Real-time feedback based self-adaptive adjusting system for clamping pressure of precision boring and milling machine

Abstract

The present application belongs to the field of automation control technology, and particularly relates to a precision boring-milling machine clamping pressure self-adaptive adjusting system based on real-time feedback. The system collects the cutting force and vibration signals naturally generated in the machining process to evaluate the stability of the clamping state in real time. Based on the performance in the steady cutting stage, the system uses a credit accumulation mechanism to automatically and gradually reduce the clamping pressure to the minimum necessary level required to maintain stable machining, and ensures the smoothness and safety of the adjusting process through slope limiting. This method effectively avoids the problems of workpiece over-clamping deformation, excessive energy consumption and clamp wear caused by traditional fixed pressure setting, while ensuring the reliability of machining, significantly improving the precision and reducing the operating cost. Moreover, the system can be integrated into existing numerical control boring-milling equipment at low cost, and has good engineering practicability.

Description

Technical Field

[0001] This application relates to the field of automation control technology, specifically to an adaptive adjustment system for clamping pressure of a precision boring and milling machine based on real-time feedback. Background Technology

[0002] In precision boring and milling, the stability of workpiece clamping directly affects machining accuracy, surface quality, and process safety. The automatic clamping system of modern CNC boring and milling machines drives the clamping action by controlling the pressure of hydraulic oil or compressed air, i.e., clamping pressure, thereby clamping the workpiece. The current mainstream clamping pressure still adopts a static preset strategy, which mainly faces two mutually restrictive technical dilemmas, forming a pair of long-standing unresolved process paradoxes: (1) In order to cover the loosening or chattering of the workpiece under the worst working conditions, the preset pressure is usually much higher than the actual needs of most finishing stages. For thin-walled, slender, or weakly rigid workpieces, after machining and unloading, uncontrollable form and position errors are generated due to springback, which seriously affects key indicators such as roundness and flatness; maintaining unnecessary high pressure consumes a lot of hydraulic or pneumatic energy, resulting in energy waste; and aggravates the mechanical wear and fatigue of clamping components, shortening their service life. (2) To address the shortcomings of static control, some high-end equipment attempts to introduce chamber pressure sensors to achieve closed-loop control. However, the sensor integration cost is high and requires customized embedding into the fixture body, which is expensive. Furthermore, the engineering complexity and cost of modifying the fixture structure are extremely high, making it uneconomical for most manufacturers. The sensor measures the nominal clamping force applied by the fixture, rather than the actual effective constraint borne by the workpiece. Due to factors such as contact nonlinearity and changes in the friction coefficient, the two differ significantly, resulting in the measured value failing to truly reflect the clamping stability. Therefore, the field has long faced a dilemma: either adopt a simple but low-performance static over-constraint strategy, bearing the cost of workpiece deformation and energy consumption; or attempt to adopt an expensive, fragile, and unreliable direct force measurement closed loop, facing high costs and implementation risks. Summary of the Invention

[0003] This application provides an adaptive intelligent adjustment method for clamping pressure that does not rely on any direct clamping force measurement sensor. The technical problem to be solved is that, in the absence of direct, reliable and economical measurement of the actual clamping force on the workpiece, the method uses signals such as vibration and cutting force naturally generated during the machining process to determine in real time whether the current clamping state is stable, and dynamically adjusts the output pressure of the fixture system accordingly, so that it is always maintained at the minimum necessary pressure level that can ensure machining safety and stability while being as low as possible, thereby reducing workpiece deformation, reducing energy consumption and extending the service life of the fixture.

[0004] In view of the above problems, this application provides an adaptive adjustment system for clamping pressure of a precision boring and milling machine based on real-time feedback, the system comprising: The signal acquisition module is used to synchronously and in real time acquire the first type of signal and the second type of signal during the boring and milling process. The first type of signal is processed to characterize the vibration amplitude of the dynamic stability of the workpiece-fixture system, and the second type of signal is processed to characterize the cutting force of the instantaneous cutting load. The status feedback module is used to calculate the ratio of the processed first type to the second type of signal in real time, and dynamically correct it by using the historical stability information of the processing process to generate a feedback signal that represents the current processing status. The pressure calculation module is used to normalize and compare the current machining state feedback signal with the feedback signal reference to obtain the deviation index characterizing the current machining stability. Based on the deviation index, the optimal clamping pressure target value matching the current machining state is calculated. The feedback signal reference is the moment when the current machining task first enters the steady-state cutting stage. The moment when the machining state feedback signal first enters the steady-state cutting stage is the start time of the first workpiece machining entering the steady-state cutting stage in the current machining task. The instruction generation module is used to apply ramp limiting processing to the calculated optimal clamping pressure target value to obtain pressure control instructions. The command output module is used to output pressure control commands to the clamping actuator to control its output of corresponding clamping pressure, thereby realizing dynamic adaptive adjustment of clamping pressure.

[0005] One or more technical solutions provided in this application have at least the following technical effects or advantages: 1) Through a steady-state credit scoring mechanism, the clamping pressure is automatically and gradually reduced to the minimum level required to maintain stability based solely on long-term stable cutting performance during machining, without manual intervention or preset thresholds. This method fundamentally avoids the problems of thin-walled part deformation, energy waste, and fixture wear caused by over-constraint in traditional fixed pressure settings. It improves machining accuracy and reduces operating costs while ensuring machining safety.

[0006] 2) By utilizing existing or easily installed general-purpose sensors on boring and milling machines, cutting force and workpiece-fixture system vibration signals are collected as the basis for judging the clamping status. This enables the adaptive clamping function to be widely applied at a lower cost and with higher reliability to the vast majority of existing boring and milling machines that do not have dedicated force sensors pre-installed, thus solving the two major technical dilemmas that currently hinder each other.

[0007] 3) The credit scoring mechanism ensures that the system's response to risks is to clear credits, suspend optimization, maintain the current state and re-observe, thus avoiding new instability caused by overreacting to short-term disturbances.

[0008] 4) Clamping pressure is no longer a fixed process parameter, but a control variable that is dynamically optimized in real time based on actual processing, thus transforming clamping control from an empirical static setting to a data-driven dynamic optimization process.

[0009] In summary, this invention provides an adjustment method for clamping pressure that does not rely on direct clamping force monitoring but can achieve adaptive optimization based on machining performance feedback. It takes into account the comprehensive requirements of high stability, high precision and high reliability for precision boring and milling, and has significant engineering practical value. Attached Figure Description

[0010] Figure 1 This is a schematic diagram of the structure of a precision boring and milling machine clamping pressure adaptive adjustment system based on real-time feedback, provided in an embodiment of this application.

