A processing and detecting system and detecting device for hydraulic bushing

Abstract

The present application relates to the field of precision machining detection, and discloses a hydraulic shaft sleeve machining detection system and detection device, which comprises a pneumatic servo control module, a signal acquisition and processing module and a thermal inversion calculation module.The pneumatic servo control module drives the overturning cylinder to apply a composite pressure waveform comprising a static reference and a periodic jitter component to the hydraulic sleeve; the signal acquisition and processing module performs frequency domain analysis on the displacement signal detected by the displacement sensor to separate the average displacement and displacement amplitude; the thermal inversion calculation module calculates the real-time dynamic stiffness according to the displacement amplitude, inversely calculates the temperature rise value using a material stiffness thermal softening model, and separates the thermal expansion and elastic deformation error through multi-dimensional error decoupling operation to calculate the standard temperature and real height data under no external force.The present application effectively solves the oil film and thermal deformation interference by using active excitation and frequency domain analysis technology, and realizes high-precision online detection and machining compensation of the thermal workpiece.

Description

Technical Field

[0001] This invention relates to the field of precision machining and inspection, specifically to a machining and inspection system and device for hydraulic bushings. Background Technology

[0002] As a critical component in a hydraulic system, the dimensional accuracy of hydraulic bushings directly affects the smoothness of valve core movement and the sealing performance of the system. Therefore, strict tolerance control is required for key dimensions such as height and bore diameter during production. In automated mass production, online inspection is typically performed immediately after workpiece machining to ensure yield rates and timely adjustment of machine tool parameters.

[0003] In actual wet machining environments, freshly machined hydraulic bushings are in a non-standard state, posing multiple challenges to high-precision measurement. First, the frictional heat generated during cutting causes a significant increase in workpiece temperature and thermal expansion, resulting in workpiece dimensions at room temperature exceeding their actual dimensions. Using natural cooling before measurement would severely reduce production cycle time. Furthermore, simple temperature compensation methods are insufficient because the workpiece surface is covered with cutting fluid and the internal temperature field is uneven; traditional infrared or contact temperature measurement methods cannot accurately reflect the overall equivalent temperature rise of the workpiece, leading to inadequate compensation accuracy.

[0004] The workpiece surface and the measurement reference surface usually have a strong adhesive film of cutting oil or coolant remaining on them. During contact measurement, the presence of this oil film introduces thickness errors. To pierce or squeeze the oil film to contact the metal body, a large contact pressure is typically required. However, hydraulic bushings are mostly thin-walled or hollow structures, and a large measuring force can cause elastic deformation of the workpiece. Further complicating matters, the elastic modulus of metallic materials changes with increasing temperature, exhibiting a stiffness-softening characteristic. This means that under the same measuring force, the amount of elastic deformation produced by a high-temperature workpiece is not the same as that of a workpiece at room temperature.

[0005] Existing detection equipment typically treats measurement as a process of acquiring a single geometric quantity, lacking the ability to analyze the aforementioned physical field coupling phenomena.

[0006] Existing technologies cannot simultaneously eliminate oil film thickness interference, thermal expansion deformation, and elastic compression errors caused by stiffness changes in a single measurement process. As a result, the measured values ​​cannot accurately reflect the dimensions of the workpiece under standard temperature and no external force conditions, ultimately affecting the accuracy of machining compensation. Summary of the Invention

[0007] The technical problem solved by this invention is that hydraulic bushings have a temperature rise effect after machining and have a machining oil film on their surface, which makes it impossible to directly obtain the true size of the workpiece under standard temperature and no external force during online inspection. Traditional inspection methods are affected by thermal expansion and elastic deformation, making it difficult to meet the requirements of high-precision real-time machining compensation.

[0008] To address the above problems, the present invention provides the following technical solution:

[0009] The first aspect of the present invention provides a machining and inspection system for hydraulic bushings, including a pneumatic servo control module, a signal acquisition and processing module, a thermodynamic inversion calculation module, a central processing module, and a compensation feedback module.

[0010] The pneumatic servo control module connects the electro-proportional valve and the tilting cylinder. It drives the electro-proportional valve by generating a composite control signal containing a DC bias voltage and an AC sinusoidal voltage, establishing a composite pressure waveform at the output of the tilting cylinder consisting of a static reference pressure component and a periodic chattering pressure component. The pneumatic servo control module sets the amplitude and frequency of the chattering pressure component based on the dynamic friction characteristics of the tilting cylinder to eliminate static friction hysteresis during the piston's movement.

[0011] The signal acquisition and processing module is connected to the sensor to synchronously acquire the time-domain displacement signal of the hydraulic sleeve under the action of the composite pressure waveform, and uses the frequency domain analysis method to separate the signal: on the one hand, the amplitude at zero frequency is extracted as the average displacement component characterizing the macroscopic height of the workpiece, and on the other hand, the spectral peak at the excitation frequency of the chatter pressure component is extracted as the displacement amplitude component.

[0012] The thermodynamic inversion calculation module performs stiffness calculation and temperature derivation based on the above data. First, the pressure modulation amplitude of the flutter pressure component is converted into a dynamic force amplitude. The real-time dynamic stiffness of the hydraulic sleeve is obtained by calculating the ratio of the dynamic force amplitude to the displacement amplitude component. Then, a preset material stiffness thermal softening model is invoked. Using the attenuation ratio of the real-time dynamic stiffness relative to the standard reference stiffness, combined with the stiffness thermal softening coefficient, the current temperature rise of the hydraulic sleeve is calculated in reverse. The standard reference stiffness is the benchmark data obtained by statistically measuring qualified samples under standard ambient temperature.

[0013] Based on this, the thermodynamic inversion calculation module performs multi-dimensional error decoupling operations: it removes the elastic compression correction caused by external force based on the static reference pressure component and real-time dynamic stiffness; it removes the thermal expansion correction caused by temperature based on the current temperature rise value and the linear thermal expansion coefficient of the material. Finally, by superimposing the elastic compression correction on the average displacement component and removing the thermal expansion correction, the true height data of the hydraulic sleeve under standard temperature and no external force conditions is obtained.

[0014] The system is also equipped with a central processing module to coordinate the detection process. Before applying the chatter excitation, the central processing module controls the tilting cylinder to perform a constant pressure pre-compaction action and monitors the settling rate of the displacement signal in real time. Once the settling rate is lower than the preset threshold, it is determined that the oil film extrusion is complete, and then the pneumatic servo control module is triggered to superimpose the chatter pressure component to ensure the stability of the measurement reference.

[0015] The compensation feedback module calculates the machining deviation based on the actual height data and the target nominal size, and converts the deviation into tool compensation data and sends it to the front-end machine tool CNC system to realize closed-loop control of machining or generate a dimensional qualification judgment result.

[0016] A second aspect of the present invention provides a machining and inspection device for hydraulic bushings, integrating the machining and inspection system described in the first aspect. The device includes a base, an inspection plate, a tilting cylinder, and sensors.

[0017] The base has a positioning plane to support the hydraulic sleeve under test; the detection plate is hinged to one side of the base; a tilting cylinder, as the actuating component, connects to the detection plate and drives it to tilt to the working position, so that the bottom surface of the detection plate abuts against the upper surface of the hydraulic sleeve. A sensor is mounted on the detection plate to detect changes in the vertical distance between the detection plate and the base in real time. During the detection process, the tilting cylinder applies a load in response to commands from the pneumatic servo control module, and the sensor transmits the collected distance signal to the signal acquisition and processing module for further processing.

[0018] This invention provides a machining and inspection system and device for hydraulic bushings. It has the following advantages:

[0019] 1. This invention utilizes superimposed chatter pressure components as detection signals and obtains the real-time dynamic stiffness of the hydraulic bushing by analyzing the amplitude components in the displacement response. Based on the physical characteristic that the stiffness of metallic materials decreases regularly with increasing temperature, the system can directly invert the overall temperature rise value of the workpiece through stiffness changes. This avoids the problem that traditional contact or infrared temperature measurement methods are easily interfered with by residual cutting fluid and oil film covering on the workpiece surface. It achieves accurate quantification of the internal thermal state of the workpiece and provides a reliable data foundation for subsequent thermal error compensation.

[0020] 2. This invention employs multi-dimensional error decoupling logic, enabling independent separation of elastic deformation caused by detection force and thermal expansion deformation caused by processing temperature rise during the measurement process. The system combines the inverted temperature rise value with real-time dynamic stiffness to calculate and eliminate thermal expansion corrections and elastic compression corrections, thereby directly reproducing the true height data of the workpiece under standard temperature and no external force conditions in a hot state before it cools down. This shortens the natural cooling time required after workpiece processing and improves the detection efficiency of the production line.

[0021] 3. This invention effectively improves the detection stability in complex processing environments by combining frequency domain analysis technology with a pre-compaction control strategy. The signal acquisition and processing module uses spectrum analysis to separate the static displacement representing height from the dynamic amplitude representing stiffness, avoiding signal aliasing. In conjunction with the pre-compaction action controlled by the central processing module, the displacement settlement rate is used as the steady-state criterion to ensure that data acquisition is carried out only after the oil film on the workpiece surface is fully squeezed and the mechanical contact reaches a stable state, thus ensuring the repeatability and signal-to-noise ratio of the measurement results. Attached Figure Description

[0022] Figure 1 This is a perspective view of the present invention;

[0023] Figure 2 This is a block diagram of the thermodynamic inversion and multiphysics decoupling logic of the present invention;

[0024] Figure 3 This is a flowchart of the dynamic excitation sampling and stiffness feature extraction control process of the present invention;

[0025] Figure 4 This is a time-domain displacement response signal diagram under composite pressure waveform excitation according to the present invention;

[0026] Figure 5 This is a schematic diagram of the spectral characteristics of the displacement signal after frequency domain analysis according to the present invention;

[0027] Figure 6 This is a schematic diagram of the material stiffness thermal softening model and temperature rise inversion principle of the present invention;

[0028] Figure 7 This is a comparison diagram of the dimensional deviation between the thermo-coupling measurement method of the present invention and the traditional measurement method.

