Turn-milling swing head driving parameter self-adaptive control method and system and electronic equipment

Abstract

The invention discloses a turn-milling swing head driving parameter self-adaptive control method and system and electronic equipment, and relates to the technical field of self-adaptive control, and the method comprises the steps: constructing a dynamic parameter matrix, carrying out tensor decomposition, and extracting swing head characteristic modal parameters; a self-adaptive adjustment driving parameter set based on a nonlinear fuzzy PID unit is configured according to the swing head characteristic modal parameters; a preset machining precision index is introduced, and a driving shaft current-rotating speed coupling control parameter is configured for dynamic optimization in combination with a self-adaptive adjustment driving parameter set; and processing load sudden change detection data, processing surface quality data and tool wear monitoring data are used as multi-dimensional feedback information sources, and a swing head driving parameter self-adaptive adjustment instruction is determined. The technical problem that in the prior art, swing head driving parameters cannot be adaptively adjusted in real time according to the machining state is solved, and the technical effect of improving swing head driving response accuracy and machining stability is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of adaptive control, in particular to a turning-milling swing head driving parameter adaptive control method, system and electronic device. BACKGROUND

[0002] In the turning-milling composite machining process, the driving parameters of the swing head mechanism directly affect the machining precision and tool life. The traditional machining control method usually adopts fixed parameters or driving strategies set based on experience, which is difficult to cope with dynamic changes such as spindle vibration, torque fluctuation and contact stress distribution in the actual machining process. Especially in the scene of load mutation, tool wear or machining precision requirement change, it is difficult to realize dynamic adjustment and rapid response of parameters, which may lead to problems such as decline in machining quality, increase in equipment energy consumption and abnormal wear. SUMMARY

[0003] The present application provides a turning-milling swing head driving parameter adaptive control method, system and electronic device, which is used to solve the technical problem that the swing head driving parameters cannot be adjusted in real time according to the machining state in the prior art.

[0004] In view of the above problems, the present application provides a turning-milling swing head driving parameter adaptive control method, system and electronic device.

[0005] In a first aspect of the present application, a turning-milling swing head driving parameter adaptive control method is provided, which comprises: Collecting multi-source dynamic data including spindle vibration spectrum, cutting torque fluctuation and contact stress distribution in the turning-milling swing head machining process to construct a dynamic parameter matrix; performing tensor decomposition based on the dynamic parameter matrix to extract swing head characteristic modal parameters; configuring an adaptive adjustment driving parameter set based on a nonlinear fuzzy PID unit according to the swing head characteristic modal parameters, the nonlinear fuzzy PID unit being cooperatively operated with a double closed-loop control architecture; introducing a preset machining precision index, combining the adaptive adjustment driving parameter set, and configuring driving shaft current-speed coupling control parameters for dynamic optimization; and simultaneously taking machining load mutation detection data, machining surface quality data and tool wear monitoring data as multi-dimensional feedback information sources to determine swing head driving parameter adaptive adjustment instructions.

[0006] In a second aspect of the present application, a turning-milling swing head driving parameter adaptive control system is provided, which comprises: The dynamic parameter matrix construction module is configured to collect multi-source dynamic data including spindle vibration spectrum, cutting torque fluctuation and contact stress distribution in the swing head machining process of the turning and milling machine, and construct a dynamic parameter matrix; the tensor decomposition module is configured to perform tensor decomposition based on the dynamic parameter matrix, and extract swing head characteristic modal parameters; the configuration module is configured to configure an adaptive adjustment driving parameter set based on a nonlinear fuzzy PID unit according to the swing head characteristic modal parameters, and the nonlinear fuzzy PID unit is cooperatively operated with a double closed-loop control architecture; the dynamic optimization module is configured to introduce a preset machining precision index, combine the adaptive adjustment driving parameter set, and configure driving shaft current-speed coupling control parameters for dynamic optimization; and the adaptive adjustment instruction determination module is configured to simultaneously use machining load mutation detection data, machining surface quality data and tool wear monitoring data as multi-dimensional feedback information sources to determine a swing head driving parameter adaptive adjustment instruction.

[0007] In a third aspect, the present application provides an electronic device, comprising: a memory configured to store executable instructions; and a processor configured to execute the executable instructions stored in the memory to implement the turning and milling swing head driving parameter adaptive control method provided by the present application.

[0008] The one or more technical solutions provided in the present application have at least the following technical effects or advantages: The present application collects multi-source dynamic data including spindle vibration spectrum, cutting torque fluctuation and contact stress distribution in the swing head machining process of the turning and milling machine, and constructs a dynamic parameter matrix; performs tensor decomposition based on the dynamic parameter matrix, and extracts swing head characteristic modal parameters; configures an adaptive adjustment driving parameter set based on a nonlinear fuzzy PID unit according to the swing head characteristic modal parameters, and the nonlinear fuzzy PID unit is cooperatively operated with a double closed-loop control architecture; introduces a preset machining precision index, combines the adaptive adjustment driving parameter set, and configures driving shaft current-speed coupling control parameters for dynamic optimization; and simultaneously uses machining load mutation detection data, machining surface quality data and tool wear monitoring data as multi-dimensional feedback information sources to determine a swing head driving parameter adaptive adjustment instruction. The present application solves the technical problem that the swing head driving parameter cannot be adaptively adjusted in real time according to the machining state in the prior art, and achieves the technical effects of improving the response accuracy and machining stability of the swing head driving through the parameter modeling and optimization control mechanism driven by multi-source dynamic data. BRIEF DESCRIPTION OF DRAWINGS

[0009] 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 labor.

[0010] Figure 1 A vehicle-milling swing head driving parameter adaptive control method flowchart provided by an embodiment of the present application is shown in the figure. Figure 2 A vehicle-milling swing head driving parameter adaptive control system structure diagram provided by an embodiment of the present application is shown in the figure. Figure 3 A structure diagram of an exemplary electronic device of the present application is shown in the figure.

[0011] Reference signs: dynamic parameter matrix construction module 11, tensor decomposition module 12, configuration module 13, dynamic optimization module 14, adaptive adjustment instruction determination module 15, bus 300, receiver 301, processor 302, transmitter 303, memory 304, bus interface 305. DETAILED DESCRIPTION

[0012] The present application provides a vehicle-milling swing head driving parameter adaptive control method, system and electronic device, which solves the technical problem that the swing head driving parameter cannot be adjusted in real time according to the processing state in the prior art, through multi-source dynamic data driven parameter modeling and optimization control mechanism, and achieves the technical effect of improving the response accuracy and processing stability of the swing head driving.