[0011] Explanation of reference numerals in the attached diagram: Signal acquisition module 10, Status feedback module 20, Pressure calculation module 30, Command generation module 40, Command output module 50. Detailed Implementation

[0012] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application.

[0013] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application will be provided in conjunction with the accompanying drawings. The described embodiments should not be considered as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0014] This application provides an adaptive adjustment system for clamping pressure of a precision boring and milling machine based on real-time feedback, such as... Figure 1 As shown, the system includes: The signal acquisition module 10 is used to synchronously and in real time acquire a first type of signal and a second type of signal during the boring and milling process. The first type of signal, after processing, is used to characterize the vibration amplitude of the dynamic stability of the workpiece-fixture system, and the second type of signal, after processing, is used to characterize the cutting force of the instantaneous cutting load.

[0015] Specifically, in this embodiment, the clamping pressure of the boring and milling machine refers to the working pressure in the hydraulic or pneumatic chamber corresponding to the normal clamping force applied by the fixture system to the workpiece surface or positioning reference surface. This pressure directly determines the magnitude of the clamping force and is used to characterize the degree to which the workpiece is securely fixed. Achieving dynamic optimization of the clamping pressure is key to improving the quality and efficiency of precision boring and milling while ensuring machining stability. The pressure sensor can be installed in the hydraulic circuit or pneumatic circuit of the cylinder to measure the clamping pressure P. However, the workpiece stability depends on the actual clamping force F, which is affected by P, piston area wear, and system efficiency. The fixture clamping force obtained by converting the monitored P cannot fully represent F. The clamping force sensor monitoring the actual clamping force F needs to be integrated into the clamping body. Industrial-grade high-precision force sensors typically cost thousands to several thousand RMB per unit. A boring and milling machine has multiple clamping points; installing sensors at all points would result in extremely high total costs for sensor hardware, custom clamping, wiring, signal conditioning modules, and installation and debugging, especially impractical for multi-station, quick-change clamping systems. This embodiment utilizes the standard sensing resources already equipped in the machine tool's CNC system—cutting force and vibration signals—to infer whether the current clamping state is sufficient. Increased cutting force and decreased vibration indicate sufficient actual clamping force F; normal cutting force and increased vibration indicate insufficient actual clamping force F. This bypasses direct dependence on clamping pressure P and clamping force F, achieving low cost and intelligent operation.

[0016] Furthermore, during the boring and milling process, the first type of signal and the second type of signal are simultaneously acquired in real time. The first type of signal, after processing, is used to characterize the vibration amplitude of the workpiece-fixture system's dynamic stability. The second type of signal, after processing, is used to characterize the cutting force of the instantaneous cutting load, including: At the initial stage of boring and milling machine machining, the following operations are performed simultaneously: By deploying vibration sensing units on the workpiece-fixture system, the original vibration signals reflecting the mechanical vibration of the workpiece-fixture system are collected in real time. By integrating a force sensing unit near the cutting edge of the tool, the original cutting force signal reflecting the interaction between the tool and the workpiece is acquired in real time. The original vibration signal is used as the original input of the first type of signal. After at least zero-point drift compensation and high-frequency noise filtering preprocessing, its features are extracted to obtain the vibration amplitude that characterizes the dynamic stability of the workpiece-fixture system. The original cutting force signal is used as the original input of the second type of signal. After the same preprocessing, its features are extracted to obtain the cutting force that characterizes the instantaneous cutting load. The data acquisition process is controlled by a unified trigger signal, ensuring physical synchronization of acquisition time.

[0017] Specifically, the workpiece-fixture system refers to the mechanically coupled body formed by the workpiece and the fixture during boring and milling. The vibration sensing unit should be located as close as possible to the workpiece-fixture interface and can be a high-sensitivity single-axis or multi-axis piezoelectric accelerometer or MEMS vibration sensor, and should never be installed at the workpiece-fixture connection interface. The force sensing unit can be a Kistler rotary force gauge, a tool holder force sensor, or indirectly obtain cutting force information through spindle motor current / power inversion. Its physical location should be as close as possible to the contact area between the cutting edge and the workpiece, such as being integrated inside the tool holder, at the front end of the spindle, or in the turret tool holder, to ensure that the collected signals can accurately reflect the instantaneous interaction force between the tool and the workpiece. The original vibration signal and the original cutting force signal serve as the input sources for the first and second types of signals, respectively. After acquisition, both signals undergo a unified preprocessing procedure: First, zero-point drift compensation is performed. Since the outputs of the vibration sensing unit and the force sensing unit may have zero-point offsets caused by temperature drift, device aging, or installation stress, a pre-set duration of the original signal is automatically acquired during the no-load operation phase after each machining task starts and before formally cutting into the workpiece. The time average of the original signal is then calculated as the dynamic zero-point reference for this machining task. In the subsequent steady-state cutting phase, all sampled values ​​are subtracted from this zero-point reference in real time to eliminate the influence of baseline drift on the vibration amplitude and cutting force feature extraction. High-frequency noise is then filtered out using a low-pass filter (the cutoff frequency is set according to the sampling rate and cutting frequency band, usually 500Hz-2kHz) or wavelet denoising methods to suppress irrelevant high-frequency components such as electromagnetic interference and mechanical shock. Based on this, the time-domain characteristics of the pre-processed vibration signal, such as RMS value, peak value, or envelope amplitude, are extracted to obtain the vibration amplitude characterizing the dynamic stability of the system. The lower this value, the higher the stiffness of the workpiece-fixture system and the more reliable the clamping. For the pre-processed cutting force signal, its composite amplitude, such as the Euclidean norm of the triaxial force or the absolute value of the main cutting direction component, is directly taken as a quantitative indicator of the instantaneous cutting load. To ensure strict alignment of the two types of signals in the time dimension, a unified hardware trigger signal from the CNC device is used to synchronously start the signal. The vibration and cutting force data acquisition module achieves physical-level time synchronization even if the original sampling periods of the two types of sensors differ. All raw data are timestamped with high precision, and in subsequent processing, the period of the lower sampling rate signal (e.g., cutting force) is used as a reference. A sliding window statistical analysis (e.g., calculating RMS every 1ms) is performed on the higher sampling rate signal (e.g., vibration), thereby generating synchronized vibration amplitude and cutting force characteristic sequences on a common time grid. This ensures a clear causal relationship between the two within each control cycle. The steady-state cutting stage does not mean the system has reached its optimal state, but rather that the cutting process is within a controllable and safe operating range without the risk of instability. Within this range, the system may still have overly conservative clamping pressure settings.The adaptive mechanism of this invention utilizes this safety window to dynamically explore the minimum necessary clamping force to meet the machining requirements without sacrificing stability, thereby improving machining accuracy, extending fixture life and reducing energy consumption.

