A method and system for controlling the cutting force of a cutter for profile tooth machining

Abstract

The application discloses a cutter cutting force control method and system for special-shaped tooth machining, and relates to the technical field of cutting force control, and comprises the following steps: acquiring the machining parameters of a special-shaped tooth workpiece, performing cutting analysis according to cutter geometric characteristic parameters combined with the machining parameter set to determine an initial cutting force threshold range; collecting dynamic cutting force data of the cutter in real time to generate a real-time cutting force frequency spectrum; dynamically comparing the real-time cutting force frequency spectrum according to the initial cutting force threshold range, extracting cutting force fluctuation characteristics for cutter wear analysis, obtaining a wear compensation coefficient, performing real-time correction of the cutting force of the special-shaped tooth workpiece, and generating a cutter cutting force control suggestion. The application solves the technical problems that the prior art cannot realize real-time sensing and effective control of the cutting force in the special-shaped tooth machining process, leading to fast cutter wear and low machining precision, and achieves the technical effects of realizing dynamic regulation and control of the cutter cutting force, prolonging the service life of the cutter, and improving the machining stability and precision of the special-shaped tooth.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cutting force control, in particular to a tool cutting force control method and system for profile tooth machining. BACKGROUND

[0002] In the precision machining process of profile tooth parts, due to the complex tooth shape structure and variable stress state, local overload and stress concentration may easily occur in the cutting process of the tool, resulting in significant cutting force fluctuation, which further aggravates tool wear and affects machining precision. Traditional machining strategies mostly rely on fixed process parameter setting, lack dynamic perception and feedback adjustment of actual cutting state, and are difficult to adapt to the changing stress characteristics and machining stages in profile tooth machining, which adversely affects machining stability and tool life. SUMMARY

[0003] The present application provides a tool cutting force control method and system for profile tooth machining, which is used to solve the technical problem that the prior art cannot realize real-time perception and effective control of the cutting force in profile tooth machining, resulting in fast tool wear and low machining precision.

[0004] In view of the above problems, the present application provides a tool cutting force control method and system for profile tooth machining.

[0005] In a first aspect of the present application, a tool cutting force control method for profile tooth machining is provided, which comprises:

[0006] Obtaining the machining parameters of the profile tooth workpiece, performing cutting analysis according to the tool geometric feature parameters combined with the machining parameter set to determine the initial cutting force threshold range; real-time acquisition of dynamic cutting force data of the tool to generate real-time cutting force spectrum; dynamic comparison of the real-time cutting force spectrum according to the initial cutting force threshold range, extraction of cutting force fluctuation characteristics for tool wear analysis to obtain a wear compensation coefficient; real-time correction of the cutting force of the profile tooth workpiece according to the wear compensation coefficient to generate tool cutting force control suggestions.

[0007] In a second aspect of the present application, a tool cutting force control system for profile tooth machining is provided, which comprises:

[0008] The cutting analysis module is used for acquiring machining parameters of the profile tooth workpiece, performing cutting analysis according to tool geometric characteristic parameters in combination with the machining parameter set, and determining an initial cutting force threshold range; the real-time data acquisition module is used for acquiring dynamic cutting force data of the tool in real time, and generating a real-time cutting force spectrum; the wear analysis module is used for dynamically comparing the real-time cutting force spectrum according to the initial cutting force threshold range, extracting cutting force fluctuation characteristics for tool wear analysis, and obtaining a wear compensation coefficient; and the real-time correction module is used for performing real-time correction of the cutting force of the profile tooth workpiece according to the wear compensation coefficient, and generating a tool cutting force control suggestion.

[0009] One or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0010] The present application acquires machining parameters of the profile tooth workpiece, performs cutting analysis according to tool geometric characteristic parameters in combination with the machining parameter set, and determines an initial cutting force threshold range; real-time acquisition of dynamic cutting force data of the tool generates a real-time cutting force spectrum; the real-time cutting force spectrum is dynamically compared according to the initial cutting force threshold range, cutting force fluctuation characteristics are extracted for tool wear analysis, and a wear compensation coefficient is obtained; the profile tooth workpiece is corrected in real time according to the wear compensation coefficient, and a tool cutting force control suggestion is generated. The present application solves the technical problem that the existing technology cannot realize real-time sensing and effective control of the cutting force during profile tooth machining, resulting in fast tool wear and low machining precision. Through cutting force modeling by combining tool geometric characteristics and machining parameters, real-time acquisition of cutting force data and dynamic comparison and analysis, calculation of a wear compensation coefficient and implementation of closed-loop correction control, dynamic regulation and control of tool cutting force are achieved, tool life is prolonged, and profile tooth machining stability and precision are improved. BRIEF DESCRIPTION OF DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort.

[0012] Figure 1 A tool cutting force control method flowchart for profile tooth machining provided by the embodiments of the present application is shown in the figure.

[0013] Figure 2 A tool cutting force control system structure diagram for profile tooth machining provided by the embodiments of the present application is shown in the figure.

[0014] Legend of the drawings: cutting analysis module 11, real-time data acquisition module 12, wear analysis module 13, real-time correction module 14. DETAILED DESCRIPTION

[0015] The present application provides a tool cutting force control method and system for special-shaped tooth machining, aiming to solve the technical problems that the prior art cannot realize real-time sensing and effective control of the cutting force in the special-shaped tooth machining process, resulting in fast tool wear and low machining precision. Through cutting force modeling combined with tool geometric characteristics and machining parameters, real-time acquisition of cutting force data and dynamic comparison and analysis, calculation of wear compensation coefficients and implementation of closed-loop correction control, the technical effects of realizing dynamic regulation and control of tool cutting force, prolonging tool life and improving the stability and precision of special-shaped tooth machining are achieved.

[0016] The technical solutions in the embodiments of the present application will be clearly and completely described in connection with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0017] It should be noted that any variation of the terms "comprise" and "have" is intended to cover non-exclusive inclusion, for example, a process, method, system, product or server comprising a series of steps or units does not have to be limited to those clearly listed steps or units, but can include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices.

[0018] Embodiment one, as shown in the present application provides a tool cutting force control method for special-shaped tooth machining, which comprises: Figure 1

[0019] Step S100: Obtain the machining parameters of the special-shaped tooth workpiece, perform cutting analysis according to the tool geometric characteristic parameters combined with the machining parameter set, and determine the initial cutting force threshold range.

[0020] In the embodiments of the present application, first, the machining parameters of the special-shaped tooth workpiece are obtained, which are pre-determined basic data including tooth shape size, material type, machining path, cutting depth and feed speed, etc.

