performed to extract the key feature quantities of the vibration amplitude, vibration frequency and the change rate of the vibration parameters over time, and combined with the coordinated change model, it is determined whether the vibration amplitude continues to increase and the vibration frequency tends to be stable. If so, the preset automatic suppression mechanism is triggered, including: A vibration sensor is used to obtain a vibration data sequence, and the vibration data sequence is decomposed into multiple scales by wavelet transform to obtain vibration amplitude data and vibration frequency data; Calculating the frequency variance within the time window according to the vibration frequency data, filtering the vibration amplitude data using a Kalman filter, and calculating the amplitude change rate from the filtered vibration amplitude data; Processing the vibration frequency data, the vibration amplitude data and the amplitude change rate through a recurrent neural network to obtain a vibration parameter prediction result; If the frequency variance is less than the stability judgment threshold and the amplitude change rate is greater than the growth judgment threshold, the suppression controller is triggered. The suppression controller selects the spindle speed adjustment amount and the feed speed adjustment amount according to the vibration parameter prediction result to generate a suppression control instruction.

8. The method according to claim 1, characterized in that The preset automatic suppression mechanism is used to adjust the spindle speed and feed rate, reduce the vibration amplitude and stabilize the vibration frequency, monitor the cutting state changes in real time, re-collect vibration parameters, and determine whether the vibration suppression effect meets the preset requirements. If the vibration amplitude drops below the threshold and the vibration frequency tends to be stable, the automatic suppression process is completed, including: A vibration sensor is used to obtain vibration frequency data and vibration amplitude data during the cutting process, and the vibration data is filtered by a Kalman filter to obtain filtered vibration data; Extracting vibration features through a recursive neural network according to the filtered vibration data, wherein the recursive neural network outputs the vibration frequency stability and the vibration amplitude change rate; The cutting force data in the cutting process are collected by a cutting force sensor, and a mapping relationship is established between the vibration characteristics and the cutting force data to obtain a vibration parameter stability interval; If the vibration frequency stability is lower than a preset stability threshold, a speed adjustment amount and a feed adjustment amount are calculated according to the vibration parameters, and the adjustment amounts are sent to a spindle controller and a feed controller via a numerical control communication interface.

Images