[0029] The components include: 1. Base; 2. Hydraulic sleeve; 3. Detection plate; 4. Displacement sensor; 5. Tilting cylinder; 6. Central processing module; 7. Pneumatic servo control module; 8. Signal acquisition and processing module; 9. Thermal inversion calculation module; and 10. Compensation feedback module. Detailed Implementation

[0030] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] Please see the appendix Figure 1This invention provides a machining and inspection system and device for hydraulic bushings. The inspection device includes a base 1, an inspection plate 3, a displacement sensor, and a tilting cylinder 5. The base 1 is constructed as a support platform with a high-rigidity surface, used to support the hydraulic bushing 2 to be inspected. The hydraulic bushing 2 is a thin-walled metal workpiece that has undergone machining processes and is placed on the positioning reference surface of the base 1 by an external robot. At the inspection station, the hydraulic bushing 2 is in an unsteady thermal environment, and its surface is coated with cutting fluid.

[0032] A tilting cylinder 5 is located on one side of the base 1, and its output end is mechanically coupled to the detection plate 3 via a linkage mechanism or direct connection. The tilting cylinder 5 is equipped with an electro-proportional pressure regulating valve or a high-frequency response pneumatic circuit assembly, configured to output adjustable pneumatic pressure. The tilting cylinder 5 drives the detection plate 3 to tilt between a non-working position and a working position. In the working position, the bottom surface of the detection plate 3 contacts the upper end face of the hydraulic sleeve 2, applying an axial load.

[0033] The displacement sensor is mounted on the detection plate 3 or on an independent bracket fixed relative to the base 1. The displacement sensor is configured as a contact displacement sensor or a high-precision non-contact ranging unit, with its probe pointing towards the displacement reference surface of the detection plate 3. The displacement sensor is used to acquire in real-time data on the change in vertical distance between the detection plate 3 and the reference surface of the base 1, which characterizes the axial height dimension of the hydraulic sleeve 2 under stress.

[0034] A machining and inspection system for hydraulic bushings is integrated into the device via electrical signal connections. It includes a central processing module 6, a pneumatic servo control module 7, a signal acquisition and processing module 8, a thermal inversion calculation module 9, and a compensation feedback module 10. The central processing module 6 communicates with each of the aforementioned functional modules via a bus, coordinating the timing actions and data flow of each module.

[0035] The pneumatic servo control module 7 is electrically connected to the pneumatic control assembly of the tilting cylinder 5. The pneumatic servo control module 7 stores a pressure modulation command sequence and is configured to send control signals to the tilting cylinder 5 to drive it to generate a composite pressure waveform during contact between the detection plate 3 and the hydraulic sleeve 2. This composite pressure waveform includes a constant static reference pressure component and a high-frequency, low-amplitude sinusoidal chatter pressure component. The pneumatic servo control module 7 eliminates static friction hysteresis during the piston movement of the tilting cylinder 5 by adjusting the chatter frequency and amplitude.

[0036] The signal acquisition and processing module 8 is electrically connected to the signal output terminal of the displacement sensor. The signal acquisition and processing module 8 is equipped with a high-frequency analog-to-digital converter and a digital filter. The signal acquisition and processing module 8 is configured to synchronously acquire the time-series displacement signal output by the displacement sensor and perform frequency domain analysis on the signal. The signal acquisition and processing module 8 uses a fast Fourier transform algorithm or a bandpass filtering algorithm to separate the displacement amplitude component corresponding to the flutter frequency output by the pneumatic servo control module 7, as well as the average displacement component representing the macroscopic height, from the original displacement signal.

[0037] The thermal inversion calculation module 9 is connected to the signal acquisition and processing module 8. The thermal inversion calculation module 9 internally stores the material physical property parameters and structural thermal softening model of the hydraulic sleeve 2. The thermal inversion calculation module 9 is configured to receive displacement amplitude components and calculate the real-time dynamic stiffness of the hydraulic sleeve 2 based on the dynamic Hooke's law. The thermal inversion calculation module 9 further utilizes the ratio of the real-time dynamic stiffness to the standard reference stiffness to deduce the current temperature rise value of the hydraulic sleeve 2.

[0038] The thermal inversion calculation module 9 is also configured to calculate the linear thermal expansion correction caused by temperature and the elastic compression correction caused by static reference pressure, based on the derived temperature rise value and average displacement component. The thermal inversion calculation module 9 performs decoupling operations and outputs the true height data of the hydraulic sleeve 2 under standard temperature and no external force conditions.

[0039] The compensation feedback module 10 is connected to the thermal inversion calculation module 9. The compensation feedback module 10 compares the calculated actual height data with a preset tolerance range. The compensation feedback module 10 is configured to generate a dimension judgment result and generate machining compensation parameters based on the deviation between the actual height data and the target dimension. These machining compensation parameters are used to feed back to the front-end machine tool control system to correct subsequent machining processes.

[0040] Please see the appendix Figure 2 The central processing module 6, serving as the system's computational core and control hub, is physically constructed using an industrial-grade programmable logic controller (PLC), industrial PC (IPC), or embedded microprocessor. The central processing module 6 establishes a bidirectional data exchange channel with the pneumatic servo control module 7, signal acquisition and processing module 8, thermal inversion calculation module 9, and compensation feedback module 10 via an industrial fieldbus or analog I / O interface. The central processing module 6 internally contains non-volatile storage media, which stores control logic programs, a standard process parameter database, and data tables containing historical stiffness-temperature mapping relationships.

[0041] The pneumatic servo control module 7 is specifically constructed as a drive circuit including a digital-to-analog converter and a PID closed-loop controller, which is electrically connected to the proportional valve in the air circuit of the tilting cylinder 5. The pneumatic servo control module 7 is configured to execute a pressure modulation strategy, i.e., to generate a composite pressure waveform on the time axis. During system operation, the pneumatic servo control module 7 establishes a static reference pressure in the air circuit according to the instructions of the central processing module 6. And on this basis, a time-varying factor is superimposed. The periodically varying high-frequency sinusoidal pressure wave component. The instantaneous pressure applied to the end face of the hydraulic sleeve. It is controlled to follow the following dynamic relationship:

[0042] ;

[0043] in, Indicates instantaneous pressure; Indicates static reference pressure; Indicates the pressure modulation amplitude; Indicates the flutter excitation frequency; Indicates time.

[0044] The signal acquisition and processing module 8 is specifically implemented as a front-end acquisition card containing a high-precision analog-to-digital converter (ADC) and a digital signal processor (DSP). The signal acquisition and processing module 8 is equipped with a hardware-triggered synchronization mechanism, configured to maintain microsecond-level synchronization with the clock signal of the pneumatic servo control module 7, thereby processing the instantaneous displacement signal output by the displacement sensor. Equal-interval sampling is performed. Given the response characteristics of the physical system, the acquired instantaneous displacement signal... It presents as a superposition of DC bias and AC response:

[0045] ;

[0046] in, Indicates the average displacement component; Indicates the displacement amplitude component; This indicates phase lag caused by material damping and mechanical clearance; Indicates the flutter excitation frequency; Indicates time.

[0047] The thermal inversion calculation module 9 is specifically implemented as the core computing unit in the central processing module 6, and it is programmed to execute the thermal-mechanical coupling decoupling algorithm. The thermal inversion calculation module 9 first executes data validity verification logic to determine the displacement amplitude components. Is it greater than the preset minimum resolution threshold? (For example, 3 times the sensor resolution). If The module triggers an alarm for abnormal contact or system freeze;

[0048] The thermal inversion calculation module 9 is configured to receive displacement amplitude components. Specifically, to eliminate the misleading effect of viscous damping caused by the cutting fluid film adhering to the workpiece surface on stiffness measurements, the displacement amplitude components used by the module... It is not the total amplitude of the original signal, but the elastic displacement component that is in phase with the excitation pressure waveform, extracted based on the principle of digital lock-in amplification (i.e., ).

[0049] If the verification passes, the real-time dynamic stiffness is calculated based on the dynamic Hooke's law under the small deformation assumption. :

[0050] ;

[0051] in, Indicates real-time dynamic stiffness; Indicates the pressure modulation amplitude; This indicates the area where the air pressure of the tilting cylinder 5 is converted into mechanical driving force. This represents the displacement amplitude component.

[0052] Subsequently, the thermal inversion calculation module 9 executes the temperature inversion logic based on the physical principle that the elastic modulus of metallic materials decreases with increasing temperature.

[0053] It should be noted that, although the axial stiffness of the workpiece Simultaneously affected by its geometric height Impact ( However, since the machining dimensional tolerances of the hydraulic sleeve 2 (typically on the micrometer scale) are relatively small compared to the change in elastic modulus caused by a temperature rise of tens of degrees Celsius, this embodiment ignores the second-order perturbation of stiffness caused by geometric dimensional deviations in the inversion model, and approximates that the stiffness change is only caused by the material's elastic modulus due to temperature. Decay is the dominant factor.

[0054] This logic calls a pre-stored material thermal softening model and derives the temperature rise value based on the following formula. :

[0055] ;

[0056] in, The thermal softening coefficient is the structural stiffness. To calibrate the reference stiffness; For real-time dynamic stiffness; This represents the temperature rise.