[0021] Then, by analyzing the preset special-shaped tooth machining parameters, the key tooth shape characteristic parameters are extracted, and the cutting prediction is performed combined with the tool geometric characteristics to calculate the theoretical cutting force data. Then, the actual cutting force data is collected by a plurality of sensors, and the theoretical value is simulated and corrected to determine the initial cutting force threshold range that meets the actual working condition.

[0022] ​Further, the method provided by the application embodiment further comprises:

[0023] The tooth profile of the profiled tooth workpiece is analyzed based on the machining parameters to obtain tooth profile characteristic parameters; cutting prediction is performed according to the tooth profile characteristic parameters and the tool geometric characteristic parameters, theoretical cutting force data is calculated; actual cutting force data is dynamically collected by a plurality of sensors, the theoretical cutting force data is simulated and corrected based on the actual cutting force data, and the initial cutting force threshold range is determined.

[0024] In the application embodiment, firstly, the spatial structure of the profiled tooth workpiece is analyzed by using a three-dimensional curved surface reconstruction method. The three-dimensional curved surface reconstruction method takes the addendum circle diameter, the dedendum circle radius, the pitch, the thickness, the modulus and the like in the machining parameters as inputs, fits the tooth profile of the workpiece by calling a NURBS modeling function in a CAD platform, and obtains a continuous spatial tooth surface expression model. On this basis, tooth profile characteristic parameters including the tooth surface principal curvature and the secondary curvature, the dedendum transition curvature radius and the profile change rate are extracted.

[0025] Subsequently, the extracted tooth profile characteristic parameters are combined with the geometric characteristic parameters (such as the rake angle, the relief angle, the cutting edge length, the helix angle and the tool tip corner radius) of the tool, the Kienzle empirical cutting force model is called to perform theoretical stress prediction. The Kienzle empirical cutting force model is based on the product relationship of the cutting force constant and the cutting thickness and width, and calculates the force value required for unit cutting volume. The cutting angle and the feed path at different tooth surface positions are combined to calculate the spindle direction force, the normal force and the radial force at the positions respectively, and the theoretical cutting force data covering the entire tooth profile is generated.

[0026] In the actual machining process, the three-axis strain gauge sensor installed on the tool clamping part and the piezoelectric vibration sensor arranged in the workpiece clamping area are used to synchronously collect multi-directional real-time cutting force data at a sampling frequency of 2 kHz or more, and the actual cutting force data containing a time tag and an axial force component are obtained. The strain gauge records the stress response in the X, Y and Z directions, and the vibration sensor captures high-frequency disturbances, effectively reflecting the dynamic change characteristics in the machining state.

[0027] Then, the theoretical cutting force data and the actually collected data are one-to-one corresponding, the difference between the two is calculated point by point, and the average value of all the differences in each tooth segment is calculated to obtain the average deviation of the tooth segment. Subsequently, the average deviation is respectively extended upwards and downwards with the theoretical cutting force value as the center to form the corresponding upper and lower boundaries. For example, the theoretical value of a tooth segment is 180N, and the average deviation is 15N, so the system determines that the cutting force threshold range of the segment is 165N to 195N. In this way, the initial cutting force threshold range conforming to the actual machining is constructed.

[0028] Step S200: Real-time acquisition of dynamic cutting force data of the tool, to generate a real-time cutting force spectrum.

[0029] In the embodiment of the present application, by arranging a three-way strain sensor array on the tool, real-time acquisition of dynamic cutting force data generated between the tool and the contact area of the profiled tooth workpiece is performed, and multiple axial cutting force components are obtained. Meanwhile, a vibration sensor is arranged on the workpiece to collect high-frequency vibration signals generated during machining. The two types of signals are fused in time and frequency domains, a cutting force time-frequency feature matrix containing time and frequency double information is constructed, and the matrix is decomposed according to a preset frequency band, the energy distribution in the corresponding frequency band is extracted, and finally a real-time cutting force spectrum reflecting the change of the stress state of the tool is drawn.

[0030] Further, in the method provided by the application, the real-time acquisition of the dynamic cutting force data of the tool and the generation of the real-time cutting force spectrum further comprise:

[0031] A three-way strain sensor array is arranged on the tool, the three-way strain sensor array is used for sensing and collecting the tool, dynamic cutting force data is obtained, and the dynamic cutting force data contains multiple axial cutting force components; a vibration sensor is arranged on the profiled tooth workpiece, the vibration sensor is used for machining sensing, and high-frequency vibration signals are obtained; the multiple axial cutting force components and the high-frequency vibration signals are fused in time and frequency domains, a cutting force time-frequency feature matrix is constructed; the cutting force time-frequency feature matrix is decomposed according to a preset frequency band, energy distribution data of the preset frequency band is extracted, and the real-time cutting force spectrum is drawn according to the energy distribution data.

[0032] In the embodiment of the present application, first, a three-way strain sensor array is uniformly arranged at the circumferential position of the tool shank, each group of sensors has X, Y and Z direction strain detection capabilities, and multiple sensors form a ring-shaped force measuring structure with an interval angle of 45°±2°, so as to realize comprehensive monitoring of the axial stress of the tool. Based on the strain resistance principle, these sensors detect the micro deformation of the tool caused by external force during machining, and then obtain dynamic cutting force data, which includes the cutting force components of the tool and the contact area of the profiled tooth workpiece in X / Y / Z three directions.

[0033] Meanwhile, a laser Doppler vibrometer is arranged on the clamp structure of the profiled tooth workpiece, the focusing direction of the laser beam of the laser Doppler vibrometer is at an angle of 30° with the tool feeding direction, and the laser Doppler vibrometer can non-contact capture the micron-level vibration displacement of the workpiece surface. The laser Doppler vibrometer relies on the laser frequency shift principle to record the dynamic response behavior of the workpiece in real time when the workpiece is disturbed by cutting, and is suitable for identifying high-frequency vibration fluctuations caused by asymmetric tooth shape, slight tool wear or material unevenness, so as to obtain high-frequency vibration signals in the machining state.

[0034] Subsequently, the plurality of axial cutting force components are fused with the high-frequency vibration signal in time-frequency domain, and the original time sequence signal is processed by using short-time Fourier transform. Specifically, each group of original signals is divided into equal-length time windows, and fast Fourier transform (FFT) is performed in each window, so that the continuous time signal is mapped to the corresponding frequency space. In this way, the frequency change information is obtained under the premise of ensuring the time resolution, and finally a three-dimensional structure composed of time, frequency and energy amplitude is generated, which is called cutting force time-frequency feature matrix.