[0013] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to 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 a person of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0014] 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 steps or units clearly listed, but can include other steps or modules that are not clearly listed or inherent to these processes, methods, products or devices.

[0015] Embodiment one, as shown in the figure, the present application provides a vehicle-milling swing head driving parameter adaptive control method, the method comprises: Figure 1 Step S100: Collecting multi-source dynamic data including spindle vibration spectrum, cutting torque fluctuation and contact stress distribution in the vehicle-milling swing head processing process, and constructing a dynamic parameter matrix.

[0016] ​In the embodiments of the present application, multi-source dynamic data including spindle vibration spectrum, cutting torque fluctuation and contact stress distribution during the swing head machining of the turning-milling machine are collected, wherein the spindle vibration spectrum is obtained by a three-axis vibration sensor array arranged radially on the swing head spindle, the cutting torque fluctuation is obtained through the instantaneous output signal of the spindle servo motor, and the contact stress distribution is measured by a strain gauge embedded at the clamping end of the tool. After wavelet packet denoising processing of the above raw data, a dynamic parameter matrix for representing the machining state is constructed. Specifically, the three types of denoised multi-source dynamic data are synchronized and calibrated according to the time axis, and after uniform sampling frequency and data length, standardized processing is performed, and the spindle vibration spectrum, cutting torque fluctuation and contact stress distribution are respectively mapped as independent feature dimensions in the matrix, and finally a dynamic parameter matrix with time window as row and three types of dynamic characteristics as column is constructed.

[0017] Further, in the method provided by the embodiments of the present application, the multi-source dynamic data including spindle vibration spectrum, cutting torque fluctuation and contact stress distribution during the swing head machining of the turning-milling machine are collected, and the dynamic parameter matrix is constructed, which further includes: A three-axis vibration sensor array is arranged radially on the swing head spindle, and a strain gauge is embedded at the clamping end of the tool to measure the dynamic stress distribution, and the instantaneous torque fluctuation data of the spindle servo motor is synchronously collected; the collected raw data is collected, and wavelet packet denoising processing is performed to obtain the multi-source dynamic data.

[0018] In the embodiments of the present application, a three-axis vibration sensor array is installed at the radial position of the swing head spindle for real-time collection of vibration signals of the spindle in X, Y and Z directions. After the signals are converted into digital signals by a data acquisition card, the vibration frequency and energy distribution of the spindle during the machining process are extracted by using a frequency domain analysis method such as fast Fourier transform (FFT), thereby forming the spindle vibration spectrum.

[0019] In order to obtain the mechanical state of the contact area between the tool and the workpiece during cutting, a strain gauge is embedded at the clamping end of the tool, the small deformation signal is converted into a voltage signal through a strain bridge, and the dynamic stress distribution data is obtained after further amplification and sampling.

[0020] At the same time, the driving current, speed and back electromotive force data of the spindle servo motor are collected through the control interface of the spindle servo motor, the real-time torque change is calculated combined with the torque constant of the motor, and the instantaneous torque fluctuation data of the spindle servo motor is obtained.

[0021] After the above three types of raw data are summarized, wavelet packet denoising processing is used for preprocessing, different frequency band sub-signals are extracted through multi-level wavelet packet decomposition of the signal, and high-frequency noise is removed combined with energy threshold identification, and main feature information is retained, thereby obtaining multi-source dynamic data.

[0022] Step S200: performing tensor decomposition based on the dynamic parameter matrix to extract the swing head characteristic modal parameters.

[0023] In the embodiment of the present application, when performing tensor decomposition based on the dynamic parameter matrix, the matrix is processed by using tensor modeling and high-order tensor decomposition technology. Specifically, the dynamic parameter matrix is first expanded into a three-order tensor structure, with the dimensions of time series, characteristic types (including vibration spectrum, torque fluctuation, stress distribution) and sensor spatial arrangement points, to retain the interaction characteristics of various data in the time series, attributes and spatial dimensions.

[0024] Subsequently, high-order singular value decomposition (HOSVD) is performed on the three-order tensor. This method extracts three groups of orthogonal matrices corresponding to the time modal matrix, the characteristic modal matrix and the spatial modal matrix by performing SVD operation on each dimension of the tensor, and generates a core tensor that preserves the principal component characteristics. In this process, the time modal matrix reveals the main vibration / stress change period in the machining process, the characteristic modal matrix describes the coupling response mode between the three types of dynamic characteristics, and the spatial modal matrix reflects the stress distribution trend between different sensor points.

[0025] By performing energy analysis and feature truncation on the core tensor, the modal components of the dominant dynamic behavior are identified, and then the swing head characteristic modal parameters that constitute the swing head operation change characteristics are extracted from the three groups of modal matrices. These parameters include the main frequency characteristics representing abnormal vibration of the main shaft, the intrinsic response mode representing cutting load fluctuation, and the principal strain direction of contact stress change, which are used to comprehensively reflect the dominant factors of the dynamic coupling state in the swing head machining process.

[0026] Step S300: configuring an adaptive adjustment driving parameter set based on a nonlinear fuzzy PID unit according to the swing head characteristic modal parameters, the nonlinear fuzzy PID unit being cooperatively operated with a double closed-loop control architecture.

[0027] In the embodiment of the present application, after extracting the swing head characteristic modal parameters, to realize real-time optimization of the driving parameters in the machining process, an adaptive adjustment driving parameter set based on a nonlinear fuzzy PID unit is further configured, and the nonlinear fuzzy PID unit is cooperatively operated with a double closed-loop control architecture to form a closed-loop dynamic adjustment mechanism.

[0028] Specifically, first, the head swing feature modal parameter is introduced as a key input variable into the nonlinear fuzzy PID unit, which has three parallel operation channels inside, respectively for processing the feature information of multi-source dynamic data, the first channel uses wavelet packet energy entropy analysis to analyze the degree of nonlinearity in the cutting process, to identify the influence of vibration intensity change on machining stability; the second channel extracts the main features representing different frequency components from the dynamic decomposition of spindle torque fluctuation data based on the intrinsic mode function extraction technology; the third channel constructs a convolutional neural network model to establish the mapping relationship between the micro damage of the cutting edge and the contact stress distribution, and improve the recognition accuracy of the change of material contact state.

[0029] The output results of the above parallel channels are used to adjust the three parameters of the nonlinear fuzzy PID controller in real time, namely the proportional coefficient, the integral coefficient and the differential coefficient, to realize the nonlinear compensation adjustment of the driving process. The control unit automatically switches the adjustment strategy under different working conditions, making the adjustment behavior more adaptable to dynamic disturbances.