[0018] The status feedback module 20 is used to calculate the ratio of the processed first type and second type signals in real time, and dynamically correct it by using the historical stability information of the processing process to generate a feedback signal that represents the current processing status.

[0019] Specifically, the vibration amplitude A, which characterizes the system response, will be used to... v The cutting force F, which characterizes the external excitation c Perform real-time ratio calculations to obtain the basic feedback quantity R'=A v / F c This ratio essentially reflects the dynamic compliance of the workpiece-fixture system at the current moment, i.e., the vibration level excited by a unit cutting force, and is a direct indicator of the system's dynamic stability. The basic feedback quantity R' is susceptible to instantaneous disturbances. A dynamic correction mechanism based on historical stability information from the machining process is introduced. When a sustained increase in high-frequency energy characteristic of system chatter is detected, even if the current R' value is not high, it is amplified through a dynamic weighting coefficient, allowing the feedback signal to more sensitively predict rather than merely respond to instability risks. The dynamically corrected signal, denoted as R, is the core feedback control signal. It is no longer a simple instantaneous physical quantity ratio, but an intelligent stability assessment indicator that integrates current dynamic performance with historical trend judgments. An increase in this indicator R clearly indicates that the current clamping constraint is becoming insufficient relative to the system's historical health state and current machining requirements; a decrease indicates that the constraint may have excessive optimization space. This provides a unique and reliable basis for subsequent decisions based on the minimum constraint criterion. In the steady-state cutting stage, by integrating the instantaneous dynamic response A... v / F c Intelligent feedback signals are constructed based on historical stability trends, and these signals drive the clamping pressure to dynamically optimize towards the direction of minimum necessary constraints.

[0020] Furthermore, a feedback signal characterizing the current processing state is generated, including: Based on the preprocessed vibration amplitude, a short-time Fourier transform is performed to obtain the current spectrum. The centroid frequency and spectral spread of the current spectrum are then calculated. The centroid frequency represents the average position of the vibration energy on the frequency axis, and the spectral spread represents the degree of dispersion of the vibration energy around the centroid frequency. Based on the current center of gravity frequency and spectral spread, the current instability trend index is calculated in real time, as shown in the following formula: ; Among them, FDI currentThe current instability trend index is represented by ε, which is a small bias to prevent division by zero. From the moment the workpiece enters the steady-state cutting stage, continuously record the current instability trend index and calculate its moving average μ within the sliding time window in real time. fdi and moving standard deviation σ fdi Calculate the adaptive weighting coefficient w fdi The formula is as follows: ; Where η is the sensitivity gain coefficient, and the value of η ranges from [0.2, 1]. The feedback signal for the current processing status is calculated using the following formula: ; Among them, R current This is a feedback signal for the current processing status.

[0021] Specifically, to overcome the lag and misjudgment risk caused by relying solely on time-domain signal ratios for state judgment in existing technologies, and to further improve the foresight and robustness of clamping constraint effectiveness assessment, an innovative instability trend prediction mechanism based on the dynamic evolution characteristics of the vibration spectrum is introduced when generating feedback signals characterizing the current processing state. This mechanism is then deeply integrated with the instantaneous dynamic response to form an intelligent feedback quantity with adaptive weights. A forward-looking instability trend quantification method that does not rely on any preset fixed thresholds is proposed and implemented for the first time. Its core lies in treating the vibration signal as a dynamically evolving system, tracking the evolution of its overall spectral shape in real time, and extracting physical characteristics highly sensitive to instability, thereby achieving early warning during the flutter incubation stage. The system uses a Hanning window with M sampling points to analyze the preprocessed vibration amplitude time series A. v [n] is used for frame segmentation, and the Hanning window w[n] is defined as: n = 0, 1, 2, ..., M-1; Where n is the local index of the sampling point within the window, and M is the length of the window, i.e., the total number of sampling points contained in one frame of signal. The preprocessed vibration amplitude time series A... v [n] is divided into several consecutive frames, and the m-th frame is denoted as . Where m is the frame number, n = 0, 1, 2, ..., M-1 is the intra-frame sample index. A windowing operation is performed on each frame: n = 0, 1, 2, ..., M-1; in, The signal is after windowing. The capability U of the window function is defined as: ; Subsequently, windowed signals were applied to each frame. Performing an M-point Discrete Fourier Transform (DFT) yields a set of M discrete complex spectra, each corresponding to one of the M discrete frequency points, also known as frequency units. The complex spectra are as follows: k = 0, 1, 2, ..., M-1; in, Here, k is the imaginary unit, and k is the frequency index, corresponding to the k-th frequency element in a set of discrete complex frequency spectra. Let be the complex spectral value of the m-th frame in the k-th frequency unit. Based on this, calculate the power spectral density estimate after windowed energy normalization and frequency resolution correction: k = 0, 1, 2, ..., M-1; in, The energy of the m-th frame in the k-th frequency unit, and the capability of the U-window function. For frequency resolution, f s The sampling frequency of the vibration signal. For power spectral density estimation, its unit is (m / s) 2 ) 2 / Hz.