[0035] After the construction of the cutting force time-frequency feature matrix is completed, based on the process rules and signal response characteristics, the frequency range is divided according to the sensitive interval of the mechanical change of the machining, and specifically the overall frequency range is divided into a low frequency band (0-500Hz), a medium frequency band (500-1500Hz) and a high frequency band (1500-5000Hz). The basis for setting the preset frequency band is that the low frequency band is mainly used to capture the basic cutting force fluctuation and process load change; the medium frequency band reflects the local contact disturbance or tooth profile jump between the tool and the workpiece; and the high frequency band is most sensitive to micro wear, impact vibration, blade collapse and other high frequency abnormalities. In each frequency band range, the energy data of the corresponding frequency interval in the cutting force time-frequency feature matrix is extracted to generate the corresponding energy distribution sequence. Then, peak detection operation is performed on the energy sequence in each frequency band to identify the main frequency peak value of the frequency band, i.e. the frequency point with the most concentrated energy, which usually reflects the main excitation frequency or structural resonance frequency of the current system. At the same time, by analyzing the frequency spectrum width near the main frequency, the bandwidth of the frequency point is calculated to measure the diffusion degree of the frequency energy, and the wider the bandwidth, the lower the signal stability or the stronger the stress disturbance. In addition, the change trend of the energy density in different time segments is tracked to determine whether there is a phenomenon of continuous energy rise or sudden aggregation in a certain frequency band. For example, during the tooth root machining stage, if there is a continuous energy rise and peak shift in the high frequency band, it may indicate that the tool has micro cracks or local blade collapse in this stage.

[0036] After the energy distribution extraction is completed, the real-time cutting force spectrum is drawn according to the energy density results of each frequency band. The spectrum diagram takes time as the horizontal axis, frequency as the vertical axis, and color intensity as the unit energy density value. The energy intensity of each frequency component at different time points is expressed by color change. The deeper the color scale, the more concentrated the cutting energy of the frequency at that time period, so that the evolution characteristics of the mechanical behavior with time and frequency in the machining process can be clearly presented.

[0037] Step S300: dynamically comparing the real-time cutting force spectrum with the initial cutting force threshold range, extracting the cutting force fluctuation characteristics for tool wear analysis, and obtaining a wear compensation coefficient.

[0038] In the embodiment of the present application, the real-time cutting force spectrum is dynamically compared with the initial cutting force threshold range, abnormal frequency bands exceeding the threshold in the spectrum are identified and calibrated, and the cutting force fluctuation feature is extracted accordingly. Then, based on the cutting force fluctuation feature, the tool stress analysis is performed, and the tool stress concentration coefficient reflecting local load concentration is calculated. On this basis, the tool degradation analysis is performed in combination with the stress history and stress concentration distribution, and the tool wear cumulative amount is quantified.

[0039] Then, the unstable change signal in the wear process is extracted to form the wear fluctuation feature reflecting the nonlinear trend of wear, and through the regression analysis of the wear fluctuation feature and the wear cumulative amount, the wear compensation coefficient for adaptive adjustment of machining parameters is generated.

[0040] Further, the method provided by the embodiment of the application further comprises:

[0041] The real-time cutting force spectrum is dynamically compared with the initial cutting force threshold range, abnormal frequency bands are identified and determined, the cutting force fluctuation feature is extracted according to the abnormal frequency bands, tool stress analysis is performed based on the cutting force fluctuation feature, a tool stress concentration coefficient is obtained, tool degradation analysis is performed according to the stress concentration coefficient, a tool wear cumulative amount is obtained, a wear fluctuation feature is obtained through feature extraction according to the tool wear cumulative amount, and the tool wear cumulative amount and the wear fluctuation feature are subjected to regression analysis to generate the wear compensation coefficient.

[0042] In the embodiment of the present application, when the real-time cutting force spectrum is dynamically compared with the initial cutting force threshold range, first, based on the multi-section structure characteristics of the spectrum, it is divided into N sub-bands according to the preset frequency band, and the energy density in unit time of each sub-band is calculated. By traversing the energy deviation degree between the N sub-bands and the upper limit of the initial threshold, a group of sub-band energy deviation indexes are obtained, and abnormal judgment identification is performed accordingly to preliminarily identify a plurality of initial abnormal frequency bands. In order to improve the integrity of the frequency domain analysis, the frequency domain continuity of these initial abnormal frequency bands is detected, and the adjacent, overlapping or short-time mutation frequency bands are merged, and finally the abnormal frequency bands for diagnosing the tool state are determined.

[0043] Next, the abnormal frequency band is further characterized. The abnormal frequency band is decomposed into a low-frequency stable component and a high-frequency disturbance component, and their typical features are extracted. The low-frequency part obtains the abnormal fundamental frequency amplitude and the abnormal harmonic energy ratio through Fourier transform, and calculates the cutting force stability index according to the above two parameters, which reflects the stress stability of the main frequency. The high-frequency part identifies the sudden peak signal, and combines the geometric feature parameters of the tool to calculate the corresponding instantaneous impact load. Finally, the cutting force fluctuation features describing the tool running stability are constructed by fusing the stability index and the impact load data.

[0044] Subsequently, based on the cutting force fluctuation features, the tool stress analysis is performed. By analyzing the abnormal fundamental frequency amplitude, the abnormal harmonic energy ratio and the instantaneous impact load extracted from the abnormal frequency band, the stress of the tool in each direction is identified. The three types of stress feature data are combined into a cutting force change diagram in three dimensions, and the von Mises equivalent stress criterion is applied to convert the complex three-axis stress state into an integrated equivalent stress value. Subsequently, the equivalent stress is compared with the yield limit extracted from the tool material database to obtain the tool stress concentration coefficient in the current state, which is used to indicate whether the tool has a significant load concentration in a specific machining stage.

[0045] On this basis, the tool degradation analysis is performed according to the tool stress concentration coefficient. In this process, the entire cutting process is divided into equal-length machining time periods (such as every 5 seconds or every tooth segment cycle), and the tool stress concentration coefficient of each period is collected and numerically accumulated to obtain the tool wear accumulation of the tool wear change in the whole cycle.

[0046] Then, according to the tool wear accumulation, the feature extraction is performed to obtain the wear fluctuation feature. Through joint analysis of the accumulated data, a multi-dimensional wear feature vector containing multiple state indicators is constructed, and the trend feature reflecting the long-term wear trend and the fluctuation intensity feature reflecting the short-term instability are extracted from the multi-dimensional wear feature vector to obtain the wear fluctuation feature. Finally, the tool wear accumulation and the wear fluctuation feature are subjected to regression analysis to generate a wear compensation coefficient.