[0030] During the operation of the nonlinear fuzzy PID unit, it cooperates with the double closed-loop control architecture. Among them, the inner loop control focuses on adjusting the spindle feed rate and the swing angle offset at high frequency, through the introduction of the multi-scale convolution kernel analysis method, the swing feature modal parameter is analyzed in time domain and frequency domain, the response relationship between parameters is established, and fast adjustment is realized; while the outer loop control takes the cutting stiffness coefficient as the core variable, combines with the change of machining environment and working condition, dynamically constrains the adjustable boundary, and uses stability theory to ensure that the control adjustment process does not deviate from the safe range.

[0031] Through the cooperative work of the nonlinear fuzzy PID unit and the double closed-loop control architecture, an adaptive adjustment driving parameter set for controlling the driving behavior is generated in real time, which includes current response rate, spindle speed regulation ratio, swing angle deviation compensation, etc. Key control parameters can effectively improve machining precision, reduce dynamic impact, and enhance the adaptability of the swing head structure to complex working conditions.

[0032] Further, the method provided by the application embodiment further comprises: The inner loop of the double closed-loop control architecture analyzes the time-frequency coupling relationship of the feature modal through the multi-scale convolution kernel, generates the cooperative constraint condition of the spindle feed rate and the swing angle offset; the outer loop of the double closed-loop control architecture dynamically optimizes the cutting stiffness coefficient, and introduces the stability theorem to construct the parameter adjustment boundary.

[0033] In the embodiments of the present application, in the double closed-loop control architecture, the inner loop control focuses on fine adjustment of rapid dynamic changes in the machining process, and a multi-scale convolution kernel method is used to deeply analyze the swing head characteristic modal parameters. The multi-scale convolution kernel is a technology in which different size convolution windows are respectively applied to time series data, and local features in the machining process can be extracted at different time resolutions. For example, a small scale convolution kernel identifies high frequency vibration or instantaneous stress peak, and a large scale convolution kernel reveals overall machining trend or slow varying fluctuation. By fusing and analyzing the features at different scales, the response mode of the modal parameters at different times and frequencies is identified, and the time-frequency coupling relationship is analyzed. Based on this relationship, the coordinated constraint condition of the spindle feed rate and the swing angle offset is further established, that is, when the vibration frequency rises or the stress concentrates, the feed rate is automatically reduced, and the swing angle offset is finely adjusted to avoid the occurrence of unstable machining or surface quality decline. For example, when the torque fluctuation frequency is detected to be enhanced and accompanied by high amplitude vibration, the feed rate is immediately reduced from 1200 mm / min to 900 mm / min, and the swing angle adjustment amplitude is limited within ±0.1°, so as to buffer the load change and ensure the machining stability.

[0034] The outer loop is responsible for adjusting the overall cutting stability, taking the cutting stiffness coefficient as the key control target. In the machining process, the current cutting stiffness performance is estimated by collecting spindle current fluctuation, cutting force change and other information, and an iterative adjustment method is used for dynamic optimization. For example, when it is found that the current cutting stiffness is low, the outer loop will gradually increase the current control proportion or optimize the speed control strategy to enhance the stress response capability of the machining area, so that the stiffness is restored to the set range. This optimization process is not completed instantaneously, but slowly approaches the target combined with multiple rounds of machining data, so as to avoid system instability caused by too fast adjustment. For example, in the rough machining of a stainless steel workpiece, it is found that the tool load change is greater than the set threshold value 0.5 Nm, and the outer loop control continuously increases the servo response sensitivity by 5% for three times, successfully compresses the stiffness error to within 5% without affecting the machining surface.

[0035] In order to ensure that each parameter update is executed within a safe range, the outer loop further introduces a stability theorem to build a parameter adjustment boundary. Specifically, by means of vibration response change, energy input fluctuation and other information in the machining state, it is judged whether the current driving parameter is likely to cause an unstable trend. Once an abnormal trend (such as vibration increase) is detected, it is judged that the safe boundary is exceeded, the current update process is terminated, and the last set of stable parameters is restored. For example, in a certain stiffness adjustment process, if the spindle appears continuous high amplitude vibration, it is identified as an unstable risk, and the update operation is refused, and the driving parameter state before adjustment is restored, and the failure tag of the adjustment behavior is recorded for subsequent optimization path analysis.

[0036] In summary, the double closed-loop control architecture realizes high-precision control of the swing head micro-motion parameters through the multi-scale convolution kernel analysis and collaborative constraint generation of the inner loop; and realizes long-term stable maintenance of the macro control target through the cutting stiffness dynamic optimization and stability boundary constraint of the outer loop, forming an adaptive adjustment mechanism with fast response and safety boundary control capability.

[0037] Further, the method provided by the application embodiment further comprises: The nonlinear fuzzy PID unit includes three parallel operation channels, wherein the first parallel operation channel uses wavelet packet energy entropy to represent the cutting nonlinearity; the second parallel operation channel extracts the intrinsic mode function of the torque signal; and the third parallel operation channel establishes a convolutional neural network of the cutting edge micro-damage and the contact stress distribution.

[0038] In the application embodiment, the nonlinear fuzzy PID unit includes three parallel operation channels, which respectively perform dynamic sensing and processing on the vibration nonlinearity, the torque fluctuation characteristics and the tool damage state. Each channel uses an algorithm method matched with the current intelligent manufacturing processing environment to extract key feature parameters, and finally acts on the real-time adjustment of the proportional coefficient (Kp), the integral coefficient (Ki) and the differential coefficient (Kd) respectively, to realize nonlinear adaptive control of the driving parameters.

[0039] In the first parallel operation channel, the wavelet packet energy entropy method is used to identify the nonlinearity degree of the vibration in the processing. First, the three-axis vibration signals of the spindle are collected, and the signals are analyzed in the multi-scale frequency band through wavelet packet decomposition to decompose a series of high-frequency and low-frequency sub-signals; then the energy proportion of each sub-signal is calculated, and the overall energy entropy is calculated based on the energy distribution. The energy entropy reflects the complexity of the signal in the frequency domain, and the energy concentration indicates that the linear regularity is strong, and the uniform energy distribution indicates that the system is in a nonlinear vibration state. For example, when the machine tool produces multiple vibration modes due to unstable cutting load, the energy is diffused in multiple frequency bands, and the entropy value increases. The entropy value is mapped as the basis for adjusting Kp, and if the energy entropy is greater than a preset threshold (such as 1.5), Kp is reduced to slow down the proportional response, so as to avoid unstable processing caused by too fast response.