[0022] Based on a set of discrete complex spectra obtained by discrete Fourier transform For k=0, 1, 2, ..., M-1, the system calculates the corresponding power spectral density estimate. To focus on the effective frequency band related to structural dynamic stability, a preset analysis frequency band [f] is set. min f max Typically, f min =10Hz, f max =2000Hz. This frequency band is a continuous physical frequency range and needs to be mapped to a discrete spectrum index: , ; Where Δf is the frequency resolution. and These are the floor function and floor function, respectively, for the frequency boundary f. min f max Mapped to a valid integer spectrum index. The current spectrum is a single frame used for feature computation. For the m-th frame signal, in the index interval k=k min k min +1, ..., k max Internally, based on a single frame Calculate the following two key morphological features: The centroid frequency f of the vibration signal in the m-th frame within the preset frequency band c Unit: Hz ; The spectral spread σ of the vibration signal in the m-th frame within the preset frequency band f Unit: Hz ; Where f[k] = k·Δf is the frequency value of the k-th frequency unit, and Δf is the frequency resolution. The centroid frequency of the vibration signal in the m-th frame is calculated. and spectral diffusivity Then, the system constructs the current instability trend index FDI based on the features of this frame. (m) , ; Here, ε is a small offset to prevent division by zero. From the moment the workpiece enters the steady-state cutting stage, the system continuously records the FDI of each frame. (m) Suppose the sliding time window contains the most recent N... w Frame, typically, N w For sampling data ranging from 0.5 to 2 seconds, when processing the m-th frame, based on the samples within the window {FDI} (m-Nw+1) FDI (m) Calculate the moving average With moving standard deviation ,Right now: , , Where i is the traversal index within the sliding time window, and its value ranges from 0 to i to N. w -1. Based on the above statistics, the adaptive weighting coefficients are dynamically calculated as follows: , Where η∈[0.2,1] is the sensitivity gain coefficient, with a preferred value of 0.5. Finally, the system generates a feedback signal R of the current processing state. (m) The calculation formula is as follows: , in, Let be the root mean square value of the vibration amplitude time series of the m-th frame. The cutting force is synchronously acquired in the m-th frame. The sensitivity gain coefficient is a key design variable used to quantify the system's response strength to instability trends. During the cutting process, the characteristic changes in the initial stage of chatter are often weak and easily masked by process noise. If the original vibration-force ratio is directly used as the feedback signal, it is difficult to achieve early warning while ensuring a low false alarm rate. Therefore, this invention introduces an adaptive weighting mechanism to standardize the deviation of the current instability trend index FDI from the historical steady-state distribution. As an enhancement basis, the amplification ratio is controlled by η. To verify the rationality of the sensitivity gain coefficient η, a systematic experiment was conducted under various typical cutting conditions, covering end milling processes of 45# steel, 6061 aluminum alloy, and Ti-6Al-4V titanium alloy, with a total of 90 valid tests completed. The results show that when η < 0.2, the system is not sensitive to early instability, with an average detection rate of less than 82%, which is difficult to meet the needs of industrial applications; when η > 1, the false alarm rate is > 6%, and it is prone to triggering false alarms due to chip impact or measurement interference in intermittent cutting or high-noise environments; when η ∈ [0.2, 1], the system can achieve a chatter detection rate of over 82% while maintaining a false alarm rate of less than 6%, with the best overall performance when η = 0.5, achieving an average detection rate of 93.4% and a false alarm rate of only 2.3%. This design eliminates the need to recalibrate the threshold for each machining scenario, significantly improving the versatility and practicality of the system. The sliding window length needs to balance the ability to capture chatter dynamic characteristics with statistical stability. During metal cutting, chatter typically develops from its inception to significant instability within 0.5–3 seconds, while transient disturbances, such as chip impact, generally last 0.3 seconds. Therefore, selecting a window of 0.5–2 seconds effectively smooths short-term disturbances and promptly reflects changes in instability trends. Simultaneously, this duration corresponds to 5–20 frames of data, sufficient to ensure the statistical reliability of the moving average and standard deviation, and meets the computational delay requirements for real-time monitoring. The preset analysis frequency band [10Hz, 2000Hz] is based on the physical characteristics of cutting vibration: components below 10Hz mainly originate from rigid body displacement or temperature drift and are unrelated to structural dynamic stability; components above 2000Hz are mostly sensor noise or transient chip impact, resulting in a low signal-to-noise ratio. The dominant modal frequencies and chatter regeneration frequencies of the machine tool-tool system are typically concentrated in the 100–1800Hz range. This frequency band covers the key frequency bands related to instability and effectively suppresses low-frequency drift and high-frequency interference, and has been adopted by international standards.

[0023] The pressure calculation module 30 is used to normalize and compare the current machining state feedback signal with the feedback signal reference to obtain a deviation index characterizing the current machining stability. Based on the deviation index, the optimal clamping pressure target value matching the current machining state is calculated. The feedback signal reference is the machining state feedback signal collected at the moment when the current machining task first enters the steady-state cutting stage. The moment when the first workpiece enters the steady-state cutting stage is the start time of the first workpiece machining in the current machining task entering the steady-state cutting stage.

[0024] Specifically, the first steady state is regarded as the process anchor point for this batch task, rather than assuming that the machining system remains absolutely unchanged throughout the task. By normalizing the deviation index, abnormal degradation trends relative to this anchor point are captured. This allows for highly sensitive detection of unexpected instability, such as sudden chatter or clamping loosening, without the need for online modeling or frequent calibration, while tolerating slow drift, such as normal tool wear.

[0025] Furthermore, the optimal clamping pressure target value matching the current processing state is calculated, including: Set the steady-state credit score C, with an initial value of zero; The feedback signal of the current processing status is acquired in real time, and its current normalized deviation exponent from the feedback signal benchmark is calculated, as follows: ; Where D is the current normalization deviation index, and R current R is the current processing status feedback signal. initial As a reference for the feedback signal, δ prevents small biases caused by division by zero; If D≤0, then |△| is accumulated to the steady-state credit score C; If D > 0, then the steady-state credit score C is reset to zero; Only when the accumulated value of C is greater than or equal to the preset credit threshold, will the current basic clamping pressure be automatically updated, and the steady-state credit score C be reset to zero, as shown in the following formula: ; in, For the updated base clamping pressure, P base The current basic clamping pressure, κ is the attenuation coefficient, with a value range of [0.002, 0.005], and the updated... Not less than 85% of the initial basic clamping pressure; Based on the updated base clamping pressure and the current normalized deviation index, the optimal clamping pressure target value is calculated using the following formula: ; Among them, P opt To achieve the optimal clamping pressure target value, The updated base clamping pressure is given by β, which is the maximum pressure increase coefficient, ranging from [0.2, 1], and γ is the response sensitivity parameter, ranging from [1, 5].