[0047] Further, in the method provided by the application embodiment, the real-time cutting force frequency spectrum is dynamically compared with the initial cutting force threshold range to identify and determine the abnormal frequency band, and the method further comprises:

[0048] The real-time cutting force spectrum is divided into N sub-bands according to the preset frequency band, N is an integer greater than 1; the N sub-bands are compared with the initial cutting force threshold range in energy density, the deviation of the energy value of each sub-band from the upper limit of the initial cutting force threshold range is calculated, and a plurality of sub-band energy deviations are obtained; abnormality judgment identification is performed according to the plurality of sub-band energy deviations, and a plurality of initial abnormal frequency bands are obtained; frequency domain continuity detection is performed based on the plurality of initial abnormal frequency bands, and the plurality of initial abnormal frequency bands are merged according to the detection result to determine the abnormal frequency band.

[0049] In the embodiments of the present application, first, the real-time cutting force spectrum is divided into N sub-bands according to the preset frequency band, where N is an integer greater than 1, and the division principle is usually based on frequency response characteristics to divide into low frequency band (0~500Hz), medium frequency band (500~1500Hz) and high frequency band (1500~5000Hz) typical machining sensitive frequency intervals, to ensure that different types of vibration and stress changes in the cutting process can be covered.

[0050] Then, the unit time energy density of the signal data in each sub-band is calculated. Specifically, first, the original time domain signal is processed based on short-time Fourier transform to obtain a cutting force time-frequency feature matrix containing time, frequency and amplitude relationship, which is used to describe the change trend of the cutting force signal at different time windows and frequency points. In the frequency interval corresponding to each sub-band, the corresponding frequency column is extracted from the matrix, and the amplitude of the frequency column in all time windows is processed point by point to obtain the energy spectrum density value sequence under the corresponding frequency. Then, in the set fixed time window (for example, 0.1 second), the energy spectrum density value of the frequency sub-band is executed numerical integration operation, and the total energy value of the sub-band in the current time period is obtained. Finally, the total energy value is divided by the length of the time window to calculate the unit time energy density of the sub-band, which reflects the stress intensity of the frequency interval under the current machining state. After obtaining the unit time energy density values of all sub-bands, energy density comparison is performed one by one, that is, the energy density of each sub-band is compared with the energy upper limit of the corresponding frequency band in the initial cutting force threshold range, the difference between the two is calculated, and the energy deviation of the sub-band is obtained, and then a plurality of sub-band energy deviations are obtained as the basic index for abnormal frequency identification.

[0051] Next, abnormality judgment identification is performed, and a fixed overrun judgment method is used to compare each sub-band energy deviation with a judgment threshold (such as 120% of the upper limit of the threshold), and if the deviation exceeds the threshold, the sub-band is marked as an initial abnormal frequency band.

[0052] To improve the continuity of frequency analysis, frequency domain continuity detection is performed on all initial abnormal frequency bands, and frequency adjacency merging is adopted to compare the center frequency difference, energy change direction and bandwidth intersection ratio of adjacent subbands. If the center frequency difference is less than 100 Hz, the energy deviation direction is consistent, and the bandwidth intersection is more than 30%, the initial abnormal frequency bands are determined to belong to the same response region, and are merged.

[0053] Further, the method provided by the application embodiment further comprises:

[0054] The abnormal frequency band is divided into a low-frequency stable component and a high-frequency disturbance component; Fourier transform is performed on the low-frequency stable component to extract an abnormal transform parameter, the abnormal transform parameter including an abnormal fundamental frequency amplitude and an abnormal harmonic energy ratio; a cutting force stability index is calculated based on the abnormal fundamental frequency amplitude and the abnormal harmonic energy ratio, and the cutting force stability index is added to the cutting force fluctuation feature; a sudden peak signal is identified in the high-frequency disturbance component, and an instantaneous impact load is calculated based on the sudden peak signal and the tool geometric feature parameter to generate the cutting force fluctuation feature.

[0055] In the application embodiment, feature structure decomposition is performed on the determined abnormal frequency band. First, the frequency threshold division method is used to divide the abnormal frequency band into a low-frequency stable component and a high-frequency disturbance component. The division is usually based on an empirical critical point of 500 Hz, and signals below 500 Hz are regarded as low-frequency stable components, reflecting periodic cutting load changes of the tool; signals above 500 Hz are defined as high-frequency disturbance components, used to capture non-stationary responses such as microscopic wear, vibration excitation or instantaneous impact.

[0056] For the low-frequency stable component, fast Fourier transform is used to reconstruct it in the frequency domain. Through FFT, the signal is converted from the time domain to the frequency domain to obtain a frequency-amplitude spectrum. Two key indicators are extracted in the spectrum to form abnormal transform parameters, one is the abnormal fundamental frequency amplitude, that is, the amplitude strength of the first main frequency point in the frequency spectrum, representing the force response strength under the main rhythm of the cutting process; the other is the abnormal harmonic energy ratio, which is calculated by taking the energy sum of the first k harmonics (usually k=5~7) and comparing it with the fundamental frequency energy to form a harmonic interference coefficient reflecting periodic consistency. The higher the ratio, the stronger the high-order harmonic interference, and the more unstable the cutting process. Subsequently, a cutting force stability index (CFSI) is calculated based on the two parameters, and the calculation formula is wherein, is the fundamental frequency amplitude, is the i-th harmonic amplitude (i=1 to k).

[0057] For high-frequency disturbance components, a sliding window peak detection algorithm is used for burst signal analysis. The sliding window width is set to 10 ms, the step is 2 ms, the local maximum value in each window is identified, and the effective impact signal is screened by setting a threshold (such as 2.5 times the mean value). Such signals usually correspond to impact events caused by tool cutting into hard spots, uneven material, or local blade collapse. After identifying the peak signal, the instantaneous impact load F is estimated using the following model combined with the geometric parameters of the tool (including tool tip corner radius, blade width, etc.), where F is the instantaneous impact load; E is the elastic modulus of the tool material, which is obtained from the material database; A is the contact area, which is estimated based on tool geometry (such as constructing a semi-ellipsoid model with corner radius and blade width); δ is the peak displacement, which is measured by the laser Doppler vibration meter and converted to micron-level displacement according to the sensitivity; L is the impact path length, which can be estimated by tool angle and sound speed, usually 1-2 mm.

[0058] Finally, the instantaneous impact load F and the aforementioned cutting force stability index CFSI constitute the cutting force fluctuation characteristics under the current abnormal frequency band.

[0059] Further, the method provided by the application embodiment includes:

[0060] Based on the tool wear cumulative amount, a multi-dimensional wear feature vector is constructed by joint analysis, wear trend features and fluctuation intensity features are extracted based on the multi-dimensional wear feature vector; wear fluctuation analysis is performed according to the wear trend features and the fluctuation intensity features to obtain wear fluctuation features; the multi-dimensional wear feature vector is processed by kernel principal component analysis for dimension reduction to obtain key wear modal components; support vector regression is performed on the key wear modal components combined with the wear fluctuation features to obtain the wear compensation coefficient.