[0040] In the second parallel operation channel, the spindle torque real-time fluctuation signal is collected, and the empirical mode decomposition (EMD) is applied to extract the intrinsic mode function (IMF). First, the extreme points of the torque signal are determined, the upper envelope and lower envelope curves are constructed, the trend item is removed after taking the average, and the iteration is completed to extract the primary IMF component; the above process is repeated to obtain a plurality of IMF sequences. Each IMF reflects different frequency components, for example, IMF1 may correspond to high-frequency mechanical impact, and IMF3 may reflect the load change trend. The energy and frequency change rate of each IMF are calculated, and when there is a large mutation (such as a 30% energy surge) in a certain IMF, the index is used to adjust the differential coefficient Kd. Increasing Kd can enhance the response ability of the controller to rapid disturbances, for example, when detecting tool impact or instantaneous load increase caused by uneven material, quickly suppressing output changes to maintain system stability.

[0041] In the third parallel operation channel, a tool wear prediction model is constructed based on a convolutional neural network (CNN) to establish the mapping relationship between the micro-damage of the cutting edge and the contact stress distribution. The model training process includes three stages of data preprocessing, feature extraction and model optimization. First, the contact stress distribution map (generated from the strain gauge two-dimensional thermal map) during processing is collected as the CNN input; at the same time, the tool damage (such as the notch depth of the cutting edge and the wear width) under the corresponding working condition is measured by the online vision system or offline microscopic imaging, which is used as the output label. Thousands of image-label pairs are used for training, and the model learns the corresponding relationship between stress map features (such as stress concentration spots and boundary mutation lines) and damage degree. In the deployment stage, after the real-time stress map is input into the CNN, the predicted tool wear value is output, such as 0.07 mm. When the predicted value exceeds the set threshold, the output result is used to adjust the integral coefficient Ki to increase Ki to enhance the controller's ability to repair long-term deviations and delay the accumulation of errors caused by wear.

[0042] The processing results of the three channels are respectively used as input items of the fuzzy controller, which are matched with a preset rule base in a fuzzy control reasoning module (for example: "energy entropy high ∧ IMF fluctuation large ∧ tool wear serious → Kp↓, Ki↑, Kd↑"), and finally a set of real-time adjusted PID control parameters are output.

[0043] Further, the method provided by the application embodiment further comprises the following steps: The state space is defined as a deviation vector of the swing head characteristic modal parameter, and the action space is defined as an adjustment amount of a drive shaft current-speed coupling control parameter; a reward function is configured based on the state space and the action space, and the reward function is associated with a surface roughness index, a tool wear index and an energy consumption efficiency index.

[0044] In the embodiment of the present application, the state space is first defined as a deviation vector of the swing head characteristic modal parameter. In the actual machining process, multi-source dynamic data such as spindle vibration spectrum, cutting torque fluctuation and contact stress distribution are collected in real time, and the corresponding swing head characteristic modal parameters are extracted through tensor decomposition and modal identification method. The real-time modal parameters are subjected to difference operation with the target reference modal value set before machining to form a deviation vector. For example, if the center frequency of the current spindle vibration spectrum deviates from the target value by 30 Hz, or the stress distribution deviates from the target form by more than 10%, the deviation will be recorded as a component of the state vector in numerical form, which is used to accurately depict whether the current running state of the swing head is within the ideal control interval.

[0045] Then, the action space is defined as the adjustment amount of the drive shaft current-speed coupling control parameter. The action space constitutes two main dimensions, one is the current setting amount ΔI, which is used to adjust the output torque in the servo drive unit, and the other is the speed setting amount Δn, which is used to fine-tune the rotation speed of the spindle. These two parameters constitute a coupling control parameter pair, which determines the cutting driving force and energy distribution rhythm respectively. A finite step strategy is used to define the discrete action set in the action space, for example, the current increment step is set to ±0.2 A, and the speed adjustment step is set to ±50 rpm, forming the action space set A={(+0.2A,-50rpm),(-0.2A,+50rpm),…}, each combination represents an executable control adjustment instruction.

[0046] On the basis of the above-mentioned state space and action space, a reward function is configured for evaluating the influence effect of each current-speed control parameter adjustment on the machining result, thereby providing feedback basis for subsequent adjustment. The reward function is designed to integrate multiple key machining performance indicators by weighting, specifically including the surface roughness indicator Ra, the tool wear indicator VB and the energy consumption efficiency indicator η, and the construction form is . Wherein, Ra is the surface roughness of the workpiece after machining, representing the machining quality, which is usually obtained by a laser probe or a contact profilometer; VB represents the maximum wear width (in mm) of the tool relief surface, which can be estimated by online wear monitoring technology (such as strain gauge signal); η is the energy consumption efficiency, which is defined as the electric energy consumed per unit volume of machining, and the calculation method is the product of current and time in the machining process divided by the volume ratio, representing the energy efficiency level under the control strategy. Ra is the surface roughness of the workpiece after machining, representing the machining quality, which is usually obtained by a laser probe or a contact profilometer; VB represents the maximum wear width (in mm) of the tool relief surface, which can be estimated by online wear monitoring technology (such as strain gauge signal); η is the energy consumption efficiency, which is defined as the electric energy consumed per unit volume of machining, and the calculation method is the product of current and time in the machining process divided by the volume ratio, representing the energy efficiency level under the control strategy. an acceptable upper limit value of the tool wear, for example, set to 0.3mm, exceeding which indicates that the tool wears too fast or needs to be replaced. a weight coefficient of the surface roughness term, a weight coefficient of the tool wear term, a weight coefficient of the energy efficiency term, and the three weight coefficients are pre-set by a technical expert.

[0047] In each control cycle, the system selects a control action combination (ΔI, Δn) from the action space according to the current state deviation vector to update the driving parameters, collects the machining quality, tool state and energy consumption information after execution, and calculates the corresponding reward value R. Through continuous iteration of the above process and combined with the historical reward trend, the optimal current-speed coupling adjustment strategy is gradually learned, so that the swing head realizes the global balance between machining quality, tool life and energy efficiency under complex load and dynamic disturbance conditions.

[0048] Step S400: Introduce a preset machining precision index, combine the adaptive adjustment driving parameter set, and configure the driving shaft current-speed coupling control parameter for dynamic optimization.