[0026] Specifically, in batch processing of identical workpieces, if a fixed clamping pressure is used for each workpiece, it is often necessary to set the clamping pressure according to the worst-case conditions, such as new tools or cold clamping. This leads to over-clamping under many normal operating conditions, wasting energy and accelerating fixture fatigue and workpiece surface damage. However, if the pressure is simply adjusted dynamically based on the current vibration level, it is easily affected by slow process drift such as tool wear and temperature rise, causing frequent pressure fluctuations or even misadjustments. To solve this contradiction, this invention proposes a progressive pressure optimization strategy based on steady-state credit accumulation. The core idea is that pressure adjustment is not decided based on instantaneous state, but on performance that is consistently better than the initial steady state as "credit." Only when the system is stable in the long term and performs better is a slight reduction in the base clamping force allowed. The system uses the feedback signal when the first workpiece enters steady-state cutting as the process anchor point, i.e., R. initial During subsequent processing, the system continuously evaluates the deviation of the current state from the anchor point. The key is that when the system performance is better than or equal to the initial steady state (i.e., the normalized deviation index Δ≤0), it is considered a "positive verification" of the current clamping strategy, and steady-state credit is accumulated. Once any signs of degradation appear (Δ>0), the credit is immediately reset to zero to prevent slow degradation from being misjudged as an optimizable window. Only when the accumulated credit reaches a preset threshold Cth is it determined that the current clamping force has a safety redundancy, allowing for conservative attenuation. The attenuation range is strictly limited to 0.2%–0.5%, and not lower than 85% of the initial pressure, thus achieving energy saving and consumption reduction while ensuring safety and stiffness. To address the risk of sudden instability, the system adopts a nonlinear pressure boosting response mechanism: the optimal clamping pressure target value increases exponentially with the deviation index. This design ensures: low sensitivity to minor disturbances (such as chip impact), avoiding frequent adjustments; rapid response to significant instability (such as the initial onset of chatter), but the pressure boosting range is constrained by an upper limit to prevent over-adjustment from causing system shock. To verify the effectiveness of the preset credit threshold Cth, a systematic comparative experiment was conducted under typical batch processing scenarios. The experiment covered end milling tasks of three materials: 45# steel, 6061-T6 aluminum alloy, and Ti-6Al-4V titanium alloy. For each working condition, 30 identical workpieces were continuously processed, and the initial basic clamping pressure was uniformly set to 5.0MPa. The control strategy was run for different Cth values ​​(0.1, 0.2, 0.3, 0.5, 0.8), and the following indicators were statistically analyzed: (1) number of pressure optimization triggers (i.e., workpiece number with C≥Cth); (2) average clamping pressure (during the entire task); (3) number of chatter occurrences (judged as a sudden increase of 3 times in vibration amplitude lasting >0.5s). The experimental results are shown in Table 1: Table 1 Analysis of the experimental results shows that when Cth∈[0.2, 0.5], the system achieves a 9%–12% reduction in clamping pressure under the premise of zero chatter (or extremely low risk), exhibiting optimal overall performance. The attenuation coefficient κ refers to the relative attenuation ratio of the basic clamping pressure during each optimization, transforming safety redundancy into a small, gradual pressure release, reducing energy consumption and fixture load without disturbing the system's dynamic characteristics. In this embodiment, three typical engineering materials—45# steel, 6061-T6 aluminum alloy, and Ti-6Al-4V titanium alloy—were used. The experimental platform was the same as before: a vertical machining center, a Kistler force gauge, and a PCB accelerometer. The initial basic clamping pressure was set to 5.0 MPa, and the preset credit threshold Cth=0.3. Thirty identical workpieces were continuously processed under each working condition. Five sets of parameters were tested at κ=0.001, 0.002, 0.003, 0.005, and 0.01 respectively. The evaluation indicators included (1) average clamping pressure, (2) number of chatter occurrences, and (3) number of micro-slippage events on the fixture contact surface (detected by high-frequency acoustic emission signals). The experimental results are shown in Table 2. Table 2 Analysis of the experimental results shows that when κ∈0.002~0.005, the system achieves a pressure reduction of 9%~10.2% under the premise of zero chatter and almost no micro-slippage, with κ=0.003 being the optimal value for comprehensive performance. This invention conducted clamping force-system stiffness correlation experiments and cutting stability limit experiments on three typical materials. Fixed process parameters were used: spindle speed 8000rpm, feed per tooth 0.1mm, axial depth of cut 3mm, initial clamping pressure 5.0MPa, gradually reduced to 4.0MPa (80%) and 4.25MPa (85%). Twenty identical thin-walled workpieces (3mm wall thickness) were continuously machined under each pressure, and instability events were recorded. The results are shown in Table 3. Table 3 Analysis of the experimental results showed that all materials could stably complete the machining of 20 pieces under 85% pressure (4.25 MPa). When the pressure was reduced to 80% (4.0 MPa), Ti-6Al-4V and 6061-T6, which are most sensitive to stiffness, exhibited instability, verifying the necessity of 85% as a safety lower limit. In the embodiment, the initial basic clamping pressure refers to the original pressure set at the start of this machining task, such as P0 = 5 MPa. It is the fundamental pressure that evolves dynamically during the task, used to calculate P for subsequent workpieces. opt The target clamping pressure P for each workpiece opt All based on the latest The updated baseline pressure is used as the pressure baseline for subsequent workpieces, while the optimal clamping pressure target value is superimposed on this baseline with a nonlinear compensation term, realizing a hierarchical control strategy of "slow baseline adjustment and fast response adjustment". This design reduces energy consumption and fixture load while ensuring machining safety.

[0027] Furthermore, the current basic clamping pressure includes: Current basic clamping pressure P base The initial clamping pressure is the initial basic clamping pressure calibrated after the machining task first enters the steady-state cutting stage, and is dynamically updated when self-optimization is satisfied.

[0028] Furthermore, the initial base clamping pressure and feedback signal references include: The initial basic clamping pressure and feedback signal reference are automatically calibrated when the machining task first enters the steady-state cutting stage; where, the machining task refers to the continuous machining of multiple identical workpieces using the same set of fixtures, tools and CNC programs; Before the start of this processing task, a target value for initial clamping pressure is preset. Through the pressure-command mapping relationship, the target value is converted into the corresponding initial pressure control command and output to the clamping actuator to complete the workpiece clamping. Upon first entering the steady-state cutting stage, within the stabilization time window, the current output pressure control command is acquired, and based on the pressure-command mapping relationship, it is converted into the corresponding clamping pressure value. The average value of this value within the stabilization time window is then calculated as the initial basic clamping pressure. The mean value of the feedback signal within the window is calculated synchronously to obtain the feedback signal reference, where the feedback signal is the ratio of the vibration amplitude to the cutting force within the window; The steady-state time window is the time interval from the start of the steady-state cutting stage to a duration of T0, where T0 is the preset reference acquisition duration, ranging from 0.5 to 2.0 seconds.

