A high-precision positioning control method and system of a servo electro-hydraulic actuator

Abstract

The application belongs to the technical field of hydraulic transmission and control, and discloses a high-precision positioning control method and system of a servo electro-hydraulic actuator; by acquiring real-time operation parameters of the servo electro-hydraulic actuator under different working frequencies, a servo valve dynamic hysteresis characteristic curve is constructed, a hysteresis characteristic change mode is extracted, and a dynamic influence model of the hysteresis characteristic change on positioning accuracy is established. Based on the model, in combination with a hydraulic system pressure fluctuation value, a frequency adaptive compensation strategy is designed to realize real-time correction of the servo valve control signal. The method innovatively adopts a differentiated strategy combining low-frequency feedforward compensation and high-frequency feedback compensation, and a smooth transition mechanism is designed to overcome the positioning accuracy fluctuation problem of the traditional control method under the frequency change working condition. The application realizes high-precision positioning control of the servo electro-hydraulic actuator in the full frequency range, and improves the positioning accuracy and stability of the system under dynamic working conditions.

Description

Technical Field

[0001] This invention relates to the field of hydraulic transmission and control technology, and more specifically, to a high-precision positioning control method and system for a servo electro-hydraulic actuator. Background Technology

[0002] Servo electro-hydraulic actuators, as key power transmission and control components in modern industrial automation, aerospace, military equipment, and precision manufacturing, are widely used in various applications requiring precise force and position control due to their high power density, fast response, and excellent load capacity. With the rapid development of intelligent manufacturing, the requirements for the control precision of servo electro-hydraulic actuators are constantly increasing, especially in fields such as aerospace steering gear, marine steering gear, construction machinery, CNC machine tools, and heavy equipment, where micron-level or even sub-micron-level positioning accuracy has become the industry standard.

[0003] However, in practical applications of servo electro-hydraulic actuators, the complex nonlinear behavior of the servo valve's hysteresis characteristics, which varies with the operating frequency, is a key bottleneck restricting the dynamic positioning accuracy of the system. Existing technologies generally employ static or quasi-static hysteresis models, neglecting the significant differences in hysteresis width, shape, and phase lag of the servo valve at different operating frequencies. This leads to the failure of compensation strategies under frequency-changing conditions. Especially in high-precision applications such as aerospace servos, CNC machine tools, and precision manufacturing equipment, when the system frequently switches operating frequencies, the statically calibrated compensation parameters cannot adapt to changes in dynamic hysteresis characteristics, causing overshoot, oscillation, or even instability. Traditional PID control and lookup table compensation methods cannot capture the intrinsic correlation between hysteresis characteristics and frequency and pressure fluctuations. Under alternating loads and frequency fluctuations, positioning accuracy deteriorates severely, with accuracy fluctuations reaching 3-5 times the rated accuracy. Furthermore, due to the lack of a dynamic mapping model between hysteresis characteristics and positioning deviation, the system cannot predict and compensate for transient errors during frequency transitions in advance. This results in significant path deviations and position jitter in valve control systems in high-dynamic response applications such as flight control systems and industrial robots. Furthermore, existing technologies fail to comprehensively consider the coupling effect of pressure fluctuations and hysteresis characteristics in hydraulic systems. Especially in the high-frequency region, the nonlinear amplification effect caused by the interaction between the two leads to a sharp decline in control accuracy. A single compensation strategy cannot simultaneously cope with different operating conditions dominated by low-frequency hysteresis and high-frequency pressure fluctuations, making it difficult for the system to balance static accuracy and dynamic performance.

[0004] In view of this, the present invention proposes a high-precision positioning control method and system for a servo electro-hydraulic actuator to solve the above problems. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a high-precision positioning control method for a servo electro-hydraulic actuator, comprising:

[0006] The real-time operating parameters of the servo electro-hydraulic actuator at different operating frequencies are obtained. The real-time operating parameters include the valve core displacement of the servo valve, the input current signal of the servo valve, the piston displacement of the hydraulic cylinder, and the pressure fluctuation value of the hydraulic system.

[0007] Based on the time-series correspondence between the valve core displacement and the input current signal of the servo valve in the real-time operating parameters, the dynamic hysteresis characteristic curve of the servo valve at each operating frequency is constructed.

[0008] Based on the differences in curve shape within adjacent operating frequency ranges in the dynamic hysteresis characteristic curve, the hysteresis characteristic variation pattern of the servo valve at different operating frequencies is extracted.