[0049] In the embodiments of the present application, in order to improve the driving control precision in the turning-milling swing head machining process, first, a preset machining precision index is introduced as a control target and a constraint condition. The machining precision index includes a surface roughness limit value, a profile error upper limit and a path deviation tolerance, which are used to comprehensively reflect the control requirements for surface quality and geometric accuracy in the machining process. These precision indexes are set before machining and compared in real time during machining, to judge whether the current driving parameters meet the expected quality target.

[0050] On this basis, combined with the previously extracted adaptive adjustment driving parameter set, the combination form of the current setting amount and the speed setting amount is selected as the candidate adjustment scheme of the current driving behavior by analyzing the corresponding relationship between the swing head characteristic modal parameters and the machining feedback results. If the machining result does not meet the preset machining precision index, it is considered that the current driving parameter configuration has deviation, and it needs to be re-optimized.

[0051] To this end, a dynamic optimization operation of configuring the drive shaft current-rotation speed coupling control parameters is performed. The optimization process first fits the thermal-mechanical behavior of the tool-workpiece contact area through a digital twin model, generates a temperature field prediction distribution atlas to identify the local high temperature area and the thermal coupling sensitive section. Subsequently, based on the temperature field prediction result, combined with the spindle temperature rise rate and the cutting chatter threshold two constraint conditions, the adjustment step of the current and the rotation speed is jointly optimized, so that the drive parameter adjustment meets the accuracy target while avoiding process risks such as thermal deformation and structural vibration. Through the above process, the adjustment behavior of the drive shaft current and rotation speed is always around the machining accuracy target for dynamic convergence, ensuring high-quality and consistent control effect under complex working conditions.

[0052] Further, the method provided by the application embodiment further comprises: The temperature field prediction distribution atlas of the tool-workpiece contact area is drafted through the digital twin model; and based on the temperature field prediction distribution atlas, the adjustment step of the drive shaft current-rotation speed coupling control parameters is optimized under the constraint conditions including the spindle temperature rise rate and the cutting chatter threshold.

[0053] In the application embodiment, to realize the dynamic optimization of the drive shaft current-rotation speed coupling control parameters, a digital twin model is introduced to model and predict the thermal response of the tool-workpiece contact area. Based on the real-time acquisition of the spindle current, cutting force, feed rate, tool displacement, contact stress and temperature and other multi-source process data in the machining site, the digital twin model constructs a simulation system highly consistent with the physical machining process, and dynamically maps the actual machining state through the method of data driving and physical model fusion.

[0054] In the thermal behavior simulation, the digital twin model uses a three-dimensional unsteady heat conduction equation, combines a Gaussian heat source model to describe the local heat transfer process between the tool and the workpiece in the cutting process, and introduces the thermal conductivity, specific heat capacity, contact thermal resistance and other physical property parameters of the workpiece material to establish a prediction relationship of temperature change with time and space. In the simulation process, the dynamic changes of the tool path and the cutting depth are considered, and the temperature field prediction distribution atlas of the tool-workpiece contact area is generated through finite element calculation, which can accurately reflect the hot spot position, temperature gradient and local heat accumulation trend in the machining process.

[0055] On the basis of temperature field prediction, process stability constraint conditions are set. One is spindle temperature rise rate constraint, through thermocouple sensor to collect local temperature data of spindle, calculate the first derivative of temperature to time, judge whether there is a trend of thermal deformation; the second is cutting chatter threshold constraint, use three-axis acceleration sensor to obtain vibration data and perform fast Fourier transform (FFT) analysis, when the main frequency falls into the preset chatter sensitive frequency band (such as 150-300Hz) and the vibration energy abnormally enhances, it is determined that there is a risk of chatter.

[0056] Under the control of the above constraint conditions, the current adjustment step (ΔI) and the speed adjustment step (Δn) of the driving shaft are dynamically optimized. When the temperature rise rate is within the safe range and no high-frequency chatter signal is detected, ΔI is allowed to be set to ±0.4A and Δn is allowed to be set to ±100rpm to improve the control response speed; when the predicted temperature field shows that the hot spots are concentrated or the chatter index is close to the threshold, ΔI is automatically contracted to ±0.1A and Δn is automatically contracted to ±20rpm, entering the thermal-vibration protection mode to avoid exciting structural deformation or vibration instability.

[0057] In summary, the temperature field prediction map obtained through the digital twin model provides real-time physical basis for the optimization of driving parameters. Under the dual constraints of spindle temperature rise rate and cutting chatter threshold, the dynamic convergence of current and speed regulation amplitude is realized, so that the machining process meets the accuracy requirements while considering thermal stability and structural safety, thereby significantly improving the intelligence and adaptability of the driving control strategy.

[0058] Step S500: At the same time, the machining load mutation detection data, the machining surface quality data, and the tool wear monitoring data are used as multi-dimensional feedback information sources to determine the swing head driving parameter adaptive adjustment instruction.

[0059] In the embodiments of the present application, in order to realize adaptive adjustment of the swing head driving parameters, multi-dimensional feedback information in the machining process is collected as the adjustment basis, mainly including machining load mutation detection data, machining surface quality data, and tool wear monitoring data. In the specific process, the cutting force signal is continuously monitored, and the absolute value of the second derivative is calculated to judge the acceleration characteristics of the force signal change, and the adaptive threshold is set according to the current machining conditions to judge whether there is a sudden load anomaly. At the same time, the chip shape information is obtained to identify abnormal cutting states such as irregular chip breaking and built-up edge generation. The load mutation recognition result and the abnormal chip working condition are jointly analyzed, and the surface roughness data and the tool wear state are combined to weight and balance all feedback information according to their change intensity and importance, and finally an optimal adjustment value for adjustment is determined. The optimal value is used to trigger the adjustment instruction of the swing head driving parameters, so that the system can respond effectively in real time when facing complex disturbances or unstable cutting states, and maintain machining accuracy and process stability.

[0060] Further, the method provided by the application embodiment further comprises: The absolute value of the second derivative of the cutting force signal is monitored, and the adaptive threshold value obtained according to the current machining condition is used to make a machining load mutation decision; meanwhile, the change in chip shape is captured, and a chip breaking abnormal condition and a built-up edge abnormal condition are identified; according to the chip breaking abnormal condition and the built-up edge abnormal condition and the machining load mutation decision information, the multi-dimensional feedback information source is weighted and balanced to determine an optimal adjustment value, and the optimal adjustment value is used to configure the swing head driving parameter adaptive adjustment instruction.