[0057] After acquiring the temperature rise data, the thermodynamic inversion calculation module 9 further performs multidimensional error compensation calculations. To eliminate the interference of real-time measurement noise on the elastic recovery calculation, the module prioritizes using calculations based on the inversion temperature. The derived theoretical stiffness is used for springback compensation. The actual height of hydraulic sleeve 2 under standard temperature and no external force conditions. The following compensation model is used for calculation:

[0058] ;

[0059] in, is the linear thermal expansion coefficient of the material of hydraulic sleeve 2; This is the actual height data; The average displacement component; This is the static reference pressure; This indicates the effective pressure-bearing area of ​​piston 5 in the tilting cylinder; For real-time dynamic rigidity; It is the linear thermal expansion coefficient; Temperature rise value.

[0060] To prevent computational overflow due to denominator singularity, the algorithm incorporates temperature boundary protection to ensure... It is always greater than 1.

[0061] The compensation feedback module 10 is configured to receive the actual height. and compare it with the preset dimensional tolerance threshold range. A comparison is performed. Based on the comparison results, the compensation feedback module 10 generates a control command containing the pass / fail determination or rework amount, and transmits the command to the CNC system of the front-end machine tool through the industrial Ethernet interface to realize closed-loop correction of the machining process.

[0062] In this embodiment, the pneumatic servo control module 7 is constructed at the hardware level as an embedded control board integrating a high-resolution digital-to-analog converter (DAC), a power amplifier drive circuit, and a signal synthesis logic unit. This module establishes a physical connection with the electro-proportional valve in the pneumatic circuit system of the tilting cylinder 5 via shielded analog signal lines and is configured to execute a micro-frequency flutter control strategy.

[0063] Given the nonlinear hysteresis (i.e., dead zone phenomenon) commonly found in industrial pneumatic components caused by the static friction of sealing rings, the signal synthesis logic unit within the pneumatic servo control module 7 is programmed to execute a waveform generation algorithm based on the superposition principle. This algorithm does not simply output a constant voltage signal, but rather synthesizes in real-time a composite control command containing both DC bias and AC excitation components. This aims to drive the electro-proportional valve to dynamically adjust the intake opening, thereby establishing a controllable dynamic pressure field within the piston chamber of the tilting cylinder 5.

[0064] Specifically, the pneumatic servo control module 7 generates time-varying parameters based on the process parameters issued by the central processing module 6. Continuously changing control voltage signal To ensure the linearity of the pneumatic output, the signal generation follows the linear modulation equation:

[0065] ;

[0066] in, This indicates the set static reference pressure value; Indicates the pressure modulation amplitude; Indicates the flutter excitation frequency; Indicates the pressure and voltage gain coefficients of an electro-proportional valve; This represents the zero-point bias voltage corresponding to zero-pressure output.

[0067] To ensure control accuracy, the above and Instead of fixed empirical values, these are calibration parameters obtained by the pneumatic servo control module 7 by calling a pre-stored valve body characteristic calibration table. This calibration table is generated based on multi-point linear fitting tests performed on the electro-proportional valve and aims to eliminate the influence of valve body manufacturing tolerances on pressure control accuracy.

[0068] Regarding excitation frequency The pneumatic servo control module 7 is equipped with automatic frequency avoidance logic. Considering the inherent resonant modes of the mechanical structure, the module internally stores amplitude-frequency characteristic data of the base 1 and the detection board 3, pre-determined through frequency sweep testing. The signal synthesis logic unit is configured to automatically adjust the frequency response based on this data. The response is set in a flat region outside the system's first natural frequency (for example, typically between 20Hz and 50Hz, avoiding the resonance peak ±5Hz). The purpose of this technique is to prevent mechanical resonance from amplifying vibration noise and to ensure that the vibration signal collected by the subsequent displacement sensor is a forced vibration directly driven by pressure fluctuations, rather than a self-excited oscillation of the structure, thereby ensuring the physical uniformity of the dynamic stiffness calculation.

[0069] At the physical level, the control signal containing a high-frequency AC component output by the pneumatic servo control module 7 causes the piston of the tilting cylinder 5 to be in a state of slight high-frequency vibration. Based on the principle of tribology, this microscopic relative motion transforms the contact friction mode between the piston seal and the cylinder wall from high-damped static friction to low-damped and relatively constant dynamic friction. This control strategy effectively linearizes the output force characteristics of the cylinder, enabling minute pressure changes to be converted into minute displacements of the detection plate 3 without hysteresis, thus providing a high signal-to-noise ratio physical basis for subsequent dynamic stiffness measurement.

[0070] Furthermore, to prevent equipment damage due to incorrect parameters, the pneumatic servo control module 7 also integrates output limiting protection logic. This logic monitors the synthesized signal in real time. When the calculated value exceeds the upper or lower limit of the rated input range of the electric proportional valve (e.g., 0-10V or 4-20mA), the module is configured to automatically clamp and clip the output signal and simultaneously report an overflow warning to the central processing module 6.

[0071] In this embodiment, the signal acquisition and processing module 8 is configured as a high-speed data acquisition and analysis unit based on a digital signal processor (DSP) or a field-programmable gate array (FPGA). At the physical level, this module includes an analog front-end conditioning circuit (AFE), a high-precision analog-to-digital converter (ADC), and a digital computing core. The module's signal input is connected to the displacement sensor via shielded twisted-pair cable, and its data output establishes a communication link with the central processing module 6 and the thermal inversion calculation module 9 via a high-speed parallel bus.

[0072] To ensure accurate extraction of micron-level displacement response from complex factory environments containing machine tool cutting vibrations and coolant flow noise, signal acquisition and processing module 8 is configured to execute a complete signal chain processing logic including analog preprocessing, synchronous sampling, and frequency domain analysis. The module integrates a cutoff frequency in the analog front-end conditioning circuit. Set as excitation frequency A fourth-order Butterworth low-pass anti-aliasing filter, 5 to 10 times (e.g., 200Hz), is used. The technical configuration of this hardware filter aims to filter out high-frequency interference signals above the Nyquist frequency, prevent spectral aliasing during sampling, and thus ensure the authenticity of the subsequent digital signal.

[0073] The analog-to-digital converter (ADC) is equipped with an external hardware trigger pin and is configured to operate under the triggering of a synchronous clock signal issued by the pneumatic servo control module 7. To satisfy the Nyquist sampling theorem and ensure the accuracy of sine wave phase reconstruction, the ADC's sampling frequency... Set as excitation frequency At least 20 times (i.e.) The signal acquisition and processing module 8 discretizes the continuous analog voltage signal into a digital time series. The length of the sequence From sampling window time With sampling frequency Joint decision ( As a preferred method, the sampling window time... It is set to an integer multiple of the excitation signal period, which is physically designed to ensure the periodic integrity of the sampled data, thereby minimizing spectral leakage effects in frequency domain analysis.

[0074] In obtaining the original digital sequence Subsequently, the digital computation core was configured to perform time-domain windowing processing. Given that the truncation effect caused by finite-length sampling can easily lead to spectral energy diffusion (i.e., sidelobe leakage), the module was configured to process the sequence... Apply Hanning window weighting. The weighted sequence. Follow the following operational logic:

[0075] ;

[0076] in, For discrete sampling point index, Total number of sampling points ( The physical significance of this windowing operation lies in smoothing the discontinuities at both ends of the signal, thereby concentrating the spectral energy in the main lobe and improving the frequency resolution of weak flutter signals.

[0077] Based on weighted sequences The signal acquisition and processing module 8 then performs a Fast Fourier Transform (FFT). This transform maps the displacement vibration signal in the time domain to the frequency domain, generating a complex spectrum sequence. :

[0078] ;

[0079] in, Indicates the first Complex spectral values ​​at each frequency point The imaginary unit; based on the spectral sequence The module is configured to execute dual-channel feature separation logic to extract the average displacement components that characterize the macroscopic dimensions. and displacement amplitude components characterizing dynamic properties .

[0080] For the average displacement component The module extracts the DC component from the spectrum (i.e., The modulus at the point is normalized and corrected based on the DC coherent gain of the window function.

[0081] For the key displacement amplitude components The module uses an index search algorithm to locate the excitation frequency in the spectrum. Corresponding specific spectral line index This index uses a relational expression. OK. The module extracts the spectral amplitude at this frequency point and calculates the physical displacement amplitude according to the following amplitude recovery formula:

[0082] ;

[0083] in, and These represent the real and imaginary parts of a complex number, respectively. This represents the window function magnitude correction factor; This represents the number of sampling points.

[0084] The frequency domain extraction method functions in the system as equivalent to constructing a narrow-bandwidth digital bandpass filter with an adjustable center frequency. This is achieved by locking only the excitation frequency. With the energy available, the system can completely eliminate static friction interference that does not change with frequency, as well as environmental noise with random frequencies (such as broadband cutting noise).

[0085] Furthermore, to prevent incorrect stiffness values ​​from being output due to sensor malfunction or poor contact, the signal acquisition and processing module 8 integrates data integrity verification logic. This logic will calculate the stiffness value... Compared with the preset background noise threshold (This threshold is set based on the sensor's noise statistics under static no-load conditions, for example, 3) The noise level is compared. If the module determines that the current signal-to-noise ratio is too low, it will automatically output an invalid flag to block the subsequent thermal inversion calculation module 9 from processing and trigger a system self-test alarm.

[0086] In this embodiment, the thermal inversion calculation module 9 is configured as a core algorithm unit embedded in the central processing module 6, or as a separate high-performance floating-point coprocessor. The data input terminal of this module receives displacement characteristic data (i.e., average displacement components) from the signal acquisition and processing module 8 via an internal high-speed bus. With displacement amplitude components Its data output is connected to the compensation feedback module 10. The thermodynamic inversion calculation module 9 integrates a non-volatile parameter data database, which contains a material physical property table for a specific batch of hydraulic sleeves 2, including the linear thermal expansion coefficient. Structural stiffness and thermal softening coefficient and standard reference stiffness .