[0029] Specifically, before the task begins, the system loads a preset initial clamping pressure value, such as 5.0 MPa, and generates corresponding initial pressure control commands, such as proportional valve current or hydraulic servo commands. Once the first workpiece enters the steady-state cutting stage, the system opens a steady-state time window. This window starts at the beginning of steady-state cutting and lasts for a duration of T0, typically ranging from 0.5 to 2.0 seconds. Within this window: the system reads the pressure control command sequence in real time, maps the read pressure control commands back to the current clamping pressure value using a pre-calibrated pressure-command mapping relationship, such as a lookup table, and calculates its average value within the steady-state time window to obtain the actual initial clamping pressure acting on the workpiece; simultaneously, it calculates the machining state feedback signal (defined as the ratio of vibration amplitude to cutting force, i.e., chatter sensitivity index) and calculates its average value within the window, using it as the feedback signal base R.initial Once the initial clamping pressure and feedback signal reference are calibrated, they serve as the reference for all subsequent workpieces (from the 2nd to the Nth) in this task, driving the adaptive adjustment of the clamping pressure. The steady-state time window is only activated once during the machining of the first workpiece to establish a task-level process anchor point; subsequent workpieces will not have their references recalibrated to ensure sensitive detection capability against abnormal degradation of the system. The steady-state time window is used only for one-time calibration of the task-level reference. Its activation is triggered by "first entry into steady-state cutting," and it is closed after a preset duration T0, permanently locking R. initial The system uses the initial baseline clamping pressure; however, before the steady-state window closes, the credit score and pressure optimization logic are temporarily disabled to avoid invalid or erroneous comparisons when the baseline is undetermined; from the end of the steady-state window, all subsequent R... current All are related to the locked R initia Normalized comparisons are performed to drive adaptive control. The steady-state time window and the sliding window time scale used to calculate the current feedback signal are consistent, a necessary design to ensure the fairness of the normalized comparison. The initial basic clamping pressure calibration relies on the mapping relationship between the pressure control command and the pre-calibration, aiming to overcome the nonlinearity of the actuator and the uncertainty of clamping, and to obtain a high-fidelity task starting pressure reference. In the subsequent adaptive control process, the system no longer relies on pressure commands or direct pressure measurement, but instead calculates the ratio of vibration amplitude to cutting force (i.e., the feedback signal R) in real time. current To evaluate the current dynamic stability and compare it with the initial baseline R initial Comparison drives credit scores and pressure optimization. This design implements a hierarchical strategy of "emphasizing accuracy during calibration and performance during control": ensuring the reliability of the benchmark while directly linking the optimization logic to the machining quality, avoiding control failure due to actuator drift or sensor loss. The pressure-command mapping relationship is obtained in advance through offline static calibration of the hydraulic system: during the equipment commissioning phase, a standard force measuring device (such as a force ring or a calibration-grade pressure sensor) is installed at the contact position between the workpiece and the fixture. In a non-cutting state, a series of known control command values ​​u, such as 0v, 2v, 4v, ..., 10v, are applied to the clamping actuator, and the actual clamping force output by the standard force measuring device is recorded and converted into an equivalent clamping pressure P, forming a discrete mapping pair set {(u1, P1), (u2, P2), ..., (u... M P M This set constitutes a pre-stored lookup table. In the actual machining task, even if no real-time pressure sensor is installed, the system can still use this pre-stored lookup table to convert the collected average control command values ​​into the corresponding initial basic clamping pressure through linear interpolation during the calibration phase.

[0030] The instruction generation module 40 is used to apply ramp limiting processing to the calculated optimal clamping pressure target value to obtain pressure control instructions.

[0031] Specifically, although the calculated optimal clamping pressure target value can theoretically be directly used as a control command, a direct step switch will cause a sudden change in the clamp-workpiece contact stiffness, which may induce micro-vibration or even chatter. Therefore, this invention modifies P before output. opt Applying a ramp limiting process means that the control command gradually changes from the current value to the target value according to a preset slope. This design may seem to introduce a delay, but it actually avoids the paradox of optimization leading to instability by maintaining the continuity of the system's dynamic characteristics.

[0032] Furthermore, pressure control commands are obtained, including: The optimal clamping pressure target value is converted into the corresponding current pressure control command, and then compared with the previously output pressure control command. If the absolute value of the difference between the two exceeds the preset single-step change limit, the current pressure control command will be limited to the previous command value plus or minus the single-step change limit. Otherwise, output the current pressure control command directly.