[0061] In the application embodiment, first, the cutting force signal is collected in real time through a force sensor during machining, and the absolute value of the second derivative of the signal is calculated using the central difference method to reflect the acceleration of load change. When the machining state is stable, the absolute value of the second derivative is small; when a hard point is cut in, the feed speed suddenly changes or the material is uneven, the cutting force changes suddenly, and the value of the second derivative significantly increases. To determine whether such a change constitutes a machining load mutation event, an adaptive threshold value is set. The threshold value is dynamically adjusted based on the current machining condition. Specifically, according to the current cutting depth, feed speed and spindle speed, a preset condition-threshold correspondence rule table is matched. When the real-time calculated second derivative value is greater than the threshold value obtained by looking up the table under the current condition, it is determined that the machining load is mutated, that is, the current swing head driving parameter may not meet the actual load requirement, and needs to be dynamically corrected.

[0062] Second, to further identify cutting abnormalities, an industrial camera is used to capture a chip image sequence, and a gray scale analysis method is used to judge the continuity and shape of the chip. When the chip in the image is obviously broken and the length change fluctuation exceeds the normal range, it is identified as a chip breaking abnormal condition; if the gray scale value of the rake face area of the tool in the image rapidly increases and the structure is blurred, it is determined as a built-up edge abnormal condition.

[0063] Meanwhile, two auxiliary dimension information are obtained, one is to scan the machining surface through a laser sensor to obtain the Ra value and judge whether the current surface roughness exceeds the standard limit value; the other is to analyze the stress change trend through a strain gauge installed on the tool shank to judge whether there is obvious tool wear (such as VB value estimation exceeding 0.3mm).

[0064] Then the cutting force mutation judgment result, the chip abnormal state, the surface roughness deviation and the tool wear degree are taken as feedback inputs, and a weighted average method is used for fusion analysis. A fixed weight is set for each feedback, such as 0.4 for cutting force mutation, 0.3 for chip state abnormality, 0.15 for roughness and 0.15 for wear, and the comprehensive disturbance score is calculated after each index is normalized.

[0065] According to the score, the corresponding optimal adjustment value is selected from a group of preset adjustment step matching intervals, which is used to correct the current current and speed regulation range of the swing head. For example, when the score is higher than 0.7, the current adjustment step is set to ±0.1A, and the speed step is set to ±20rpm; when the score is lower than 0.3, the adjustment step is widened to ±0.4A and ±100rpm.

[0066] Finally, the optimal adjustment value is taken as a core parameter to generate a swing head driving parameter adaptive adjustment instruction, complete dynamic compensation of the driving output, and realize stable control and precision guarantee of the machining process under complex disturbance.

[0067] Further, the method provided by the application embodiment further comprises: After the machining is completed, the deviation distribution of the workpiece surface topography data and the standard workpiece surface topography data is extracted, and the deviation distribution is mapped to the swing head driving parameter adjustment history record to establish a quality association knowledge base; based on the quality association knowledge base, the initial strategy of the nonlinear fuzzy PID unit is updated and strengthened through transfer learning.

[0068] In the application embodiment, after the workpiece machining is completed, the workpiece surface is scanned by a white light interferometer for high-precision three-dimensional scanning to obtain the workpiece surface topography data of the workpiece. At the same time, the standard workpiece surface topography data set in advance is called as a reference model. The spatial alignment processing is performed on the two groups of point cloud data through a least square point cloud registration algorithm (ICP algorithm), the spatial difference between each pair of corresponding points is calculated after the coordinates are unified by using the Euclidean distance calculation method, the deviation distribution covering the entire workpiece area is formed, the micro error of the surface at different positions is represented in numerical form, and the actual machining deviation between the workpiece and the standard topography is reflected.

[0069] Subsequently, the deviation distribution is matched with the swing head driving parameter adjustment history record in the entire machining process. The history record includes the current setting amount ΔI, the speed setting amount Δn, and the control output of the nonlinear fuzzy PID unit, that is, the real-time adjusted proportional coefficient (Kp), integral coefficient (Ki) and differential coefficient. Through a position-time synchronous mapping algorithm, the local area with larger error in the deviation distribution is matched with the corresponding machining time period, and then the deviation feature is mapped with the corresponding control parameter sequence.

[0070] After the above mapping is completed, a quality association knowledge base is constructed to store instance entries composed of "error distribution + swing drive parameter adjustment history" in a structured manner. The knowledge base summarizes typical error distribution and control behavior patterns in multi-batch processing, such as "low Kp in the Ra overrun area, Ki response delay, and frequent Kd fluctuations", or "severe Δn jump in the profile overcut area, and insufficient current adjustment", etc. These records are indexed in a searchable manner, with data accumulation, comparative evaluation and experience transfer capabilities.

[0071] In subsequent processing tasks, the quality association knowledge base is called based on the current working condition characteristics, and the control strategy is updated through a transfer learning method. Specifically, the K-Nearest Neighbor algorithm (KNN) is used to select the historical case with the highest similarity to the current working condition, extract its control parameter behavior, and initialize the nonlinear fuzzy PID unit at the present stage through a parameter transfer strategy, i.e. reset the initial proportional coefficient (Kp), integral coefficient (Ki) and differential coefficient (Kd) range or reference value, so that it is more suitable for the current working condition requirements. In the early stage of processing, these transfer parameters serve as the starting point for optimization, improving the initial accuracy and response efficiency of the control strategy.

[0072] Through the above process, a control optimization path based on error distribution and swing drive parameter adjustment history is realized, the control strategy is transformed from empirical rules to data-driven, and the self-learning ability and strategy adaptability of the nonlinear fuzzy PID unit in complex processing environment are improved.

[0073] In the embodiments of the present application, as described above, the embodiments of the present application have at least the following technical effects: The present application collects multi-source dynamic data including spindle vibration spectrum, cutting torque fluctuation and contact stress distribution in the process of turning and milling swing, constructs a dynamic parameter matrix, performs tensor decomposition based on the dynamic parameter matrix, extracts swing feature modal parameters, configures an adaptive adjustment drive parameter set based on a nonlinear fuzzy PID unit according to the swing feature modal parameters, and the nonlinear fuzzy PID unit cooperates with the double closed-loop control architecture. The preset machining precision index is introduced, the drive shaft current-speed coupling control parameter is dynamically optimized by combining the adaptive adjustment drive parameter set, and the swing drive parameter adaptive adjustment instruction is determined by taking the machining load mutation detection data, the machining surface quality data and the tool wear monitoring data as multi-dimensional feedback information sources. The present application solves the technical problem that the swing drive parameter cannot be adjusted in real time according to the processing state in the prior art, and through the parameter modeling and optimization control mechanism driven by multi-source dynamic data, the technical effects of improving the response accuracy and processing stability of the swing drive are achieved.