[0087] The thermal inversion calculation module 9 is configured to execute a decoupling algorithm based on a strong coupling physical mechanism between stiffness and temperature-size. To ensure the numerical stability of the calculation results and prevent program crashes, the module has a strict pre-checking logic for input data. The module will receive displacement amplitude components... Compared with the minimum resolution threshold set by the system Perform a comparison. Minimum resolution threshold. The setting is determined based on the physical resolution limit of the displacement sensor and the statistical value of the background noise of environmental micro-vibrations (e.g., set to 0.5 micrometers). If detected... The module determines that no effective dynamic deformation has occurred (this state usually corresponds to mechanical jamming or workpiece stiffness exceeding the range), and then outputs an abnormal status code and terminates subsequent calculations. The technical purpose of this logic is to prevent floating-point division by zero overflow errors in subsequent stiffness calculation formulas from the source.

[0088] After passing the data pre-check, the thermodynamic inversion calculation module 9 calculates the real-time dynamic stiffness of the hydraulic sleeve 2 based on the dynamic Hooke's law under small deformation conditions. The calculation process follows the following dynamic equations:

[0089] ;

[0090] in, Indicates real-time dynamic stiffness; Indicates the pressure modulation amplitude; This indicates the area where the air pressure of the tilting cylinder 5 is converted into mechanical driving force. This represents the displacement amplitude component.

[0091] Based on the physical property that the elastic modulus of metallic materials monotonically decreases with increasing temperature (i.e., thermal softening effect), the thermodynamic inversion calculation module 9 is configured to perform temperature field inversion calculations. The module calls a pre-stored linear thermal softening model and utilizes real-time dynamic stiffness... relative to standard reference stiffness The attenuation ratio is used to deduce the current overall average temperature rise of hydraulic sleeve 2. The inversion calculation follows the following state equations:

[0092] ;

[0093] in, The thermal softening coefficient is the structural stiffness. To be at standard reference temperature and standard static reference pressure Calibrate reference stiffness; For real-time dynamic stiffness; This represents the temperature rise.

[0094] Considering that noise during industrial field measurements can easily affect stiffness observations Occasionally greater than (That is, calculating the fluctuation of negative temperature rise that violates the laws of physics) The module is equipped with numerical clamping logic: if the calculation result Module forced correction And record the event as a low-confidence measurement.

[0095] In acquiring temperature rise data Subsequently, the thermal inversion calculation module 9 performs the final multiphysics error decoupling operation. This operation aims to decouple macroscopic measurements containing thermal and mechanical deformations. Restored to the true geometric dimensions under standard conditions The module comprehensively considers the linear thermal expansion effect (positive error) caused by temperature rise and the effect of static reference pressure. The resulting elastic compression effect (negative error). Based on the superposition principle of mechanics of materials, the thermodynamic inversion calculation module 9 calculates the true height using the following constitutive relation model. :

[0096] ;

[0097] in, is the linear thermal expansion coefficient of the material of hydraulic sleeve 2; This is the actual height data; The average displacement component; This is the static reference pressure; This indicates the effective pressure-bearing area of ​​piston 5 in the tilting cylinder; For real-time dynamic rigidity; It is the linear thermal expansion coefficient; Temperature rise value.

[0098] To ensure the robustness of the system, the module modifies the denominator term. Perform a singularity protection check. Despite physical... Positive value and After processing, the value becomes non-negative, ensuring the denominator is always greater than 1. However, logically, the module still monitors whether the value falls within the reasonable range of [1.0, 1.05] to ensure absolute safety of the calculation. The final calculated value is... The data is encapsulated into a digital message with a timestamp and a quality flag, and transmitted to the compensation feedback module 10 for subsequent tolerance determination and tool compensation generation.

[0099] To ensure the physical authenticity of the measurement data and prevent false detections caused by contact surface separation, chip inclusion, or mechanical impact, the central processing module 6 in this embodiment is equipped with a hierarchical contact state management logic unit. This unit serves as a low-level real-time daemon process and runs between the data streams of the pneumatic servo control module 7 and the signal acquisition and processing module 8. It is configured to monitor and regulate the mechanical coupling quality between the detection plate 3 and the hydraulic sleeve 2 in real time.

[0100] In the initial stage of establishing physical contact through pneumatic loading, the central processing module 6 is configured to execute soft landing trajectory planning logic. To avoid micro-indentations on the workpiece surface or sensor zero-point drift caused by instantaneous impact between rigid metal contact surfaces, this logic unit does not directly send step pressure commands, but instead generates a pressure-time ramp curve with an adjustable slope based on a preset damping model. The pneumatic servo control module 7 drives the electro-proportional valve according to this curve, controlling the tilting cylinder 5 to drive the detection plate 3 downwards with controlled acceleration. When the system detects that the rate of change of the displacement sensor output value approaches zero, and the pressure sensor feedback value in the tilting cylinder 5 system reaches the preset contact sensing threshold... (As a preferred approach, the threshold is set to 0.05 MPa to 0.1 MPa.) When this threshold is reached, the logic unit determines that physical contact has been established, and then the instruction system smoothly transitions to the pressure holding mode until the pressure stabilizes and transitions to the static reference pressure. .

[0101] After establishing static contact, the central processing module 6 automatically invokes the contact rigidity verification logic. Given that metal shavings or oil residue are common in industrial environments, these foreign objects, if embedded between the contact surfaces, can create a false soft contact state, leading to subsequent height measurements being significantly overestimated and height measurements being severely underestimated. Therefore, this logic unit is configured to identify foreign objects by monitoring the stability of the static position. Specifically, the system... The collection period is maintained for a certain duration. A displacement data sequence (e.g., 200 ms) is generated, and its standard deviation is calculated. .like Exceeding the preset micro-creep threshold (set based on the statistical characteristics of surface roughness under clean conditions) indicates inelastic plastic settling at the contact interface (i.e., crushing of the shear layer or oil film). In this situation, the system is configured to trigger a foreign object cleaning sequence, controlling the cylinder to perform several high-frequency, high-amplitude pulse impacts to attempt to break up or expel the inclusions using impact force; if the retry fails... If the test still exceeds the limit, report a contact failure and suspend testing.

[0102] Upon entering the dynamic measurement phase, the contact state management logic unit switches to executing the anti-separation monitoring algorithm. This is because the chatter excitation occurs under static pressure. An alternating pressure wave is superimposed on top. If the modulation amplitude Excessive or external vibrations can cause instantaneous total pressure Upon falling below zero, the detection plate 3 will undergo microscopic separation from the hydraulic sleeve 2. This separation not only interrupts the force transmission path but also introduces strong nonlinear collision noise, completely undermining the basis for dynamic stiffness calculation. To prevent this, the central processing module 6 calculates the dynamic contact safety factor in real time based on the following inequality constraint model.

[0103] ;

[0104] in, This is the static reference pressure; This is the pressure modulation amplitude; This is an estimate of the instantaneous random pressure disturbance; Minimum contact pressure margin is maintained.

[0105] The contact status management logic unit is configured for real-time comparison. Relationship with the critical value of 1.0. If detected during the measurement process... The system detected a risk of contact separation and immediately sent an adjustment command to the pneumatic servo control module 7 to dynamically increase the static reference pressure. Or reduce the flutter amplitude value Force the system state to return to the safe contact zone ( .

[0106] In addition, this module also integrates overload protection logic. This is to prevent static reference pressure... Setting the displacement too high caused irreversible plastic deformation of the hydraulic sleeve 2. The central processing module 6 set a maximum allowable displacement limit based on Hooke's Law. The system pre-stores the yield strength limit of the hydraulic sleeve 2 material. And based on the effective contact area Calculate the maximum allowable pressure At any given moment, if the pressure sensor feedback value approaches... At 90% capacity, the system is configured to forcibly cut off the air supply and open the exhaust valve at the hardware level. This logic is based on the rapid unloading characteristics of the pneumatic circuit, aiming to physically protect the workpiece and detection device from overload damage.

[0107] To eliminate the nonlinear friction dead zone between the piston seal ring and the cylinder wall inside the tilting cylinder 5, the system in this embodiment is equipped with a dead zone elimination algorithm based on tribological principles at the control logic level. This algorithm is not a separate physical hardware, but is embedded in the collaborative operation mechanism of the pneumatic servo control module 7 and the central processing module 6, aiming to transform the friction state of the contact interface from high-damped, uncertain static friction to low-damped, relatively constant dynamic friction.

[0108] In traditional static or quasi-static pneumatic loading, due to the viscoelasticity and interference fit characteristics of the rubber seals, the cylinder piston must overcome the maximum static friction force at the moment of startup. Before this force can be overcome, changes in input pressure cannot be translated into macroscopic piston displacement, resulting in an unresponsive dead zone on the force-displacement curve. The existence of this dead zone causes minute stiffness changes to be masked by frictional noise, severely affecting measurement accuracy.

[0109] To overcome this physical challenge, the pneumatic servo control module 7 is configured to superimpose a high-frequency AC chatter signal onto a DC bias pressure. This chatter signal not only serves as an excitation source for detecting stiffness but also acts as a friction linearization carrier in terms of physical mechanism. Specifically, the system sets the excitation parameters based on the Stribeck friction model, ensuring that the instantaneous velocity of the piston at any given moment is... The values ​​are not zero, or the dwell time at the zero point is short, thus preventing the friction pair from re-entering the adhesive state.

[0110] To ensure that the flutter strategy can fully cover the friction dead zone, the central processing module 6 has a preset minimum drive amplitude verification logic. This logic defines the pressure modulation amplitude based on the following dynamic constraint inequality. The lower limit:

[0111] ;

[0112] in, This is the pressure modulation amplitude; This is the maximum static friction force; It is kinetic friction; This indicates the effective pressure-bearing area of ​​piston 5 in the tilting cylinder; To overcome the safety factor in the friction dead zone; This is a compensation item for pressure loss along the pressure path.