[0009] Based on the variation pattern of hysteresis characteristics and the deviation distribution of piston displacement of hydraulic cylinder, a dynamic influence model of servo valve hysteresis characteristics variation on positioning accuracy is constructed.

[0010] Based on the dynamic influence model and combined with the pressure fluctuation value of the hydraulic system, a frequency adaptive compensation strategy for the servo valve is designed to obtain the compensation control signal of the servo valve at each working frequency.

[0011] The input current signal is corrected in real time based on the compensation control signal to obtain the corrected servo valve control signal.

[0012] The servo electro-hydraulic actuator is driven by the modified servo valve control signal to control the piston displacement of the hydraulic cylinder and achieve high-precision positioning.

[0013] Furthermore, the steps for constructing the dynamic hysteresis characteristic curve of the servo valve at each operating frequency include:

[0014] The valve core displacement and input current signal of the servo valve in the real-time operating parameters are time-aligned to obtain the time data pairs of valve core displacement and input current signal at each operating frequency;

[0015] The timing data pairs are segmented, and each operating frequency timing data pair is divided into multiple time segments, with each time segment corresponding to one action cycle of the servo valve.

[0016] Within each time segment, calculate the displacement difference between the positive and negative change ranges of the valve core displacement, and record it as the local hysteresis value for each time segment.

[0017] Based on the local hysteresis value of each time segment and the rate of change of the amplitude of the input current signal, a scatter distribution of the local hysteresis value as a function of the rate of change of amplitude is constructed.

[0018] Curve fitting is performed on the scattered distribution to obtain the dynamic hysteresis characteristic curve of the servo valve at each operating frequency. The horizontal axis of the dynamic hysteresis characteristic curve is the rate of change of the amplitude of the input current signal, and the vertical axis is the local hysteresis value.

[0019] Furthermore, the steps for extracting the hysteresis characteristic variation pattern of the servo valve at different operating frequencies include:

[0020] The dynamic hysteresis characteristic curve is decomposed in the frequency domain to obtain the spectral components of the dynamic hysteresis characteristic curve at each operating frequency.

[0021] The amplitude and phase differences of the spectral components of the dynamic hysteresis characteristic curve within adjacent operating frequency intervals are calculated and denoted as the spectral change characteristics.

[0022] Based on the spectral variation characteristics, the morphological change trend of the dynamic hysteresis characteristic curve in adjacent operating frequency ranges is extracted. The morphological change trend includes the curve slope change rate, the number of curve inflection points, and the curve oscillation period.

[0023] Based on the trend of morphological changes, a retracement characteristic change pattern is constructed. The retracement characteristic change pattern uses the operating frequency as the independent variable and the quantitative indicators of the morphological change trend as the dependent variable. The quantitative indicators include the mean of the rate of change of the curve slope, the increment of the number of curve inflection points, and the ratio of the curve oscillation period.

[0024] Furthermore, the steps for constructing a dynamic impact model of the servo valve hysteresis characteristic variation on positioning accuracy include:

[0025] The deviation between the actual value and the target value of the piston displacement of the hydraulic cylinder is recorded as the positioning deviation.

[0026] Time-frequency analysis of positioning deviation is performed to obtain the time-domain fluctuation characteristics and frequency-domain distribution characteristics of positioning deviation at each operating frequency;

[0027] Based on the quantitative indicators of the morphological change trend in the hysteresis characteristic change pattern, the correlation coefficient between the quantitative indicators and the time-domain fluctuation characteristics of the positioning deviation is calculated and denoted as the time-domain influence weight.

[0028] Based on the quantitative index of the morphological change trend in the hysteresis characteristic change pattern, calculate the correlation coefficient between the quantitative index and the frequency domain distribution characteristics of the positioning deviation, and denote it as the frequency domain influence weight.

[0029] A dynamic influence model is constructed based on the time-domain influence weight and the frequency-domain influence weight. The dynamic influence model takes the operating frequency as input and the predicted value of the positioning deviation as output. The predicted value is determined by the weighted sum of the time-domain influence weight and the frequency-domain influence weight.

[0030] Furthermore, the steps for designing a frequency adaptive compensation strategy for the servo valve include:

[0031] Based on the dynamic influence model, predict the positioning deviation at each operating frequency;

[0032] Based on the pressure fluctuation value of the hydraulic system, calculate the periodic and random components of the pressure fluctuation value at each working frequency, and denot them as pressure fluctuation characteristics;

[0033] Based on the characteristics of pressure fluctuations, the gain effect of pressure fluctuation values ​​on positioning deviation is extracted and denoted as the pressure gain factor.