[0061] In the application embodiment, first, based on the wear cumulative amount of the tool in the continuous machining process, joint analysis is performed to construct a multi-dimensional wear feature vector. This feature vector contains multiple feature parameters from different monitoring sources, including wear increment per unit time, tool stress mean value, impact load peak value, vibration response amplitude, cutting temperature change rate, etc. Each feature can be obtained by real-time recording by the sensor system during machining to form a complete wear evolution description sequence.

[0062] Subsequently, based on the multi-dimensional wear feature vector, linear regression analysis method is used to fit the time series data of each dimension, and the wear trend features representing long-term change trend, such as wear growth slope, are extracted. Combined with the sliding window standard deviation analysis technology, the variance, standard deviation and coefficient of variation of each feature sequence in the sliding window are calculated to extract the fluctuation intensity features representing the short-term fluctuation amplitude. These trend and fluctuation features are used to identify the long-term stability and short-term disturbance degree of the tool running state.

[0063] After obtaining the above two types of features, wear fluctuation analysis is performed, and a composite index that can represent both trend stability and local impact, i.e., wear fluctuation feature, is extracted by combining and mapping. For example, in the stage where high trend slope and high fluctuation coefficient appear at the same time, it is determined that the tool has a rapid and unstable wear process, which is used as a key analysis index for subsequent modeling.

[0064] In order to reduce the feature dimension and improve the regression efficiency, kernel principal component analysis is used to reduce the dimension of the original feature vector. This method uses radial basis function as the kernel function, maps the original nonlinear data to a high-dimensional feature space, and extracts the first several principal components whose cumulative contribution rate is higher than 90% as the key wear modal components representing the core structure of tool wear change.

[0065] Finally, the above key wear modal components and wear fluctuation features are input into the support vector regression model. The model is trained based on multiple machining samples with known wear compensation labels. Each training sample includes sample key wear modal components, sample wear fluctuation features and their corresponding wear compensation coefficients. The wear compensation coefficient is the output of the model, and the sample key wear modal components and sample wear fluctuation features are the input of the model. During the training phase, the model learns the mapping relationship between the input features and the wear compensation coefficient to establish a prediction model. After training, the newly collected key wear modal components and wear fluctuation features are input into the model to predict and output the current corresponding wear compensation coefficient in real time.

[0066] Step S400: Real-time correction of cutting force of special-shaped tooth workpiece according to the wear compensation coefficient to generate tool cutting force control suggestion.

[0067] In the embodiment of the present application, when the cutting force of the profiled tooth workpiece is real-time corrected according to the wear compensation coefficient, the corresponding spindle speed compensation amount and feed amount correction value are dynamically calculated according to the wear compensation coefficient obtained by wear analysis, and the control parameter correction amount is obtained by stage weight distribution of the above correction parameters through extraction of the real-time cutting stage of the profiled tooth workpiece in combination with the actual machining state of the profiled tooth workpiece. Then, the multi-axis linkage control path of the numerical control system is real-time adjusted according to the correction amount, the control optimization instruction in the machining process is generated, and the tool movement trajectory is dynamically updated in a closed loop, and finally the tool cutting force control suggestion is output, so as to realize the precise control of the cutting force and the wear adaptability compensation.

[0068] Further, the method provided by the embodiment of the present application further comprises the following steps:

[0069] The wear compensation coefficient is analyzed to determine the spindle speed compensation amount and the feed amount correction value; the real-time cutting stage of the profiled tooth machining position is extracted, and the spindle speed compensation amount and the feed amount correction value are weight distributed according to the real-time cutting stage to determine the control parameter correction amount; the profiled tooth workpiece is controlled by multi-axis linkage based on the control parameter correction amount to generate a control optimization instruction; the profiled tooth machining trajectory is controlled in a closed loop by executing the control optimization instruction to generate the tool cutting force control suggestion.

[0070] In the embodiment of the present application, first, the wear compensation coefficient is analyzed, and a linear proportional coefficient model is used to map it into the spindle speed compensation amount and the feed amount correction value. The spindle speed compensation amount is calculated by the adjustment formula , and the feed amount correction value is calculated by , represents the wear compensation coefficient, , are the speed and feed adjustment coefficients set based on the tool material and process experience.

[0071] Secondly, the profiled tooth machining path is real-time calibrated to identify the current cutting stage. By matching the time stamp and spatial coordinates of each trajectory point in the CNC code, in combination with the set geometric feature positions such as the tool contact point or the tooth profile boundary, the entire machining process is divided into three stages of cutting-in section, tooth surface stable cutting section and tooth root transition section. The system calls the preset weight set according to the machining stage to weight distribute the above compensation amount, and finally obtains the control parameter correction amount of the current machining stage, that is, the weighted spindle speed correction value and the feed rate adjustment value.

[0072] Then, the current multi-axis machining trajectory is updated with the control parameter correction amount as input. Specifically, intermediate control points are inserted between the original tool path points by a cubic spline interpolation algorithm to generate a new trajectory with smooth transitions, thereby avoiding trajectory jumps and unstable cutting caused by parameter mutations, and outputting a new tool position point sequence and spindle / feed control parameters to form a new control optimization instruction.

[0073] Then, the control optimization instruction is executed to perform closed-loop control on the profile tooth machining trajectory. In this process, first, machining analysis is performed according to the control optimization instruction to extract the actual machining trajectory of the profile tooth workpiece and compare it with the original planned trajectory to obtain the offset information in the trajectory execution. Then, the trajectory inverse solving method is used to back out the path offset based on the actual path. In combination with the local curvature distribution of the profile tooth profile, dynamic optimization analysis is performed on the offset area to identify multiple dynamic optimization nodes with compensation value. These nodes are embedded into the original trajectory to generate an optimized machining trajectory through interpolation reconstruction. Finally, the optimized trajectory is uploaded to the numerical control controller to build a closed-loop control logic between the machining path and the feedback response, and output the corresponding tool cutting force control suggestion.

[0074] Further, the method provided by the application embodiment further comprises the following steps:

[0075] According to the control optimization instruction, machining analysis is performed on the profile tooth workpiece to obtain the profile tooth machining trajectory. Based on the control optimization instruction, the profile tooth machining trajectory is inversely solved to determine the path offset. According to the path offset and the tooth profile curvature characteristics of the profile tooth workpiece, optimization analysis is performed to obtain multiple dynamic optimization nodes. The multiple dynamic optimization nodes are inserted into the profile tooth machining trajectory for reconstruction to generate an optimized machining trajectory. The optimized machining trajectory is updated to the numerical control system for closed-loop control to generate the tool cutting force control suggestion.