[0113] Based on the above principle, when the dynamic pressure fluctuations applied by the system meet the above conditions, the piston seal ring will always be in a sliding state at the microscale.

[0114] Furthermore, to address the drift in frictional characteristics caused by seal wear, the central processing module 6 is equipped with adaptive dead-zone compensation logic. This logic monitors the spectral purity of the displacement response signal in real time. Based on signal and system theory, when the input is a pure sine wave and the system exhibits symmetrical nonlinearity (such as a Coulomb friction dead zone), the output signal will produce odd-order harmonic distortion.

[0115] Therefore, if a significant increase in the amplitude of odd harmonic components (such as the 3rd and 5th harmonics) is detected in the displacement spectrum, the system determines that this is likely to cause waveform clipping at the top or zero-crossing hysteresis due to increased friction. To address this anomaly, the logic unit is configured to automatically increase the pressure in 0.01 MPa increments. The odd harmonic distortion rate will continue to fall below a preset linearity threshold (e.g., 1%) until the odd harmonic distortion rate falls below that threshold.

[0116] Through this active friction dead zone elimination mechanism, this embodiment effectively transforms the highly nonlinear pneumatic actuator into a quasi-linear mechanical exciter, enabling the detection device to achieve micron-level dynamic stiffness precision measurement using low-cost industrial cylinders without introducing expensive hydrostatic bearings or linear motors. The fundamental theory of the Stribeck curve and the tribological properties of rubber seals are well-known to those skilled in the art, and will not be elaborated upon here.

[0117] To accurately quantify the mechanical impedance characteristics of the hydraulic sleeve 2 in complex environments including cutting fluid flow noise and machine tool background vibration, the central processing module 6 is equipped with an advanced spectrum analysis unit. Unlike simple time-domain peak extraction, this unit is configured to construct the system's transfer function using a cross-power spectral density algorithm based on statistical signal processing theory, thereby achieving statistical separation of signal and noise in the frequency domain.

[0118] The signal acquisition and processing module 8 acquires a time-stricken aligned pressure excitation signal via a hardware synchronous clock. With displacement response signal Subsequently, the spectrum analysis unit performs a Fast Fourier Transform (FFT) to convert it to the frequency domain. Considering the physical fact that the displacement sensor in this embodiment is susceptible to the influence of oil film and micro-roughness on the workpiece surface, resulting in a slightly lower signal-to-noise ratio compared to the pressure sensor, the system employs... The frequency response function is calculated using an estimation method. The mathematical principle of this method assumes that noise is mainly superimposed on the system output (i.e., displacement data), and eliminates uncorrelated random noise at the output through cross-spectral calculation.

[0119] The system calculates discrete frequency points based on the following formula. Complex frequency response function at point

[0120] ;

[0121] in, It is the complex frequency response function; This represents the one-sided cross-power spectral density of the pressure signal and the displacement signal. This represents the one-sided self-power spectral density of the pressure signal; These are discrete frequency points.

[0122] Based on the calculated complex frequency response function The system extracts complex dynamic stiffness based on the principle of mechanical impedance. Since the frequency response function physically represents dynamic flexibility (i.e., the displacement response produced by a unit dynamic force), the complex dynamic stiffness... Defined as the reciprocal of the frequency response function and the effective pressure area The product of:

[0123] ;

[0124] in, For complex dynamic stiffness; The cross-sectional area is the annular area under pressure. It is a complex frequency response function; Energy storage modulus; The imaginary unit; This is the loss modulus. The system extracts complex numbers... The real part is used to determine the final dynamic stiffness value:

[0125] ;

[0126] in, For real-time dynamic stiffness; The cross-sectional area is the annular area under pressure. This represents the one-sided self-power spectral density of the pressure signal; This represents the one-sided cross-power spectral density of the pressure and displacement signals.

[0127] like The small magnitude (approaching 0) means that the system exhibits non-physical characteristics of infinite stiffness (usually due to the sensor not being connected or the signal being shielded). The system is also equipped with numerical clamping logic to prevent division by zero errors.

[0128] To ensure the physical reliability of the above calculation results, the central processing module 6 executes coherence verification logic in parallel. This logic utilizes a coherence function. To evaluate the strength of the linear causal relationship between input pressure and output displacement, the calculation model of the coherence function is as follows:

[0129] ;

[0130] in, It is a coherence function; This represents the one-sided cross-power spectral density of the pressure signal and the displacement signal. This represents the one-sided self-power spectral density of the pressure signal; Let be the self-power spectral density of the displacement signal, and The value range is [0, 1].

[0131] The verification logic has a strict confidence threshold. If the calculated result This means that the current displacement response contains a large number of nonlinear disturbances (such as loose mechanical parts impacting the system, strong external vibrations, or nonlinear friction). In this case, the calculated stiffness value does not possess the physical meaning of a linear system. To address this, the system is configured to automatically discard the current measurement and trigger a remeasurement mechanism.

[0132] In addition, to further distinguish between the elastic deformation of metallic materials and the viscous deformation of oil films, the module is also equipped with a phase hysteresis angle. The monitoring logic. Phase angle is achieved through... Calculation of the ratio of the imaginary part to the real part: Based on the principles of viscoelastic mechanics, for purely elastic steel, stress and strain are essentially in phase. It should approach 0 degrees; however, for operating conditions with thick oil films or soft connections, viscous damping will cause displacement to lag significantly behind force, making... Increase. Therefore, the system has a phase angle threshold. (e.g., 5°). If detected The system will determine that there is a serious soft contact phenomenon at the contact interface and issue a maintenance prompt to the user to clean the contact surface. This detection strategy based on complex domain feature separation ensures that the obtained stiffness index can purely reflect the material properties of the hydraulic sleeve 2 body, rather than the illusion of the interface coupling state.

[0133] To provide a definite theoretical constraint for the system to perform thermodynamic inversion calculations, this embodiment constructs a multiphysics coupling model based on the small deformation assumption during the initialization phase. This model is configured to describe the macroscopic deformation behavior of the hydraulic sleeve 2 under the combined action of aerodynamic load and temperature field. Its core purpose is to establish a quantitative functional relationship between the material's elastic modulus decay and microscopic thermal expansion, thereby mathematically decoupling the single-dimensional observation of displacement into two independent physical quantities: geometric dimensions and thermal state.

[0134] In the parameter definition system of this embodiment, the hydraulic sleeve 2 is abstracted as an equivalent Kelvin-Voyt model element with temperature-dependent stiffness characteristics (simplified to a spring after ignoring the viscous term). A standard reference stiffness is pre-stored within the system. This parameter characterizes the workpiece under standard thermodynamic conditions (defined as temperature). And without preload, it has the ability to resist axial elastic deformation. The value does not depend on a single random measurement, but is based on the principles of statistical mechanics, obtained by performing multiple dynamic loading tests on the same batch of cold-state standard samples calibrated by a coordinate measuring machine (CMM). The physical meaning of this definition is to establish a zero-point reference that does not include thermal errors, so that all subsequent stiffness observation deviations can be uniquely attributed to temperature effects or material defects.

[0135] Based on the lattice vibration theory in metal physics, Young's modulus of metallic materials The stiffness exhibits an approximately linear decreasing trend with increasing temperature. It should be noted that the linear model described in this section is based on the fundamental physical principle of thermodynamic inversion. In preferred embodiments where higher accuracy is required, the system will employ a nonlinear higher-order mapping model for specific calculations. This equation describes the real-time dynamic stiffness. Attenuation law relative to reference stiffness:

[0136] ;

[0137] in, The real-time average temperature of the workpiece; Standard reference temperature; The thermal softening coefficient is the structural stiffness. To calibrate the reference stiffness.

[0138] As a preferred method, The slope normalization value was determined by performing a gradient temperature increase loading experiment on a standard workpiece in a constant temperature chamber from 20℃ to 60℃, and then performing a least-squares linear fitting on the resulting stiffness-temperature curve. The typical range of this coefficient is 2.5 × 10⁻⁶. -4 Up to 4.0×10 -4 .

[0139] Furthermore, to establish a mapping from macroscopically measured height to actual geometric dimensions, a thermo-mechanical superposition constitutive model was constructed. This model, based on the principle of linear superposition, assumes that the total axial deformation vector of the workpiece is equal to the algebraic sum of the positive elongation vector of thermal expansion and the negative vector of force-induced elastic compression. Its forward evolution formula is as follows:

[0140] ;

[0141] in, This is the absolute average height reading; This is the actual height data; It is the linear thermal expansion coefficient; The real-time average temperature of the workpiece; Standard reference temperature; This is the static reference pressure; This indicates the effective pressure-bearing area of ​​piston 5 in the tilting cylinder; For real-time dynamic stiffness.

[0142] Considering the division operation in the formula, the system is configured with denominator protection logic to prevent calculation overflow. The system verifies the denominator value before performing the calculation. Is it greater than the minimum physical stiffness threshold? .like This usually means that the workpiece has suffered severe structural yielding or fracture, and the system will immediately terminate the calculation and output a structural failure alarm.

[0143] The above constitutive model constitutes a model with two unknowns ( and The equations of the system. Since the dynamic stiffness measurement mechanism in this embodiment has independently solved the equations. This allows for logical sequential decoupling of the aforementioned equations. The system is configured to first solve for the temperature state using the inverse function of the stiffness and temperature state equations. Then Substitute into the thermo-mechanical superposition constitutive model to solve. To ensure boundary safety of the model computation, this embodiment also defines the validity domain of the parameters. Since the linearized model only has high confidence within the temperature range without phase transition, the system presets the effective temperature range of the model. (e.g., 10°) up to 80° If the temperature calculated by inversion If the material exceeds this range, it means that it is prone to entering the softening region where nonlinearity increases sharply. At this point, the model will output a confidence overflow alarm to prevent incorrect compensation caused by the failure of the physical model.