[0033] Specifically, ramp limiting in the digital control system is achieved through single-step change limiting: after converting the calculated optimal clamping pressure target value into the current pressure control command, it is compared with the previous output command. If the change exceeds the preset limit value, the maximum allowable step size is output; otherwise, the command is output directly. This method is equivalent to applying ramp constraints to the pressure command, ensuring smooth actuator operation. The currently calculated optimal clamping pressure target value P... opt Based on the pre-stored pressure-command mapping relationship, it is converted into the corresponding current pressure control command u. target , will u target The pressure control command u output from the previous optimal clamping pressure target prev Compare them and calculate the absolute value of their difference |u targe -u prev |. if |u targe -u prev |>Δu max The control command output this time is limited to u. out =u prev ±Δu max , when u target >u prev Take the "+" sign when u target <u prev， Use the "-" sign. Otherwise, i.e., |u targe -u prev |≤Δu max Then directly output u out =u target . will u outThe final pressure control command for this adjustment is sent to the actuator and recorded as the value for the next adjustment. prev Δu max This is a single-step variation limit value, which can be preset based on the hydraulic system response characteristics, workpiece material, and structural stiffness. Typically, its value is chosen to ensure that the rate of change of clamping pressure does not exceed 1.0 MPa / s, thus preventing micro-slippage or stiffness jumps at the clamp-workpiece contact interface due to sudden command changes, thereby preventing abnormal phenomena such as chatter induced by the optimized action itself. This mechanism essentially achieves a ramp-limiting effect in the discrete-time domain: even if the optimal target value changes significantly, the actual control command gradually approaches the target with controlled step sizes, ensuring a smooth transition of the system's dynamic characteristics and significantly improving the robustness and safety of the adaptive control process. Single-step variation limit value Δu max The parameter is determined by a lookup table. The system pre-stores a "limiting parameter configuration table," which directly outputs the corresponding Δu based on a joint index of the workpiece material type and structural stiffness level (determined by wall thickness, support method, etc.). max The value. The amplitude limiting parameter configuration table was pre-calibrated through systematic offline stability experiments. The specific calibration process is as follows: 1. Based on the typical processing tasks of the enterprise, the workpieces are classified according to the following two dimensions: Material type: including Ti-6Al-4V titanium alloy, 6061-T6 aluminum alloy, 45# steel, etc.; Structural stiffness level: based on geometric characteristics such as wall thickness, cantilever length, and number of support points, they are divided into "high rigidity" (such as solid blocks), "medium rigidity" (such as ribbed structures), and "low rigidity" (such as thin-walled cantilever parts with a wall thickness ≤3mm). For each type of combination, standard test pieces are prepared to ensure that their clamping method and positioning reference are completely consistent with actual production. 2. For each test piece, under steady-state cutting conditions (e.g., constant spindle speed, feed, and depth of cut), perform the following operations: Starting from the current clamping pressure, apply a step-down pressure command, corresponding to a control command change of Δu; gradually increase Δu (e.g., starting from 0.001V, increasing by 0.001V each time), continuously machining 5 workpieces per test; monitor for any of the following abnormalities: a sudden increase in acoustic emission signal (>45dB, determined as micro-slip), vibration acceleration RMS value exceeding 3 times the steady-state baseline, periodic vibration marks or chatter on the surface causing machine shutdown. Record the Δu at the first occurrence of the abnormality. crit This is the critical single-step variation limit value for this operating condition. 3. Take the safety limit value as 70%~80% of the critical value, that is: Δu max =β·Δu crit β∈[0.7,0.8], let Δu max Associated with the corresponding "material + structural stiffness level" combination, this is written into the limiting parameter configuration table. For example, for a solid block of 45# steel, Δu crit =0.025V, take Δu max=0.020V. This configuration table is calibrated during the equipment commissioning or process development phase and is fixed as system parameters in the CNC controller. Before the formal machining task begins, the operator or CAM system inputs the workpiece material and structure type, and the system automatically matches and loads the corresponding Δu. max value.

[0034] Furthermore, the steady-state cutting stage includes: The steady-state cutting stage is a machining state in which the cutting process has moved away from the transient impact of the initial entry, and the cutting force and vibration remain stable without significant fluctuations.

[0035] Specifically, in lathe machining, a single tool pass includes three stages: entry, steady-state cutting, and exit. Data from the entry stage is susceptible to transient interference and cannot reflect the true system stiffness, making adaptive control unsuitable. Instability in the exit stage is not caused by insufficient clamping force but is a boundary effect, similarly unsuitable for adaptive control. The steady-state cutting stage refers to the period after the tool has completed its initial entry and entered continuous, uniform cutting. During this stage, the cutting force tends to stabilize, the vibration level is stable, and the chip shape is uniform. At this time, the dynamic stiffness of the fixture-workpiece system is no longer disturbed by the initial impact and can truly reflect the stability state under the current clamping pressure, thus making it suitable as the observation starting point for adaptive control. This invention uses a deterministic criterion based on CNC program timing to identify this stage: when the system detects that a constant feed cutting command (such as the G01 command) is currently being executed, and the duration since the command's initiation has been ≥ the preset entry buffer time T... dela If so, it is determined that the steady-state cutting stage has been entered. Buffer time T dela Used to avoid transient impact during cutting, its typical value is 100–300 milliseconds, which can be calibrated through process testing based on the workpiece material and diameter. For example, when machining a 50 mm diameter 45# steel shaft, T... dela =200ms can effectively avoid the entry impact zone. While force or vibration signals are not monitored in real time, because turning is a continuous cutting process and the entry transient duration is short (typically <200 ms), setting T appropriately can help avoid this impact. dela It can ensure with a high probability that the steady-state window falls within the true steady-state range without increasing the cost of the sensor.

[0036] The instruction output module 50 is used to output pressure control instructions to the clamping actuator to control it to output the corresponding clamping pressure, thereby realizing dynamic adaptive adjustment of the clamping pressure.

[0037] Specifically, the system outputs pressure control commands to a clamping actuator with high dynamic response (such as a proportional control hydraulic cylinder or a high-speed electro-pneumatic valve), driving it to adjust the output pressure according to the command during the steady-state cutting phase of the current workpiece, thereby achieving online adaptive adjustment of the clamping force. The actuator operates in open-loop control mode and does not rely on the clamping chamber pressure sensor; its actual output pressure is indirectly determined by the pre-calibrated pressure-command mapping relationship. To ensure adjustment safety, the system only triggers pressure adjustment when the following conditions are met: (1) the current steady-state cutting phase is reached, and the remaining cutting time is ≥ the preset safety window (e.g., 2 seconds); (2) the credit score meets the standard, and the calculated optimal clamping pressure target value is lower than the current value (only a small pressure reduction is allowed); (3) the command change has been processed by single-step limiting to ensure that the pressure transition rate does not exceed the safety threshold. After the command is issued, the actuator smoothly changes the output pressure according to its inherent dynamic characteristics. After the adjustment is completed, the system continues to collect vibration and cutting force signals and calculate the feedback index R. current This is used to determine whether the current clamping pressure can still maintain dynamic performance at or above the initial steady state. If R current ≤R initial This indicates that even with reduced clamping force, the system remains in a safe and stable state, and the current pressure is acceptable; if R curren >R initial If the clamping force is below the critical threshold, the system stiffness is insufficient, and optimization must be stopped and a higher pressure strategy should be implemented. This mechanism achieves adaptive optimization of the minimum necessary clamping pressure while ensuring machining reliability.

[0038] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A precision boring and milling machine clamping pressure adaptive adjustment system based on real-time feedback, characterized in that, The system includes: The signal acquisition module is used to synchronously and in real time acquire the first type of signal and the second type of signal during the boring and milling process. The first type of signal is processed to characterize the vibration amplitude of the dynamic stability of the workpiece-fixture system, and the second type of signal is processed to characterize the cutting force of the instantaneous cutting load. The status feedback module is used to calculate the ratio of the processed first type to the second type of signal in real time, and dynamically correct it by using the historical stability information of the processing process to generate a feedback signal that represents the current processing status. The pressure calculation module is used to normalize and compare the current machining state feedback signal with the feedback signal reference to obtain the deviation index characterizing the current machining stability. Based on the deviation index, the optimal clamping pressure target value matching the current machining state is calculated. The feedback signal reference is the moment when the current machining task first enters the steady-state cutting stage. The moment when the machining state feedback signal first enters the steady-state cutting stage is the start time of the first workpiece machining entering the steady-state cutting stage in the current machining task. The instruction generation module is used to apply ramp limiting processing to the calculated optimal clamping pressure target value to obtain pressure control instructions. The command output module is used to output pressure control commands to the clamping actuator to control its output of corresponding clamping pressure, thereby realizing dynamic adaptive adjustment of clamping pressure.