[0074] Example 2, based on the same inventive concept as the adaptive control method for milling and turning head drive parameters in the previous examples, such as... Figure 2 As shown, this application provides an adaptive control system for the drive parameters of a milling and turning head. The system and method embodiments in this application are based on the same inventive concept. The system includes: The dynamic parameter matrix construction module 11 is used to collect multi-source dynamic data, including spindle vibration spectrum, cutting torque fluctuation, and contact stress distribution, during the milling head machining process, and construct a dynamic parameter matrix. The tensor decomposition module 12 is used to perform tensor decomposition based on the dynamic parameter matrix to extract the characteristic modal parameters of the milling head. The configuration module 13 is used to configure an adaptive adjustment drive parameter set based on a nonlinear fuzzy PID unit according to the characteristic modal parameters of the milling head. The nonlinear fuzzy PID unit operates in conjunction with a dual closed-loop control architecture. The dynamic optimization module 14 is used to introduce a preset machining accuracy index and, in combination with the adaptive adjustment drive parameter set, configure the drive shaft current-speed coupling control parameters for dynamic optimization. The adaptive adjustment command determination module 15 is used to simultaneously use machining load change detection data, machining surface quality data, and tool wear monitoring data as multi-dimensional feedback information sources to determine the adaptive adjustment command for the milling head drive parameters.

[0075] Furthermore, the system is also used to implement the following functions: The inner loop of the dual closed-loop control architecture analyzes the time-frequency coupling relationship of the feature modes through multi-scale convolution kernels to generate the collaborative constraint conditions of the spindle feed rate and the swing angle offset; the outer loop of the dual closed-loop control architecture dynamically optimizes the cutting stiffness coefficient and introduces the stability theorem to construct the parameter adjustment boundary.

[0076] Furthermore, the system is also used to implement the following functions: A triaxial vibration sensor array is arranged radially on the spindle, and the dynamic stress distribution at the tool clamping end is measured by embedded strain gauges. The instantaneous torque fluctuation data of the spindle servo motor is collected synchronously. The collected raw data is processed by wavelet packet denoising to obtain the multi-source dynamic data.

[0077] Furthermore, the system is also used to implement the following functions: The nonlinear fuzzy PID unit includes three parallel computing channels. The first parallel computing channel uses wavelet packet energy entropy to characterize the cutting nonlinearity; the second parallel computing channel extracts the intrinsic mode function of the torque signal; and the third parallel computing channel establishes a convolutional neural network for the micro-damage of the cutting edge and the distribution of contact stress.

[0078] Furthermore, the system is also used to implement the following functions: Define the state space as the deviation vector of the swing head characteristic modal parameter, and the action space as the adjustment amount of the drive shaft current-speed coupling control parameter; configure a reward function based on the state space and the action space, and the reward function is associated with the surface roughness index, the tool wear index and the energy consumption efficiency index.

[0079] Further, the system is also used to implement the following functions: Through the digital twin model, a temperature field prediction distribution atlas of the tool-workpiece contact area is formulated; based on the temperature field prediction distribution atlas, the adjustment step of the drive shaft current-speed coupling control parameter is optimized under the constraint conditions including the spindle temperature rise rate and the cutting chatter threshold.

[0080] Further, the system is also used to implement the following functions: The absolute value of the second derivative of the cutting force signal is monitored, and the adaptive threshold obtained according to the current machining condition is used for machining load mutation decision; at the same time, the chip shape change is captured, and the chip breaking abnormal condition and the built-up edge abnormal condition are identified; according to the chip breaking abnormal condition and the built-up edge abnormal condition, and the machining load mutation decision information, the multi-dimensional feedback information source is weighted and balanced to determine an optimal adjustment value, and the optimal adjustment value is used to configure the swing head drive parameter adaptive adjustment instruction.

[0081] Further, the system is also used to implement the following functions: After the machining is completed, the deviation distribution of the workpiece surface topography data and the standard workpiece surface topography data is extracted, and the deviation distribution is mapped to the swing head drive parameter adjustment history record to establish a quality association knowledge base; based on the quality association knowledge base, the initial strategy of the nonlinear fuzzy PID unit is updated and strengthened through transfer learning.

[0082] Embodiment three, based on the inventive concept of the swing-milling head drive parameter adaptive control method in the foregoing embodiments, the present application also provides an electronic device, comprising: at least one processor; a memory in communication connection with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the steps of any one of the methods in the foregoing embodiment one.

[0083] Figure 3 The structure schematic diagram of the exemplary electronic device of the present application is shown in FIG. 1. Figure 3In particular embodiments, bus 300 represents what can be one or more busses (such as an address bus, data bus, and control bus) and bridges (such as an address bridge, data bridge, and control bridge) that are used to connect the various circuits of the device together. Bus 300 would connect, for example, the processor 302 and the memory 304, which can represent one or more processors and one or more memories, respectively. Bus 300 can also connect various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art, and thus, will not be described further. Bus interface 305 provides an interface between bus 300 and receiver 301 and transmitter 303. Receiver 301 and transmitter 303 can be the same device, i.e., a transceiver, providing a unit for communicating with various other apparatus over a transmission medium. Processor 302 is responsible for managing bus 300 and general processing, while memory 304 can be used to store data for use by processor 302 during execution of operations.

[0084] It should be noted that the foregoing merely illustrates the principles of the application. Various modifications and adaptations can occur to those skilled in the art in view of the above teachings without departing from the scope of the application. For example, while the application has been described specifically with respect to particular embodiments, the terms used are intended to include both explicit and implied printings and equivalents. Other variations and modifications of the examples disclosed herein can be made based on the description disclosed herein, and the description is to be regarded as illustrative rather than restrictive of the disclosure. It is the intention that all such modifications and variations that fall within the scope of the disclosure are to be regarded as within the scope of the disclosure. Thus, numerous specific details are set forth in this disclosure to provide a thorough understanding of the application. However, it is to be understood that the application can be practiced without these specific details. In other instances, well-known methods, procedures, components, and networks have not been described in detail so as not to unnecessarily obscure aspects of the application. It is intended that the application not be limited to the details shown and described herein, but include all such modifications, alterations, and variations that fall within the scope of the application.