[0144] Although the aforementioned physical model provides a linearized state equation based on lattice vibration theory, in actual industrial measurement scenarios, the viscosity of the sealing grease inside the hydraulic sleeve 2 decreases exponentially with temperature. Combined with the nonlinear thermal softening effect of the metal material at the microscopic level, a single linear coefficient often cannot meet the micrometer-level inversion accuracy requirements across the entire temperature range (e.g., 10℃ to 80℃). Therefore, the central processing module 6 in this embodiment is equipped with a high-precision stiffness-temperature mapping engine. This engine is configured to perform an accurate inverse solution from dynamic stiffness to medium temperature based on a pre-calibrated polynomial regression model.

[0145] To establish an accurate physical reference, a basic mapping database needs to be built before the system is put into operation. Under calibration configuration, the measuring device equipped with a standard hydraulic sleeve 2 is placed in a precision constant-temperature environment chamber, and multi-point contact high-precision thermocouples are arranged on the sleeve surface as the true temperature reference. The system controls the environment chamber to perform quasi-static heating at a relatively slow heating rate to ensure that there is no obvious thermal gradient artifact inside the workpiece. During this process, whenever a temperature change reaches a preset step size, the system automatically triggers a dynamic stiffness measurement sequence and records the current true temperature value. The calculated stiffness value Thus constructing a Discrete sample set consisting of data points .

[0146] Based on the acquired discrete sample set, the mapping engine is configured to perform model parameter identification. To balance computational efficiency with goodness of fit to nonlinear physical characteristics, this embodiment preferably employs a second-order polynomial model to describe the evolution of stiffness with temperature. Compared to linear models, second-order models can more accurately capture the accelerated softening characteristics of materials at high temperatures caused by intensified grain boundary slip. The system constructs the following least-squares optimization objective function to minimize the mean square error:

[0147] ;

[0148] in, This is the stiffness value; These are the second-order curvature coefficients; This is the true value of the temperature. It is a first-order softening system; This is the nominal reference stiffness.

[0149] The system uses matrix-based least squares to solve the above optimization problem. During the calculation process, to prevent the Vandermonde matrix from exhibiting singular or ill-conditioned characteristics due to excessively concentrated temperature distribution at sampling points (e.g., a malfunction in the incubator causing no temperature change), the system incorporates condition number verification logic. If the calculated matrix condition number exceeds a preset threshold, the system determines the calibration data is invalid and automatically prompts the operator to check the environmental chamber status and resample. The calculated coefficient vector... It will be permanently stored in non-volatile memory.

[0150] During the real-time measurement phase, the mapping engine executes temperature inversion logic based on the fixed parameters. The signal processing unit then outputs the current real-time dynamic stiffness. Subsequently, the system no longer relies on simple proportional conversions, but instead solves for temperature... Using a quadratic equation to obtain the equivalent temperature estimate The inversion formula is as follows:

[0151] ;

[0152] in, These are the second-order curvature coefficients; This is the true value of the temperature. It is a first-order softening system; The nominal reference stiffness; For real-time dynamic stiffness.

[0153] To ensure the computational security and physical uniqueness of this inversion formula, the system has implemented the following multiple constraints and protection logic:

[0154] Discriminant nonnegativity constraint: To prevent transient impulse noise from causing... The abnormal increase makes the discriminant under the square root... If the value is less than zero (i.e., a mathematical imaginary root appears), the system checks before taking the square root. .like Physically, this means that the measured stiffness exceeds the maximum physical boundary of the theoretical model. At this point, the system is configured with forced values. The results are clamped to the physical limit boundary, and the current data quality is marked as low confidence.

[0155] Selection of physical roots: A quadratic equation in mathematics has two roots. Since the physical stiffness monotonically decreases with temperature within the effective temperature range, the system uses logical judgment to select roots falling within the effective temperature range. The root within is taken as the valid solution, thus eliminating spurious solutions that have no physical meaning.

[0156] Division protection and model degradation: Considering the high linearity of some materials, the fitted data may exhibit certain degradation. When the coefficient approaches 0, performing division directly at this point will cause overflow. The system has a threshold value set for the absolute value of the coefficient. .like The system automatically switches to the linear inversion model. This ensures the robustness of the algorithm.

[0157] Furthermore, to further eliminate measurement random errors, the mapping engine also integrates a sliding time window filtering algorithm. The system does not directly output the instantaneous inversion temperature, but instead maintains a time window of length... (For example A first-in, first-out (FIFO) queue. The final output temperature. For queue The weighted average of the inverted values ​​is assigned according to a time decay function (i.e., newer data has a higher weight). This processing method is based on the principle of thermal inertia, which means that the macroscopic temperature of the workpiece is not prone to abrupt changes within a few seconds. This effectively filters out temperature reading glitches caused by electromagnetic interference, ensuring the smoothness and stability of the subsequent thermal error compensation signal.

[0158] To achieve metrological-level reconstruction from raw sensor readings to the actual geometric dimensions of the workpiece, the compensation engine in this embodiment does not employ simple linear correction. Instead, it is configured to perform inverse operations based on the superposition principle of a multiphysics decoupling model. This engine aims to extract from macroscopic observations the elastic deformation error caused by contact force, the material thermal expansion error caused by workpiece temperature, and the detection frame system error caused by ambient temperature fluctuations, thereby restoring the essential properties of the workpiece under standard conditions.

[0159] At the input end of the data processing link, the compensation engine first receives multi-source heterogeneous data from various front-end subsystems through a hardware-level time alignment interface. To ensure the physical consistency of causality, the system uses the displacement acquisition time as a reference to measure the real-time contact pressure fed back by the pneumatic servo loop. The equivalent internal temperature of the workpiece obtained by the aforementioned mapping engine and the ambient temperature provided by the environmental monitoring unit Interpolation synchronization is performed. This strict time alignment logic eliminates phase misalignment caused by differences in signal transmission delay, preventing spurious deformation calculation errors during dynamic pressure fluctuations.

[0160] Based on the synchronized data stream, the arithmetic logic unit (ALU) within the compensation engine calculates the true height of the hydraulic sleeve 2 according to the following full-throughput comprehensive compensation equation.

[0161] ;

[0162] in, This is the actual height data; This represents the raw reading of the displacement sensor at the current moment; To effectively apply pressure; This indicates the effective pressure-bearing area of ​​piston 5 in the tilting cylinder; This represents the real-time dynamic stiffness of the workpiece at the current inversion temperature. The structural thermal drift coefficient reflects the amount of zero-point drift of the sensor caused by the thermal expansion and contraction of the C-frame or column. ; is the linear thermal expansion coefficient of the hydraulic sleeve metal material; Thermal expansion normalization factor, used as a denominator term, is used to eliminate the geometric elongation caused by the thermal expansion of materials; and These are the ambient air temperature and the international standard metrological reference temperature measured inside the chamber, respectively.

[0163] Furthermore, considering that sensors are prone to slight nonlinear residual errors across their entire measurement range, a residual lookup table correction module is cascaded at the output of the compensation engine. This module internally stores a systematic residual map obtained from high-level gauge block calibration. For the preliminary residuals calculated by the formula... The module automatically queries and adds a small correction amount based on the range of the value. (usually in) Within a certain range, the nonlinearity error of the grating ruler or LVDT itself is eliminated. This layered and progressive compensation architecture, which combines physical model and data correction, ensures that the device can always output true dimensional data with metrological accuracy under complex operating conditions of varying temperature, force, and environment.

[0164] Please see the appendix Figure 3 To ensure that subsequent dynamic stiffness measurements and thermal inversion processes are based on defined physical boundary conditions, the control unit in this embodiment is configured to execute a strict timing control program before formal data acquisition. The core purpose of this program is to eliminate mechanical transmission chain gaps, squeeze out excess oil film at the contact interface, and bring the workpiece material to a micro-stress steady state through controlled mechanical loading, thereby ensuring the physical repeatability of the measurement data.

[0165] At the beginning of the control sequence, the central processing module 6 performs system-level self-diagnosis. This diagnostic logic is not a simple electrical continuity check, but rather involves sending micro-amplitude high-frequency test pulses to the pneumatic servo valve and simultaneously monitoring the transient response characteristics of the pressure and displacement sensors. The system calculates the dead time and rise time of the step response. If the response delay exceeds a preset safety threshold (e.g., 20ms), or the signal-to-noise ratio of the feedback signal is lower than a predetermined standard, the system will determine that there is a blockage in the air path or a loose sensor coupling, thereby triggering a hardware fault lockout. This self-checking mechanism based on dynamic response ensures the frequency response fidelity of the actuator in subsequent micron-level control.

[0166] After passing the self-test, the system enters the rapid approach and flexible contact phase. To balance detection efficiency and workpiece protection, the servo control unit is configured to execute a dual-speed segmented approximation strategy. Detection board 3 first feeds at a set high speed... (e.g., 20 mm / s) moves downwards to quickly cross the free travel; when the displacement sensor reading shows the detection plate is at the theoretical workpiece height... Enter safe buffer distance When the speed is within a range of 2mm (e.g.), the controller automatically and smoothly switches to a low-speed creep mode. (e.g., 0.5 mm / s). During this process, the system monitors the pressure feedback signal in real time with a high sampling rate. Once detected Exceeding the contact trigger threshold (For example, 0.05 MPa) indicates that physical contact has been established, and the servo valve immediately switches to force control mode to maintain this weak contact force. This contact logic based on position-pressure hybrid triggering effectively avoids rigid collision impacts caused by workpiece dimensional deviations or misalignment.