[0034] Based on the predicted positioning deviation and the pressure gain factor, a frequency adaptive compensation strategy is designed, which includes:

[0035] When the operating frequency is lower than the preset frequency threshold, a feedforward compensation signal is designed based on the predicted value of the positioning deviation. The amplitude of the feedforward compensation signal is proportional to the predicted value of the positioning deviation.

[0036] When the operating frequency is greater than the preset frequency threshold, a feedback compensation signal is designed based on the pressure gain factor. The phase of the feedback compensation signal is out of phase with the periodic component of the pressure fluctuation value.

[0037] The feedforward compensation signal and the feedback compensation signal are weighted and fused to obtain the compensation control signal. The weights of the weighted fusion are determined by the normalized value of the operating frequency.

[0038] Furthermore, the step of real-time correction of the input current signal based on the compensation control signal includes:

[0039] The compensation control signal is smoothed in the time domain to obtain the smoothed compensation control signal. The time domain smoothing process adopts an adaptive window length based on the operating frequency, and the adaptive window length is inversely proportional to the operating frequency.

[0040] The smoothed compensation control signal is superimposed on the input current signal to obtain the superimposed current signal;

[0041] The superimposed current signal is subjected to amplitude limiting processing to obtain the corrected servo valve control signal. The upper and lower limits of the amplitude limiting processing are determined by the rated operating range of the servo valve.

[0042] Furthermore, the step of curve fitting to the scatter distribution includes:

[0043] Outlier removal is performed on the scatter distribution. The criteria for outlier removal are: the difference between the local hysteresis value and the mean of the scatter distribution is greater than a preset outlier threshold.

[0044] The distribution of scattered points after outliers is piecewise fitted, where the segment points of the piecewise fitting are determined by the inflection points of the rate of change of the amplitude of the input current signal.

[0045] Within each segment, a spline interpolation-based curve fitting method is used to obtain the segmented fitting curve;

[0046] All piecewise fitted curves are spliced ​​together to obtain the dynamic hysteresis characteristic curve. The smoothness of the splicing is constrained by the consistency of the endpoint slopes of adjacent piecewise fitted curves.

[0047] Furthermore, methods for obtaining time-domain fluctuation characteristics include:

[0048] Wavelet decomposition is performed on the positioning error to obtain the wavelet coefficients of the positioning error at different time scales;

[0049] Calculate the energy proportion of wavelet coefficients at each time scale, and denote it as the time-domain fluctuation energy distribution;

[0050] Based on the energy distribution of time-domain fluctuations, the time-domain fluctuation characteristics of the positioning deviation are extracted. These characteristics include the peak time position of the wavelet coefficients, the variance of the energy proportion of the wavelet coefficients, and the periodic components of the wavelet coefficients.

[0051] Furthermore, methods for obtaining pressure fluctuation characteristics include:

[0052] Perform a Fourier transform on the pressure fluctuation value to obtain the spectral components of the pressure fluctuation value;

[0053] Based on the spectral components, the periodic components of the pressure fluctuation value are extracted, wherein the frequency range of the periodic components is determined by the harmonic components of the operating frequency.

[0054] Subtract the periodic component from the pressure fluctuation value to obtain the random component of the pressure fluctuation value;

[0055] The amplitude of the periodic component and the variance of the random component are calculated and denoted as the pressure fluctuation characteristics.

[0056] A high-precision positioning control system for a servo electro-hydraulic actuator, comprising a high-precision positioning control method for the servo electro-hydraulic actuator, including:

[0057] The parameter acquisition module is used to acquire the real-time operating parameters of the servo electro-hydraulic actuator at different operating frequencies. The real-time operating parameters include the valve core displacement of the servo valve, the input current signal of the servo valve, the piston displacement of the hydraulic cylinder, and the pressure fluctuation value of the hydraulic system.

[0058] The dynamic hysteresis modeling module is used to construct the dynamic hysteresis characteristic curve of the servo valve at each operating frequency based on the time-series correspondence between the valve core displacement and the input current signal in the real-time operating parameters.

[0059] The hysteresis change analysis module is used to extract the hysteresis characteristic change pattern of the servo valve at different operating frequencies based on the curve shape differences in adjacent operating frequency ranges in the dynamic hysteresis characteristic curve.

[0060] The influence model building module is used to construct a dynamic influence model of the servo valve hysteresis characteristic change on positioning accuracy based on the deviation distribution of the hysteresis characteristic change pattern and the piston displacement of the hydraulic cylinder.