[0076] In the application embodiment, first, machining analysis is performed on the profile tooth workpiece according to the control optimization instruction. In this process, the displacement code value fed back by the servo system in the numerical control machining process is collected, and the actual and theoretical trajectories are aligned by using a synchronous timestamp algorithm in combination with the theoretical machining trajectory instruction (such as the tool displacement path defined by the G code) to form the real motion trajectory of the tool in the three-dimensional space. The trajectory points take the tool tip as the reference center to record the position change sequence in each machining time segment, thereby obtaining the profile tooth machining trajectory.

[0077] Then, the profiled tooth machining trajectory is reversely solved based on the control optimization instruction. The Euclidean distance between each actual machining trajectory point and its corresponding theoretical path point is calculated by using the space vector distance calculation, and the deviation value in the three-axis direction, i.e., the path offset, is obtained. By traversing each trajectory segment, the path offset curve on the entire machining path is finally output.

[0078] Next, the path offset is combined with the tooth profile curvature characteristics of the profiled tooth workpiece for joint optimization analysis. First, based on the three-dimensional tooth profile modeling data of the profiled tooth workpiece, the local curvature values of each point on the machining path are calculated using the finite difference method, and the tooth profile curvature change curve is obtained. Then, the path offset curve and the curvature change curve are compared and analyzed. By setting the offset gradient threshold and the curvature change rate threshold, the region where both the offset and the curvature change suddenly change is identified as the precision risk point. Representative trajectory points are selected in these regions, and their spatial positions, corresponding curvature values and offsets are extracted as the basic data of the dynamic optimization nodes, and finally a plurality of dynamic optimization nodes are obtained.

[0079] Then, the plurality of dynamic optimization nodes are inserted into the profiled tooth machining trajectory for reconstruction. The trajectory reconstruction adopts the cubic B-spline interpolation method to maintain the smoothness and derivative continuity of the original trajectory. Smooth curve segments are generated between the original trajectory segments and the optimization nodes, and the control point sequence of the path is updated. This method can effectively avoid machining vibration and precision decline caused by trajectory mutation, and finally generate an optimized machining trajectory to ensure stable stress and high-quality forming of the tool during profiled tooth surface machining.

[0080] Finally, the optimized machining trajectory is updated to the numerical control system for closed-loop control to generate tool cutting force control suggestions. The control suggestions are based on the offset trend between the optimized trajectory and the original trajectory and real-time machining parameters (such as spindle load, current fluctuation, etc.) for strategy adjustment, including real-time adjustment of spindle speed, feed rate and acceleration limit, and other key parameters. At the same time, the optimized machining trajectory is executed as the main path, and prediction and control suggestions are provided for subsequent similar tooth segment machining to realize dynamic adjustment and intelligent feedback.

[0081] In the embodiments of the present application, as described above, the embodiments of the present application have at least the following technical effects:

[0082] This application obtains the machining parameters of irregular-shaped tooth workpieces, performs cutting analysis based on tool geometric features and machining parameter sets, and determines the initial cutting force threshold range; it collects dynamic cutting force data of the tool in real time and generates a real-time cutting force spectrum; it dynamically compares the real-time cutting force spectrum with the initial cutting force threshold range, extracts cutting force fluctuation characteristics for tool wear analysis, and obtains a wear compensation coefficient; it then performs real-time correction of the cutting force of the irregular-shaped tooth workpiece according to the wear compensation coefficient, generating tool cutting force control suggestions. This invention solves the technical problem of existing technologies being unable to perceive and effectively control the cutting force during the machining of irregular-shaped teeth in real time, leading to rapid tool wear and low machining accuracy. By combining tool geometric features and machining parameters to model the cutting force, collecting cutting force data in real time and dynamically comparing and analyzing it, calculating the wear compensation coefficient, and implementing closed-loop correction control, it achieves dynamic regulation of tool cutting force, extends tool life, and improves the stability and accuracy of irregular-shaped tooth machining.

[0083] Example 2, based on the same inventive concept as the cutting force control method for machining irregularly shaped teeth in the foregoing examples, such as... Figure 2 As shown, this application provides a tool cutting force control system for machining irregularly shaped teeth. The system and method embodiments in this application are based on the same inventive concept. The system includes:

[0084] The cutting analysis module 11 is used to acquire the machining parameters of the irregular-shaped tooth workpiece, perform cutting analysis according to the tool geometric feature parameters and the machining parameter set, and determine the initial cutting force threshold range; the real-time data acquisition module 12 is used to acquire the dynamic cutting force data of the tool in real time and generate a real-time cutting force spectrum; the wear analysis module 13 is used to dynamically compare the real-time cutting force spectrum according to the initial cutting force threshold range, extract the cutting force fluctuation characteristics for tool wear analysis, and obtain the wear compensation coefficient; the real-time correction module 14 is used to correct the cutting force of the irregular-shaped tooth workpiece in real time according to the wear compensation coefficient and generate tool cutting force control suggestions.

[0085] Furthermore, the system is also used to implement the following functions:

[0086] Based on the machining parameters, the tooth profile of the irregular tooth workpiece is analyzed to obtain tooth profile feature parameters; cutting prediction is performed based on the tooth profile feature parameters and the tool geometric feature parameters to calculate theoretical cutting force data; actual cutting force data is dynamically collected through multiple sensors, and the theoretical cutting force data is simulated and corrected based on the actual cutting force data to determine the initial cutting force threshold range.

[0087] Furthermore, the system is also used to implement the following functions:

[0088] A three-dimensional strain sensor array is arranged on the tool, and the tool is sensed and collected by the three-dimensional strain sensor array to obtain dynamic cutting force data, which includes multiple axial cutting force components; a vibration sensor is arranged on the profiled tooth workpiece, and a machining sensor is performed by the vibration sensor to obtain a high-frequency vibration signal; the multiple axial cutting force components and the high-frequency vibration signal are fused in time and frequency domains to construct a cutting force time-frequency feature matrix; based on the cutting force time-frequency feature matrix, a preset frequency band is decomposed to extract energy distribution data of the preset frequency band, and the real-time cutting force spectrum is drawn according to the energy distribution data.

[0089] Further, the system is also used to realize the following functions:

[0090] The real-time cutting force spectrum is dynamically compared with the initial cutting force threshold range to identify and determine an abnormal frequency band, and the cutting force fluctuation feature is extracted according to the abnormal frequency band; based on the cutting force fluctuation feature, tool stress analysis is performed to obtain a tool stress concentration coefficient; according to the stress concentration coefficient, tool degradation analysis is performed to obtain a tool wear cumulative amount; according to the tool wear cumulative amount, a feature is extracted to obtain a wear fluctuation feature, and the tool wear cumulative amount and the wear fluctuation feature are subjected to regression analysis to generate the wear compensation coefficient.