2. The precision boring and milling machine clamping pressure adaptive adjustment system based on real-time feedback as described in claim 1, characterized in that, During the boring and milling process, two types of signals are simultaneously acquired in real time. The first type of signal, after processing, is used to characterize the vibration amplitude of the workpiece-fixture system's dynamic stability. The second type of signal, after processing, is used to characterize the cutting force of the instantaneous cutting load, including: At the initial stage of boring and milling machine machining, the following operations are performed simultaneously: By deploying vibration sensing units on the workpiece-fixture system, the original vibration signals reflecting the mechanical vibration of the workpiece-fixture system are collected in real time. By integrating a force sensing unit near the cutting edge of the tool, the original cutting force signal reflecting the interaction between the tool and the workpiece is acquired in real time. The original vibration signal is used as the original input of the first type of signal. After at least zero-point drift compensation and high-frequency noise filtering preprocessing, its features are extracted to obtain the vibration amplitude that characterizes the dynamic stability of the workpiece-fixture system. The original cutting force signal is used as the original input of the second type of signal. After the same preprocessing, its features are extracted to obtain the cutting force that characterizes the instantaneous cutting load. The data acquisition process is controlled by a unified trigger signal, ensuring physical synchronization of acquisition time.

3. The precision boring and milling machine clamping pressure adaptive adjustment system based on real-time feedback as described in claim 1, characterized in that, Generate feedback signals characterizing the current processing state, including: Based on the preprocessed vibration amplitude, a short-time Fourier transform is performed to obtain the current spectrum. The centroid frequency and spectral spread of the current spectrum are then calculated. The centroid frequency represents the average position of the vibration energy on the frequency axis, and the spectral spread represents the degree of dispersion of the vibration energy around the centroid frequency. Based on the center of gravity frequency and spectral spread, the current instability trend index is calculated in real time, as shown in the following formula: ； Among them, FDI current The current instability trend index is represented by ε, which is a small bias to prevent division by zero. From the moment the workpiece enters the steady-state cutting stage, continuously record the current instability trend index and calculate its moving average μ within the sliding time window in real time. fdi and moving standard deviation σ fdi Calculate the adaptive weighting coefficient w fdi The formula is as follows: ； Where η is the sensitivity gain coefficient, and the value of η ranges from [0.2, 1]. The feedback signal for the current processing status is calculated using the following formula: ； Among them, R current This is a feedback signal for the current processing status.

4. The precision boring and milling machine clamping pressure adaptive adjustment system based on real-time feedback as described in claim 1, characterized in that, The optimal clamping pressure target value matching the current machining state is calculated, including: Set the steady-state credit score C, with an initial value of zero; The feedback signal of the current processing status is acquired in real time, and its current normalized deviation index from the feedback signal benchmark is calculated, as follows: ； Where D is the current normalization deviation index, and R current R is the current processing status feedback signal. initial As a reference for the feedback signal, δ prevents small biases caused by division by zero; If D≤0, then |D| is accumulated to the steady-state credit score C; If D > 0, then the steady-state credit score C is reset to zero; Only when the accumulated value of C is greater than or equal to the preset credit threshold, will the current basic clamping pressure be automatically updated, and the steady-state credit score C be reset to zero, as shown in the following formula: ； in, For the updated base clamping pressure, P base The current basic clamping pressure, κ is the attenuation coefficient, and its value ranges from [0.002, 0.005]. Based on the updated base clamping pressure and the current normalized deviation index, the optimal clamping pressure target value is calculated using the following formula: ； Among them, P opt To achieve the optimal clamping pressure target value, The updated base clamping pressure is given by β, which is the maximum pressure increase coefficient, ranging from [0.2, 1], and γ is the response sensitivity parameter, ranging from [1, 5].

5. The precision boring and milling machine clamping pressure adaptive adjustment system based on real-time feedback as described in claim 4, characterized in that, Current basic clamping pressure includes: Current basic clamping pressure P base The initial clamping pressure is the initial basic clamping pressure calibrated after the machining task first enters the steady-state cutting stage, and is dynamically updated when self-optimization is satisfied.

6. The precision boring and milling machine clamping pressure adaptive adjustment system based on real-time feedback as described in claim 5, characterized in that, Initial basic clamping pressure and feedback signal reference, including: The initial basic clamping pressure and feedback signal reference are automatically calibrated when the machining task first enters the steady-state cutting stage; where, the machining task refers to the continuous machining of multiple identical workpieces using the same set of fixtures, tools and CNC programs; Before the start of this processing task, a target value for initial clamping pressure is preset. Through the pressure-command mapping relationship, the target value is converted into the corresponding initial pressure control command and output to the clamping actuator to complete the workpiece clamping. Upon first entering the steady-state cutting stage, within the stabilization time window, the current output pressure control command is acquired, and based on the pressure-command mapping relationship, it is converted into the corresponding clamping pressure value. The average value of this value within the stabilization time window is then calculated as the initial basic clamping pressure. The mean value of the feedback signal within the window is calculated synchronously to obtain the feedback signal reference, where the feedback signal is the ratio of the vibration amplitude to the cutting force within the window; The steady-state time window is the time interval from the start of the steady-state cutting stage to a duration of T0, where T0 is the preset reference acquisition duration, ranging from 0.5 to 2.0 seconds.

7. The precision boring and milling machine clamping pressure adaptive adjustment system based on real-time feedback as described in claim 6, characterized in that, The steady-state cutting stage includes: The steady-state cutting stage is a machining state in which the cutting process has moved away from the transient impact of the initial entry, and the cutting force and vibration remain stable without significant fluctuations.

8. The adaptive adjustment system for clamping pressure of a precision boring and milling machine based on real-time feedback as described in claim 1, characterized in that, The pressure control commands received include: The optimal clamping pressure target value is converted into the corresponding current pressure control command, and then compared with the previously output pressure control command. If the absolute value of the difference between the two exceeds the preset single-step change limit, the current pressure control command will be limited to the previous command value plus or minus the single-step change limit. Otherwise, output the current pressure control command directly.

Images