[0061] The compensation strategy design module is used to design a frequency adaptive compensation strategy for the servo valve based on the dynamic influence model and the pressure fluctuation value of the hydraulic system, and to obtain the compensation control signal of the servo valve at each working frequency.

[0062] The control signal correction module is used to correct the input current signal in real time according to the compensation control signal, and obtain the corrected servo valve control signal.

[0063] The high-precision positioning and execution module is used to drive the servo electro-hydraulic actuator based on the corrected servo valve control signal, control the piston displacement of the hydraulic cylinder, and achieve high-precision positioning.

[0064] The technical effects and advantages of the high-precision positioning control method and system for a servo electro-hydraulic actuator of the present invention are as follows:

[0065] This invention maintains good control performance across the entire frequency range, improving positioning accuracy and stability under dynamic conditions. In practical applications, it reduces system position tracking errors and oscillations during dynamic response, resulting in smoother and more precise actuator movement, effectively extending equipment lifespan and reducing energy consumption. Particularly in industrial scenarios with high dynamic and high precision requirements, it enhances the system's resistance to external interference and load variations, enabling the equipment to maintain stable and reliable performance output even in complex and changing working environments. Furthermore, this invention achieves seamless and precise control across different operating frequencies, eliminating performance gaps in the frequency transition region. Attached Figure Description

[0066] Figure 1 This is a schematic diagram of a high-precision positioning control method for a servo electro-hydraulic actuator according to the present invention;

[0067] Figure 2 This is a schematic diagram of a high-precision positioning control system for a servo electro-hydraulic actuator according to the present invention. Detailed Implementation

[0068] The technical solutions of 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.

[0069] This application provides a high-precision positioning control method and system for a servo electro-hydraulic actuator. The execution entities of the system include, but are not limited to, the following: a hydraulic servo controller, an electro-hydraulic actuator system monitoring platform, a high-precision position control device, and a servo valve dynamic characteristic analyzer, which can be considered as general control nodes in this application. The high-precision positioning control system includes, but is not limited to, at least one of the following: a servo electro-hydraulic actuator frequency adaptive control unit, a distributed hydraulic system dynamic compensator, and an intelligent servo valve hysteresis characteristic modeler.

[0070] This invention provides a high-precision positioning control method and system for a servo electro-hydraulic actuator. By acquiring the operating parameters of the servo electro-hydraulic actuator in real time at different operating frequencies, a precise dynamic hysteresis characteristic curve and hysteresis characteristic change pattern of the servo valve are constructed. A dynamic influence model is generated by combining the hydraulic cylinder piston displacement and hydraulic system pressure fluctuation values. A frequency adaptive compensation strategy is used to achieve precise correction and high-precision positioning of the servo valve control signal. It has high adaptability, capable of optimizing control parameters in real time according to the system operating frequency and hysteresis characteristics, significantly improving positioning accuracy and reducing system oscillation.

[0071] Please see Figure 1 In this embodiment of the invention, the detailed implementation steps of a high-precision positioning control method for a servo electro-hydraulic actuator include:

[0072] Real-time operating parameters of the servo electro-hydraulic actuator at different operating frequencies are acquired. These parameters include key indicators such as the servo valve spool displacement, the servo valve input current signal, the hydraulic cylinder piston displacement, and the hydraulic system pressure fluctuation, all collected in real time via a sensor network. The servo valve spool displacement and input current signal directly reflect the dynamic response characteristics of the servo valve, the hydraulic cylinder piston displacement characterizes the system's positioning effect, and the hydraulic system pressure fluctuation describes the system's dynamic characteristics during operation. These data provide a foundation for hysteresis characteristic analysis and the formulation of compensation strategies, ensuring the accuracy and effectiveness of control.

[0073] Based on the time-series correlation between the servo valve spool displacement and the input current signal in the real-time operating parameters, a dynamic hysteresis characteristic curve for the servo valve at each operating frequency is constructed. This dynamic hysteresis characteristic curve quantifies the nonlinear characteristics of the servo valve at different operating frequencies, directly determining the upper limit constraint of the system's positioning accuracy. This curve is obtained through time-series analysis of the spool displacement and the input current signal, and calculation of the hysteresis value, ensuring that the control strategy matches the actual characteristics of the servo valve.