[0091] Further, the system is also used to realize the following functions:

[0092] The real-time cutting force spectrum is divided into N sub-bands according to the preset frequency band, N is an integer greater than 1; the N sub-bands are compared with the initial cutting force threshold range in energy density, the deviation degree of each sub-band energy value from the upper limit of the initial cutting force threshold range is calculated, and multiple sub-band energy deviation degrees are obtained; according to the multiple sub-band energy deviation degrees, abnormality judgment is performed to obtain multiple initial abnormal frequency bands; based on the multiple initial abnormal frequency bands, frequency domain continuity detection is performed, the multiple initial abnormal frequency bands are combined according to the detection result, and the abnormal frequency band is determined.

[0093] Further, the system is also used to realize the following functions:

[0094] The abnormal frequency band is divided into a low-frequency stable component and a high-frequency disturbance component; the low-frequency stable component is subjected to Fourier transform to extract an abnormal transform parameter, the abnormal transform parameter includes an abnormal fundamental frequency amplitude and an abnormal harmonic energy ratio; based on the abnormal fundamental frequency amplitude and the abnormal harmonic energy ratio, a cutting force stability index is calculated, the cutting force stability index is added to the cutting force fluctuation feature; based on the high-frequency disturbance component, a burst peak signal is identified, an instantaneous impact load is calculated according to the burst peak signal combined with the tool geometric feature parameter, and the cutting force fluctuation feature is generated.

[0095] Further, the system is also used to implement the following functions:

[0096] Based on the tool wear accumulation, joint analysis is performed to construct a multi-dimensional wear feature vector, and based on the multi-dimensional wear feature vector, wear trend features and fluctuation intensity features are extracted; wear fluctuation analysis is performed according to the wear trend features and the fluctuation intensity features to obtain wear fluctuation features; kernel principal component analysis is used to perform dimension reduction processing on the multi-dimensional wear feature vector to obtain key wear modal components; support vector regression is performed on the key wear modal components in combination with the wear fluctuation features to obtain the wear compensation coefficient.

[0097] Further, the system is also used to implement the following functions:

[0098] The wear compensation coefficient is analyzed to determine a spindle speed compensation amount and a feed rate correction value; a real-time cutting stage in which a profile tooth machining position is located is extracted, and weight distribution is performed on the spindle speed compensation amount and the feed rate correction value according to the real-time cutting stage to determine a control parameter correction amount; multi-axis linkage control is performed on the profile tooth workpiece based on the control parameter correction amount to generate a control optimization instruction; and closed-loop control is performed on the profile tooth machining trajectory by executing the control optimization instruction to generate the tool cutting force control suggestion.

[0099] Further, the system is also used to implement the following functions:

[0100] According to the control optimization instruction, machining analysis is performed on the profile tooth workpiece to obtain the profile tooth machining trajectory; the profile tooth machining trajectory is inversely solved based on the control optimization instruction to determine a path offset amount; according to the path offset amount in combination with tooth profile curvature features of the profile tooth workpiece, optimization analysis is performed to obtain a plurality of dynamic optimization nodes; the plurality of dynamic optimization nodes are inserted into the profile tooth machining trajectory for reconstruction to generate an optimized machining trajectory, and the optimized machining trajectory is synchronously updated to a numerical control system for closed-loop control to generate the tool cutting force control suggestion.

[0101] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. The above describes a specific embodiment of the present application. The processes depicted in the drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are possible or can be advantageous.

[0102] The above only describes the preferred embodiments of the present application and does not limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0103] The specification and drawings are, of course, to be regarded in an illustrative rather than a restrictive sense. It is to be understood that any such modifications, variations, combinations or equivalents that fall within the scope of the application are intended to be embraced herein.

Claims

1. A method for controlling the cutting force of a tool used in machining irregularly shaped teeth, characterized in that, The method includes: Obtain the machining parameters of the irregular toothed workpiece, perform cutting analysis based on the tool geometric feature parameters and the machining parameter set, and determine the initial cutting force threshold range; Real-time acquisition of dynamic cutting force data of the tool, generating a real-time cutting force spectrum; The real-time cutting force spectrum is dynamically compared according to the initial cutting force threshold range, and the cutting force fluctuation characteristics are extracted for tool wear analysis to obtain the wear compensation coefficient. Based on the wear compensation coefficient, the cutting force of the irregular tooth workpiece is corrected in real time, and tool cutting force control suggestions are generated. Real-time acquisition of dynamic cutting force data of the tool to generate a real-time cutting force spectrum, including the following methods: A triaxial strain sensor array is deployed on the cutting tool, and the cutting tool is sensed and collected through the triaxial strain sensor array to obtain dynamic cutting force data, which includes multiple axial cutting force components. Vibration sensors are installed on irregularly shaped toothed workpieces to obtain high-frequency vibration signals through processing sensing. The multiple axial cutting force components are fused with the high-frequency vibration signal in the time-frequency domain to construct a cutting force time-frequency feature matrix. Based on the cutting force time-frequency feature matrix, the energy distribution data of the preset frequency band is extracted and the real-time cutting force spectrum is plotted according to the energy distribution data. The method involves dynamically comparing the real-time cutting force spectrum according to the initial cutting force threshold range, extracting cutting force fluctuation characteristics for tool wear analysis, and obtaining the wear compensation coefficient. The real-time cutting force spectrum is dynamically compared with the initial cutting force threshold range to identify abnormal frequency bands, and the cutting force fluctuation characteristics are extracted based on the abnormal frequency bands. Based on the cutting force fluctuation characteristics, tool stress analysis is performed to obtain the tool stress concentration factor; Tool degradation analysis is performed according to the stress concentration factor to obtain the cumulative tool wear. Feature extraction is performed based on the cumulative tool wear to obtain wear fluctuation characteristics. Regression analysis is then performed between the cumulative tool wear and the wear fluctuation characteristics to generate the wear compensation coefficient.

2. The cutting force control method for machining irregularly shaped teeth as described in claim 1, characterized in that, To obtain the machining parameters of the irregularly shaped toothed workpiece, and to perform cutting analysis based on the tool geometry parameters and the machining parameter set, the initial cutting force threshold range is determined. The methods include: Based on the aforementioned processing parameters, tooth profile analysis is performed on the irregular tooth workpiece to obtain tooth profile feature parameters; Based on the tooth profile characteristic parameters and the tool geometric characteristic parameters, cutting prediction is performed, and theoretical cutting force data is calculated. The actual cutting force data is dynamically acquired by multiple sensors, and the theoretical cutting force data is simulated and corrected based on the actual cutting force data to determine the initial cutting force threshold range.