[0167] After confirming contact, the system performs the core pre-compaction operation. Considering that the surface of the hydraulic sleeve 2 usually has residual rust-preventive oil or cutting fluid, and that the contact surface between the workpiece and the detection plate 3 has randomly distributed rough peaks and valleys at the microscopic level, if dynamic measurement is performed directly, the squeezing flow of the oil film and the plastic crushing of the rough peaks will produce significant nonlinear displacement drift, seriously interfering with the calculation of dynamic stiffness. Therefore, the system controls the cylinder to apply a constant pre-compaction pressure. .

[0168] As a preferred method, this Set to a static reference pressure slightly higher than that used in subsequent dynamic tests. (For example, set to) The physical purpose of this overload preloading strategy is to use high stress to accelerate the rheological equilibrium process at the contact interface, enabling the contact state to quickly cross the unstable region.

[0169] Maintaining precompaction pressure During the process, the system initiates a steady-state monitoring algorithm. This algorithm no longer relies on empirical fixed delays, but instead uses the sliding window statistical principle to dynamically determine whether the system has reached thermodynamic equilibrium by calculating the real-time settlement rate of the displacement signal. The determination logic is based on the following steady-state convergence formula:

[0170] ;

[0171] in, These are values ​​calculated by the steady-state monitoring algorithm. The length of the sliding observation window; It is an instantaneous displacement signal; The discrete sampling time interval; This is the creep rate threshold.

[0172] The system performs the above calculations in a loop, once Once the threshold condition is met, pre-compaction is considered complete. To prevent false numerical stability (e.g., the reading remains unchanged) due to mechanical jamming or sensor drift, the system also incorporates a maximum timeout protection period. (e.g., 5 seconds) and minimum fluctuation detection logic. If in If the internal convergence is still not achieved, the system will report an alarm indicating an abnormal contact surface condition (which usually indicates the presence of foreign objects or excessive oil on the workpiece surface).

[0173] After the pre-compaction stage is successfully completed, the system does not completely unload immediately, but instead smoothly adjusts the pressure back to the working reference pressure. The displacement reading at this moment is recorded as the relative zero-point reference for this measurement cycle. This step, by eliminating the influence of system hysteresis loops, ensures that subsequent dynamic excitation is performed under a tight, linear, and stable physical contact state, thus providing the necessary initial conditions for high signal-to-noise ratio stiffness feature extraction. The PID control principle and sliding filter technique in the above-mentioned pneumatic servo control are well-known techniques to those skilled in the art and will not be elaborated upon here.

[0174] After the system completes pre-compaction and establishes a steady-state baseline, the control logic automatically switches to dynamic measurement mode. In this mode, the processing engine is configured to induce the workpiece's micro-elastic response by actively emitting physical excitations of a specific spectrum, and to accurately extract dynamic stiffness characteristics from a strong noise background using digital signal processing technology, thereby driving the aforementioned temperature inversion and size compensation logic.

[0175] To obtain the dynamic stiffness that characterizes the bulk properties of the material, the pneumatic servo control unit is configured to operate at the current static reference pressure. A continuous sinusoidal micro-perturbation signal is superimposed on top of this. The mathematical form of the excitation signal is defined as follows: Among them, the excitation frequency The choice of is crucial, and in this embodiment, it is preferably set to 5. Up to 15 The low-frequency range (as a preferred method, set to 10) The physical reason for choosing this frequency band is that this frequency is much higher than the first natural frequency of the mechanical frame of the detection device (usually lower than 2). This effectively avoids modal gain errors introduced by structural resonance; simultaneously, the frequency is low enough that the viscous damping and inertial effects within the workpiece are secondary to the elastic restoring force, ensuring that the measurement results primarily reflect the static stiffness characteristics of the material. Micro-perturbation amplitude Usually set to The range of 5% to 10% is chosen to balance the signal-to-noise ratio (SNR) and linearity, ensuring that the induced displacement response has a detectable amplitude while preventing large deformations from deviating from the linear Hooke's Law range of the material.

[0176] Simultaneously with the application of the excitation, the data acquisition unit synchronously acquires the air pressure signal at a sampling rate much higher than the Nyquist frequency (e.g., 1 kHz). With displacement signal Given the significant presence of broadband mechanical noise in industrial environments, the signal processing module in this embodiment does not employ the easily interfered peak-to-peak extraction method. Instead, it incorporates a single-frequency feature extraction algorithm based on the digital lock-in amplification principle. This algorithm utilizes the orthogonality principle of sine functions to accurately separate the amplitude and phase information of components at the same frequency by performing cross-correlation operations between the sampled signal and reference sine and cosine waves. The integral transform formula is as follows:

[0177] ;

[0178] ;

[0179] ;

[0180] in, The input discrete-time series signals correspond to the real-time acquired pressure data sequences. or displacement data sequence ; This represents the total number of sampling points within the integration period; This refers to the dominant amplitude of the demodulated signal at the excitation frequency. , These are the in-phase component and the quadrature component, which together constitute the vector characteristics of the signal; This is the flutter excitation frequency; The sampling frequency.

[0181] Based on the extracted pure amplitude, the computing unit calculates the real-time dynamic stiffness according to the ratio of effective stress to strain. To improve the accuracy of the physical model, a phase correction term was introduced into the calculation formula:

[0182] ;

[0183] in, This refers to the real-time dynamic stiffness (specifically, the measured value after phase correction). This indicates the effective pressure-bearing area of ​​piston 5 in the tilting cylinder; This represents the amplitude of the pressure fundamental wave. This represents the amplitude of the displacement fundamental wave. This is due to phase lag. Key parameters. The phase angle by which the displacement signal lags behind the pressure signal is determined by... Calculated. Introduced. The physical purpose of this method is to eliminate the imaginary contribution caused by the viscous damping of the hydraulic oil film, retaining only the real stiffness characterizing the elastic energy storage properties of the workpiece, thereby achieving physical decoupling of stiffness and damping. To ensure the physical validity and numerical safety of the measurement, the system is equipped with the following dual protection logic:

[0184] Non-zero denominator protection: If Below the minimum displacement response threshold (e.g., 1) (This value is usually determined by the system resolution). If this indicates a high risk of workpiece jamming or sensor failure, the system will stop calculations and report an error to prevent division by zero overflow. Phase angle validity check: The system has a phase angle threshold. (For example If the calculated Physically, this means that the system has experienced severe hydraulic film slippage or mechanical loosening (at which point the damping effect dominates), and the calculated stiffness value no longer represents the material properties. The system will automatically discard this set of data and attempt to reload.

[0185] In obtaining high confidence Then, the data flow enters the multi-dimensional decoupling and inversion logic. The system will... The temperature is passed as input to the aforementioned stiffness and temperature mapping engine to solve for the internal temperature of the workpiece. Then, the multi-dimensional error compensation engine is invoked to... Real-time displacement mean Real-time average pressure and ambient temperature Substituting into the full throughput compensation equation, the true height of the workpiece is calculated. .

[0186] To further improve the robustness of the results, this embodiment employs iterative convergence logic. The process from excitation to solution described above is configured to be executed continuously while maintaining contact. One cycle (e.g.) The system performs real-time calculations. Second-rate Standard deviation of results Only when Less than the preset precision metrology level convergence threshold (e.g.) Only when the current measurement condition is stable will the system determine that the current measurement condition is stable and output the last calculated value as the final result; if the error exceeds the tolerance, it indicates that the system is affected by environmental vibration or thermal instability. The system will automatically trigger the retest mechanism to extend the pre-compaction time or increase the number of excitation cycles.

[0187] The actual geometric height of the workpiece under standard conditions is calculated by the aforementioned multi-dimensional error compensation engine. Afterward, the processing flow enters the final decision-making and execution stage. In this embodiment, the judgment execution module is not limited to a single pass / fail screening, but is configured to execute advanced quality management logic based on statistical process control (SPC), aiming to achieve graded matching of precision parts and dynamic correction of upstream processing parameters through intelligent distribution of data flow.

[0188] To address the risk of false acceptance or rejection due to measurement uncertainty at tolerance boundaries in traditional binary decision logic, this system introduces a dynamic confidence interval decision algorithm with a guard band. The decision unit is pre-loaded with a nominal target value. Upper tolerance limit and lower tolerance limit Based on this, the system considers the expanded uncertainty of the device itself. As a preferred method, this value is determined by GR&R repeatability and reproducibility testing, for example, set as the effective acceptable range for automatic shrinkage of 0.5 μm. Only when the measurement result falls within this shrinkage range is the workpiece statistically considered securely acceptable. The specific judgment logic follows the following set of inequalities:

[0189] ;

[0190] in, The result is the size determination. The actual dimensions of the workpiece output by the aforementioned multi-dimensional error compensation engine; The maximum and minimum permissible limit dimensions specified in the product drawings; To expand the uncertainty; For the protection zone coefficient; , These correspond to the states of excessively large and excessively small deviations, respectively.

[0191] For workpieces deemed OK, in order to meet the selection and assembly requirements of precision-fitting components such as hydraulic spool valves, the grouping module further performs micron-level boxing operations. For example, each 1 As a group. The system is based on... The specific numerical calculation of its grouping index. The calculation formula is as follows:

[0192] ;

[0193] in This is the preset packet bandwidth; This indicates the floor function; This is the minimum permissible limit size; For the protection zone coefficient; This is the actual height data; To expand the uncertainty; This represents the total number of sampling points.

[0194] In addition to real-time processing of the current workpiece, the central processing module in this embodiment is also equipped with a trend monitoring and feedback correction engine. This engine sends compensation commands to the front-end CNC machining equipment (such as cylindrical grinding machines or honing machines), thereby forming a closed-loop manufacturing system of machining measurement and correction. This engine does not react to deviations in individual workpieces (to avoid system oscillations caused by single-point random errors), but instead calculates the systematic process drift of the machining dimensions based on a moving average model. :

[0195] ;

[0196] In the formula, To adjust the sliding window size; The nominal target value; This is the actual height data.