[0074] Based on the differences in curve shape within adjacent operating frequency ranges in the dynamic hysteresis characteristic curve, this module extracts the hysteresis characteristic variation patterns of the servo valve at different operating frequencies. First, it performs frequency domain decomposition on the dynamic hysteresis characteristic curve to identify the variation patterns of hysteresis characteristics at different operating frequencies. Then, by quantifying the trend of hysteresis change, it constructs the hysteresis characteristic variation pattern, laying the foundation for subsequent influence model construction.

[0075] Based on the variation pattern of hysteresis characteristics and the deviation distribution of piston displacement in hydraulic cylinders, a dynamic influence model of servo valve hysteresis characteristics on positioning accuracy is constructed. This module analyzes the correlation between hysteresis characteristics and positioning deviation, quantifies the degree of influence of hysteresis characteristics on positioning accuracy, establishes a predictive model, and provides a theoretical basis for compensation strategies.

[0076] Based on a dynamic influence model and considering the pressure fluctuations in the hydraulic system, a frequency-adaptive compensation strategy for the servo valve is designed to obtain the compensation control signal for the servo valve at each operating frequency. This module designs targeted feedforward and feedback compensation mechanisms according to the characteristic differences in different operating frequency ranges, achieving precise adjustment of the servo valve control signal.

[0077] The input current signal is corrected in real time based on the compensation control signal to obtain the corrected servo valve control signal. This module employs smoothing and limiting measures to ensure the stability and usability of the corrected signal.

[0078] The servo electro-hydraulic actuator is driven by the modified servo valve control signal to control the piston displacement of the hydraulic cylinder and achieve high-precision positioning.

[0079] In this embodiment of the invention, the detailed implementation steps for constructing the dynamic hysteresis characteristic curve of the servo valve at each operating frequency based on the time-series correspondence between the valve core displacement and the input current signal in the real-time operating parameters include:

[0080] Timing alignment is performed on the servo valve spool displacement and input current signal in real-time operating parameters to obtain time-series data pairs of spool displacement and input current signal at each operating frequency. Timing alignment is a prerequisite for analyzing hysteresis characteristics, ensuring accurate correspondence between displacement and current signals in the time dimension. The alignment process employs cross-correlation analysis, determining the optimal matching position by calculating the cross-correlation function of the two signal sequences to compensate for potential phase delays in the system. The result is time-synchronized displacement-current data pairs, providing an accurate data foundation for subsequent hysteresis analysis. For high-frequency signals, a small-time-window moving average preprocessing method is used to eliminate high-frequency noise interference; for low-frequency signals, trend extraction technology is employed to identify the main trends in signal variation. This multi-scale preprocessing ensures the accuracy and robustness of timing alignment.

[0081] The time-series data pairs are segmented, dividing each operating frequency pair into multiple time segments, each corresponding to one action cycle of the servo valve. Segmentation is a crucial step in analyzing local hysteresis characteristics, discretizing the continuous signal into periodic units with clear physical meaning. The processing is based on the signal's periodicity, identifying each complete action cycle through peak detection or zero-crossing analysis. For different operating frequencies, the segment window length is dynamically adjusted to ensure each time segment contains a complete action cycle. The segmentation results provide the basic units for subsequent local hysteresis value calculation; each segment is both independent and interconnected, collectively forming a complete description of the hysteresis characteristics.

[0082] Within each time segment, the displacement difference between the positive and negative ranges of the valve core displacement is calculated and recorded as the local hysteresis value for each time segment. The local hysteresis value is a direct indicator of hysteresis characteristics, reflecting the degree of nonlinearity of the servo valve within a single action cycle. The calculation process first identifies the direction of valve core displacement change, dividing it into a positive range (displacement increase) and a negative range (displacement decrease); then, at points where the input current signal has the same value, the displacement difference between the two directions is calculated to obtain the local hysteresis value. The calculation formula is:

[0083] ;

[0084] in, For the first Local lag values ​​for each time segment, For the positive change range in current value Displacement value at that location, For the reverse change interval at the same current value Displacement value at that location, This represents the number of sampling points.

[0085] The calculation of local hysteresis eliminates the uncertainty of single-point measurement and improves the reliability of hysteresis estimation. Local hysteresis values ​​are the basic data points for subsequent characteristic curve construction, accurately reflecting the hysteresis characteristics of the servo valve under different operating states.