3. The cutting force control method for machining irregularly shaped teeth as described in claim 1, characterized in that, The method involves dynamically comparing the real-time cutting force spectrum with the initial cutting force threshold range to identify and determine abnormal frequency bands. The real-time cutting force spectrum is divided into N sub-bands according to the preset frequency band, where N is an integer greater than 1; The energy density of each of the N sub-bands is compared with that of the initial cutting force threshold range. The deviation of the energy value of each sub-band from the upper limit of the initial cutting force threshold range is calculated to obtain the energy deviation of multiple sub-bands. Anomalies are identified based on the energy deviation of the multiple sub-bands to obtain multiple initial abnormal frequency bands; Frequency domain continuity detection is performed based on the multiple initial abnormal frequency bands, and the multiple initial abnormal frequency bands are merged according to the detection results to determine the abnormal frequency band.

4. The cutting force control method for machining irregularly shaped teeth as described in claim 3, characterized in that, The method for extracting the cutting force fluctuation characteristics based on the abnormal frequency band includes: The abnormal frequency band is divided into a low-frequency stable component and a high-frequency disturbance component; Perform a Fourier transform on the low-frequency stable component to extract the abnormal transform parameters, which include the abnormal fundamental frequency amplitude and the abnormal harmonic energy ratio. The cutting force stability index is calculated based on the ratio of the abnormal fundamental frequency amplitude to the abnormal harmonic energy, and the cutting force stability index is added to the cutting force fluctuation characteristics. Based on the identification of sudden peak signals in the high-frequency disturbance components, the instantaneous impact load is calculated according to the sudden peak signals and the tool geometric feature parameters to generate the cutting force fluctuation characteristics.

5. The cutting force control method for machining irregularly shaped teeth as described in claim 1, characterized in that, Based on the cumulative tool wear, feature extraction is performed to obtain wear fluctuation characteristics. Regression analysis is then performed between the cumulative tool wear and the wear fluctuation characteristics to generate the wear compensation coefficient. The method includes: Based on the cumulative tool wear, a joint analysis is performed to construct a multidimensional wear feature vector, and wear trend features and fluctuation intensity features are extracted based on the multidimensional wear feature vector. Wear fluctuation analysis is performed based on the wear trend characteristics and the fluctuation intensity characteristics to obtain wear fluctuation characteristics; Kernel principal component analysis was used to reduce the dimensionality of the multidimensional wear feature vector to obtain the key wear mode components. The wear compensation coefficient is obtained by combining the key wear mode components with the wear fluctuation characteristics through support vector regression.

6. The cutting force control method for machining irregularly shaped teeth as described in claim 1, characterized in that, The method includes real-time correction of the cutting force of the irregular-shaped tooth workpiece according to the wear compensation coefficient, and generation of tool cutting force control suggestions. Analyze the wear compensation coefficient to determine the spindle speed compensation and feed rate correction values; Extract the real-time cutting stage of the machining position of the irregular tooth, and allocate the spindle speed compensation and feed correction value according to the real-time cutting stage to determine the control parameter correction amount; Based on the control parameter correction, multi-axis linkage control is performed on the irregular tooth workpiece to generate control optimization instructions; The control optimization instructions are executed to perform closed-loop control on the machining trajectory of the irregular teeth, and the tool cutting force control suggestions are generated.

7. The cutting force control method for machining irregularly shaped teeth as described in claim 6, characterized in that, The method includes executing the control optimization instructions to perform closed-loop control on the machining trajectory of irregular teeth and generating the tool cutting force control suggestions, wherein the control optimization instructions are executed. The machining trajectory of the irregular tooth workpiece is obtained by performing machining analysis on the irregular tooth workpiece according to the control optimization instructions. Based on the control optimization instructions, the machining trajectory of the irregular tooth is solved in reverse to determine the path offset; Based on the path offset and the tooth curvature characteristics of the irregular tooth workpiece, an optimization analysis is performed to obtain multiple dynamic optimization nodes; The multiple dynamic optimization nodes are inserted into the machining trajectory of the irregular tooth to reconstruct the machining trajectory, generating an optimized machining trajectory. The optimized machining trajectory is then synchronously updated to the CNC system for closed-loop control, generating the tool cutting force control suggestion.

8. A tool cutting force control system for machining irregularly shaped teeth, characterized in that, The system includes: The cutting analysis module is used to obtain the machining parameters of irregularly shaped toothed workpieces, perform cutting analysis based on the tool geometric feature parameters and the machining parameter set, and determine the initial cutting force threshold range. The real-time data acquisition module is used to acquire dynamic cutting force data of the tool in real time and generate a real-time cutting force spectrum; The wear analysis module is used to dynamically compare the real-time cutting force spectrum according to the initial cutting force threshold range, extract the cutting force fluctuation characteristics to perform tool wear analysis, and obtain the wear compensation coefficient. The real-time correction module is used to correct the cutting force of the irregular tooth workpiece in real time according to the wear compensation coefficient and generate tool cutting force control suggestions. Real-time acquisition of dynamic cutting force data of the tool to generate a real-time cutting force spectrum, including the following methods: A triaxial strain sensor array is deployed on the cutting tool, and the cutting tool is sensed and collected through the triaxial strain sensor array to obtain dynamic cutting force data, which includes multiple axial cutting force components. Vibration sensors are installed on irregularly shaped toothed workpieces to obtain high-frequency vibration signals through processing sensing. The multiple axial cutting force components are fused with the high-frequency vibration signal in the time-frequency domain to construct a cutting force time-frequency feature matrix. Based on the cutting force time-frequency feature matrix, the energy distribution data of the preset frequency band is extracted and the real-time cutting force spectrum is plotted according to the energy distribution data. The method involves dynamically comparing the real-time cutting force spectrum according to the initial cutting force threshold range, extracting cutting force fluctuation characteristics for tool wear analysis, and obtaining the wear compensation coefficient. The real-time cutting force spectrum is dynamically compared with the initial cutting force threshold range to identify abnormal frequency bands, and the cutting force fluctuation characteristics are extracted based on the abnormal frequency bands. Based on the cutting force fluctuation characteristics, tool stress analysis is performed to obtain the tool stress concentration factor; Tool degradation analysis is performed according to the stress concentration factor to obtain the cumulative tool wear. Feature extraction is performed based on the cumulative tool wear to obtain wear fluctuation characteristics. Regression analysis is then performed between the cumulative tool wear and the wear fluctuation characteristics to generate the wear compensation coefficient.

Images