[0085] The above description is intended to be illustrative and not restrictive. Many other modifications within the scope of the application will be apparent to those of skill in the art upon reviewing the above description. The scope of the application should, therefore, be determined not with reference to the above description, but instead with reference to the appended claims, along with their full scope of equivalents.

[0086] The description and drawings are merely illustrative of the application and do not limit the scope of the application. Variations and modifications can be made to the application without departing from the spirit and scope of the application. It is intended that the application encompass all such variations and modifications as fall within the scope of the application.

Claims

1. A method for adaptive control of turning and milling swing drive parameters, characterized in that, The method comprises: Collecting multi-source dynamic data including spindle vibration frequency spectrum, cutting torque fluctuation and contact stress distribution during the swing head machining of the milling machine, and constructing a dynamic parameter matrix; Performing tensor decomposition based on the dynamic parameter matrix to extract swing head characteristic modal parameters; According to the swing head characteristic modal parameters, configuring an adaptive adjustment driving parameter set based on a nonlinear fuzzy PID unit, and the nonlinear fuzzy PID unit is cooperatively operated with a double closed-loop control architecture; Introducing a preset machining precision index, combining the adaptive adjustment driving parameter set, and configuring driving shaft current-speed coupling control parameters for dynamic optimization; Meanwhile, taking machining load mutation detection data, machining surface quality data and tool wear monitoring data as multi-dimensional feedback information sources, swing head driving parameter adaptive adjustment instructions are determined.

2. The adaptive control method of the turning and milling swing head driving parameters according to claim 1, characterized in that, The inner loop of the double closed-loop control architecture analyzes the time-frequency coupling relationship of the characteristic modal through a multi-scale convolution kernel to generate a cooperative constraint condition of the spindle feed rate and the swing angle offset; The outer loop of the double closed-loop control architecture dynamically optimizes the cutting stiffness coefficient, and introduces a stability theorem to construct a parameter adjustment boundary.

3. The adaptive control method of a turning and milling swing drive parameter according to claim 1, characterized in that, Collecting multi-source dynamic data including spindle vibration frequency spectrum, cutting torque fluctuation and contact stress distribution during the swing head machining of the milling machine, and constructing a dynamic parameter matrix, the method comprises: Arranging a three-axis vibration sensor array radially on the swing head spindle, and measuring the dynamic stress distribution of the tool clamping end through embedded strain gauges, and synchronously collecting the instantaneous torque fluctuation data of the spindle servo motor; Collecting the collected raw data, and performing wavelet packet denoising processing to obtain the multi-source dynamic data.

4. The adaptive control method of a turning and milling swing drive parameter according to claim 3, characterized in that, The nonlinear fuzzy PID unit includes three parallel operation channels, wherein the first parallel operation channel uses wavelet packet energy entropy to represent the cutting nonlinearity; the second parallel operation channel extracts the intrinsic modal function of the torque signal; and the third parallel operation channel establishes a convolutional neural network of the cutting edge micro-damage and the contact stress distribution.

5. The adaptive control method of a turning and milling swing drive parameter according to claim 4, characterized in that, According to the swing head characteristic modal parameters, configuring an adaptive adjustment driving parameter set based on a nonlinear fuzzy PID unit, the method further comprises: Defining a state space as a deviation vector of the swing head characteristic modal parameters, and an action space as an adjustment amount of the driving shaft current-speed coupling control parameters; Based on the state space and the action space, a reward function is configured, and the reward function is associated with surface roughness index, tool wear index and energy efficiency index.

6. The adaptive control method of a turning and milling swing drive parameter according to claim 5, characterized in that, Configuring driving shaft current-speed coupling control parameters for dynamic optimization, the method comprises: Through a digital twin model, a temperature field prediction distribution map of the tool-workpiece contact area is drafted; Based on the temperature field prediction distribution map, under the constraint condition including the spindle temperature rise rate and the cutting chatter threshold, the adjustment step of the driving shaft current-speed coupling control parameters is optimized.

7. The adaptive control method of a turning-curling drive parameter according to claim 1, wherein, Taking machining load mutation detection data, machining surface quality data and tool wear monitoring data as multi-dimensional feedback information sources, swing head driving parameter adaptive adjustment instructions are determined, the method comprises: Monitoring the absolute value of the second derivative of the cutting force signal, and making machining load mutation decisions according to the adaptive threshold obtained according to the current machining condition; Meanwhile, the chip morphology change is captured, and the chip breaking abnormal working condition and the chip accumulation abnormal working condition are identified; According to the chip breaking abnormal working condition, the chip accumulation abnormal working condition, and machining load mutation decision information, the multi-dimensional feedback information source is weighted and balanced to adjust to determine an optimal adjustment value, and the optimal adjustment value is used to configure the swing head driving parameter adaptive adjustment instruction.

8. The adaptive control method of a turning and milling swing drive parameter according to claim 7, characterized in that, The method further includes: After the machining is completed, a deviation distribution of workpiece surface topography data and standard workpiece surface topography data is extracted, and the deviation distribution is mapped to a swing head driving parameter adjustment history record to establish a quality association knowledge base; Based on the quality association knowledge base, the initial strategy of the nonlinear fuzzy PID unit is updated and strengthened through transfer learning.

9. A system for adaptive control of turning and milling head drive parameters, characterized by The system includes: A dynamic parameter matrix construction module is configured to collect multi-source dynamic data including spindle vibration frequency spectrum, cutting torque fluctuation, and contact stress distribution in the turning-milling-swinging process to construct a dynamic parameter matrix; A tensor decomposition module is configured to perform tensor decomposition based on the dynamic parameter matrix to extract swing head characteristic modal parameters; A configuration module is configured to configure an adaptive adjustment driving parameter set based on a nonlinear fuzzy PID unit according to the swing head characteristic modal parameters, and the nonlinear fuzzy PID unit is cooperatively operated with a double closed-loop control architecture; A dynamic optimization module is configured to introduce a preset machining precision index, combine the adaptive adjustment driving parameter set, and configure a driving shaft current-speed coupling control parameter for dynamic optimization; An adaptive adjustment instruction determination module is configured to simultaneously use machining load mutation detection data, machining surface quality data, and tool wear monitoring data as multi-dimensional feedback information sources to determine a swing head driving parameter adaptive adjustment instruction.

10. An electronic device, comprising: The electronic device includes: A memory is configured to store executable instructions; A processor is configured to execute the executable instructions stored in the memory to implement the turning-milling-swinging driving parameter adaptive control method in any one of claims 1-8.

Images