[0197] At the physical execution level, based on the above calculations, the system controller drives the sorting robot or pneumatic diversion valve to feed the workpiece into the corresponding discharge chute. Simultaneously, the laser marking unit is triggered, engraving a unique serial number and measured dimensions onto the workpiece's end face. And a two-dimensional barcode with a grouping code. This action establishes a single-piece-level product traceability link, enabling subsequent assembly stages to directly obtain the precise microscopic dimensions of the sleeve by scanning the code, without the need for secondary measurement.

[0198] Specific application examples:

[0199] This embodiment selects an HV-200 automotive automatic transmission hydraulic valve sleeve as the test object. The material is 20CrMnTi carburized steel, with a nominal height of 50.0000 mm and a tolerance requirement of ±0.0030 mm. The workpiece has just come off the honing process, with a randomly thick cutting oil film adhering to its surface, and the workpiece temperature fluctuates between 30°C and 70°C due to processing heat. The testing system is set with a static reference pressure. The pressure is 0.6 MPa, and the flutter excitation frequency is... 25Hz, pressure modulation amplitude The pressure is 0.06 MPa, and the sampling frequency is set to 1000 Hz. After the measurement is started, the system first acquires the raw data from the displacement sensor. Figure 4 As shown, the time-domain waveform records the dynamic response of the sensor under a reference pressure of 0.6 MPa. It can be seen from the figure that the displacement signal is not a straight line, but exhibits periodic sinusoidal fluctuations around a certain average value, and the waveform is superimposed with random environmental noise and low-frequency mechanical drift. The system then performs a Fast Fourier Transform on this time-domain signal, as shown... Figure 5 As shown in the figure, the solid line clearly depicts the spectral characteristics of the displacement signal; there is a significant DC component at 0 Hz, the amplitude of which corresponds to the average displacement component. A distinct flutter signal characteristic peak exists at 25Hz. The sharpness of this peak indicates a good excitation response, and its corresponding dynamic amplitude... The reading is approximately 3.50 mm. The system follows the formula... Calculate the real-time dynamic stiffness of the workpiece, using a preset value. (approximately 1256mm) 2 Substituting the stress amplitude into the calculation, we obtain the dynamic stiffness value under the current state. It is approximately 15072 N / mm.

[0200] After obtaining the stiffness data, the system calls... Figure 6 The stiffness versus temperature calibration curves shown are used for thermodynamic inversion. This figure establishes and visualizes the physical law governing the change of material stiffness with temperature: the scatter points represent calibration sampling points of the standard part at different temperatures, and the solid line is a second-order fitting model generated based on the sampling points, showing that the stiffness monotonically decreases with increasing temperature. The standard stiffness at 20℃ is known. It is 15250 N / mm, according to Figure 6 The slope of the curve indicates that the stiffness value of 197 kN / mm measured by the current inversion derivation path corresponds to an internal workpiece temperature of approximately 61.5℃. This suggests that the workpiece has experienced a temperature rise relative to standard room temperature. .

[0201] Finally, the system uses the formula Multidimensional error decoupling is performed. This calculation incorporates the material's linear thermal expansion coefficient. The numerator compensates for the elastic compression caused by the 0.6MPa static pressure, while the denominator eliminates the thermal expansion caused by the 30℃ temperature rise. Figure 7 The figure visually demonstrates a comparison of measurement results between the method of this embodiment and the traditional method. The horizontal axis represents sample numbers with gradually increasing temperature, and the vertical axis represents the dimensional deviation values. It can be seen that the traditional pneumatic measurement curve marked with squares shows a clear upward trend with increasing sample temperature, with many data points exceeding ±3... The tolerance zone is limited because this method only deducts elastic deformation at room temperature and does not consider the effects of thermal expansion and oil film. Conversely, the thermo-coupling measurement curve of this invention, marked with a circular mark below, remains consistently near 0, indicating the true size after decoupling calculation. The measured height (approximately 50.0577 mm) is consistent with the true height measured by a high-precision grating length measuring machine after the workpiece has been cleaned and cooled, verifying that the actual size of the workpiece is within the acceptable tolerance range and effectively solving the problem of misjudgment under variable temperature environment.

Claims

1. A machining and inspection system for hydraulic bushings, characterized in that, include: The pneumatic servo control module (7) is used to control the tilting cylinder (5) to apply a composite pressure waveform containing a static reference pressure component and a periodic chatter pressure component to the hydraulic sleeve (2); The signal acquisition and processing module (8) is used to acquire the time-domain displacement signal of the hydraulic sleeve (2) detected by the displacement sensor (4), and separate the average displacement component and the displacement amplitude component corresponding to the flutter frequency through frequency domain analysis. The thermal inversion calculation module (9) calculates the real-time dynamic stiffness based on the displacement amplitude component and pressure component, uses the material stiffness thermal softening model to back-calculate the current temperature rise value from the stiffness attenuation amount, and combines the average displacement component, the static reference pressure component, the real-time dynamic stiffness and the current temperature rise value to calculate the real height data of the hydraulic sleeve (2) under standard temperature and no external force. The compensation feedback module (10) is used to receive the actual height data and generate a size qualification judgment result or output processing compensation parameters based on the actual height data.

2. The machining and inspection system for a hydraulic bushing according to claim 1, characterized in that, The pneumatic servo control module (7) is connected to the electric proportional valve to generate a composite control signal containing a DC bias voltage and an AC sinusoidal voltage. The DC bias voltage drives the electric proportional valve to output the static reference pressure component, and the AC sinusoidal voltage drives the electric proportional valve to output the chatter pressure component. The pneumatic servo control module (7) sets the amplitude and frequency of the chatter pressure component based on the dynamic friction characteristics of the tilting cylinder (5).

3. The machining and inspection system for a hydraulic bushing according to claim 1, characterized in that, The signal acquisition and processing module (8) performs windowing truncation and spectrum analysis on the time-domain displacement signal, extracts the amplitude at zero frequency in the spectrum as the average displacement component, and extracts the peak value of the spectrum at the excitation frequency of the flutter pressure component as the displacement amplitude component.

4. The machining and inspection system for a hydraulic bushing according to claim 1, characterized in that, The process by which the thermal inversion calculation module (9) calculates the real-time dynamic stiffness includes: Obtain the pressure modulation amplitude of the flutter pressure component, calculate the dynamic force amplitude by multiplying the pressure modulation amplitude with the effective pressure area of ​​the tilting cylinder (5), and calculate the ratio of the dynamic force amplitude to the displacement amplitude component to obtain the real-time dynamic stiffness.

5. The machining and inspection system for a hydraulic bushing according to claim 1, characterized in that, The material stiffness thermal softening model includes the stiffness thermal softening coefficient of the hydraulic sleeve (2) material; The thermal inversion calculation module (9) calculates the ratio of the real-time dynamic stiffness to the standard reference stiffness, and solves the current temperature rise value based on the mapping relationship between the ratio and the stiffness thermal softening coefficient.

6. The machining and inspection system for a hydraulic bushing according to claim 1, characterized in that, The logic of the thermal inversion calculation module (9) for calculating the true height data includes: The elastic compression correction caused by external force is calculated using the static reference pressure component, the effective pressure area of ​​the tilting cylinder (5), and the real-time dynamic stiffness. The thermal expansion correction caused by temperature is calculated using the current temperature rise value and the linear thermal expansion coefficient of the hydraulic sleeve (2) material; The average displacement component is added to the elastic compression correction and the thermal expansion correction is subtracted to obtain the true height data.

7. The machining and inspection system for a hydraulic bushing according to claim 1, characterized in that, It also includes a central processing module (6), which is connected to the pneumatic servo control module (7) and the signal acquisition and processing module (8). The central processing module (6) controls the pneumatic servo control module (7) to drive the tilting cylinder (5) to perform a constant pressure pre-compaction action, and receives the settlement rate of the displacement signal fed back by the signal acquisition and processing module (8). When the settling rate is less than a preset threshold, the central processing module (6) sends an instruction to the pneumatic servo control module (7) to superimpose and output the flutter pressure component.

8. The machining and inspection system for a hydraulic bushing according to claim 5, characterized in that, The standard reference stiffness is data obtained by applying an excitation signal consistent with the composite pressure waveform parameters to a calibrated hydraulic sleeve (2) sample under standard ambient temperature and performing multiple measurements and statistics.

9. The machining and inspection system for a hydraulic bushing according to claim 1, characterized in that, The compensation feedback module (10) is specifically used for: Compare the actual height data with the preset standard tolerance range; If the actual height data is within the standard tolerance range, the dimension acceptance result is generated; If the actual height data exceeds the standard tolerance range, the difference between the actual height data and the target nominal size is calculated to obtain the size deviation value, and the size deviation value is converted into tool compensation data as the machining compensation parameter and sent to the front-end machine tool CNC system.

10. A machining and inspection device for hydraulic bushings, characterized in that, A machining and inspection system for a hydraulic bushing according to any one of claims 1-9 includes a base (1), a hydraulic sleeve (2) to be tested is placed on the top of the base (1), a detection plate (3) is provided on one side of the base (1), a tilting cylinder (5) is fixedly connected to the top of the base (1), the detection plate (3) is fixedly connected to the output end of the tilting cylinder (5) for driving the detection plate (3) to tilt and make the bottom surface of the detection plate (3) abut against the upper surface of the hydraulic sleeve (2), and a displacement sensor (4) is provided on the outside of the detection plate (3) for detecting the distance signal relative to the base (1).

Images