[0086] Based on the local hysteresis value and the rate of change of the input current signal amplitude for each time segment, a scatter distribution of the local hysteresis value as a function of the rate of change of amplitude is constructed. The scatter distribution is a visual representation of the hysteresis characteristics, revealing the potential relationship between the hysteresis value and the rate of change of current. The construction process first calculates the rate of change of the input current signal amplitude within each time segment, i.e., the speed at which the current value changes with time; then, the calculated rate of change and the corresponding local hysteresis value are paired and plotted on a two-dimensional plane. The scatter distribution presents the dynamic variation law of the hysteresis characteristics, providing data support for subsequent curve fitting. This rate-of-change-based analysis method overcomes the limitations of traditional static hysteresis analysis and can capture the nonlinear characteristics of servo valves under dynamic operating conditions.

[0087] Curve fitting was performed on the scattered data to obtain the dynamic hysteresis characteristic curve of the servo valve at each operating frequency. The horizontal axis of the dynamic hysteresis characteristic curve represents the rate of change of the input current signal amplitude, and the vertical axis represents the local hysteresis value. Curve fitting is a crucial step in obtaining a continuous hysteresis characteristic model, transforming discrete experimental data into a mathematical model that can be used for prediction and compensation. The fitting process employs a robust nonlinear regression method, which can resist the interference of outliers and extract the main trends in the data. The fitting model uses a polynomial or piecewise polynomial function, dynamically adjusting the model complexity according to the complexity of the scattered data distribution. This avoids both insufficient accuracy due to underfitting and decreased generalization ability due to overfitting. The fitting results are unique dynamic hysteresis characteristic curves for each operating frequency, intuitively demonstrating the variation of the hysteresis value with the rate of change of current, providing a basic model for subsequent extraction of hysteresis characteristic change patterns.

[0088] In this embodiment of the invention, the detailed implementation steps for extracting the hysteresis characteristic variation pattern of the servo valve at different operating frequencies based on the curve shape differences within adjacent operating frequency ranges in the dynamic hysteresis characteristic curve include:

[0089] Frequency domain decomposition is performed on the dynamic hysteresis characteristic curve to obtain the spectral components of the curve at each operating frequency. Frequency domain decomposition is an effective method for analyzing the essential characteristics of the curve, transforming complex time-domain characteristics into a frequency-domain expression, which facilitates the identification of the main influencing factors. The decomposition process uses Discrete Fourier Transform (DFT), treating the dynamic hysteresis characteristic curve as a discrete signal and calculating its spectral characteristics. The calculation formula is as follows:

[0090] ;

[0091] in, These are discrete sampling points for the dynamic hysteresis characteristic curve. Its spectral components, The number of sampling points. For frequency index or frequency component number, here It represents the imaginary unit.

[0092] The spectral components consist of both amplitude and phase spectra, which together describe the frequency domain characteristics of the hysteresis. Low-frequency components reflect the basic trends of the hysteresis, while high-frequency components correspond to subtle variations. The frequency domain decomposition results provide a new perspective and data foundation for subsequent analysis of characteristic variations.

[0093] The amplitude and phase differences of the spectral components of the dynamic hysteresis characteristic curve within adjacent operating frequency intervals are calculated and denoted as spectral variation characteristics. Spectral variation characteristics are a direct indicator of the morphological differences in the curves, reflecting the law governing the variation of hysteresis characteristics with operating frequency. The calculation process compares the amplitude and phase of the spectral components for each pair of adjacent operating frequencies to obtain the difference index. The amplitude difference calculation uses a relative rate of change, eliminating the influence of absolute values; the phase difference calculation considers the periodicity of the phase, avoiding calculation errors caused by 2π jumps. Spectral variation characteristics provide an objective quantitative basis for extracting morphological change trends, ensuring the accuracy and reliability of the change patterns.

[0094] Based on the spectral variation characteristics, the morphological change trend of the dynamic hysteresis characteristic curve within adjacent operating frequency ranges is extracted. This morphological change trend includes the rate of change of the curve slope, the number of inflection points, and the oscillation period. The morphological change trend is a direct expression of the changes in hysteresis characteristics, quantifying the changes in curve shape through several key geometric features. The extraction process, based on spectral variation characteristics and combined with the geometric property analysis of the curve, systematically evaluates three key indicators: the rate of change of the curve slope reflects the change in the sensitivity of the hysteresis characteristics, calculated using the rate of change of the first derivative; the number of inflection points characterizes the change in the complexity of the hysteresis characteristics, statistically analyzed using the zero-points of the second derivative; and the oscillation period describes the periodic changes in the hysteresis characteristics, determined by the position of spectral peaks. These three indicators describe the morphological changes of the hysteresis characteristics from different perspectives, collectively constituting a comprehensive description of the change trend.

[0095] Based on the morphological change trend, a hysteresis characteristic change pattern is constructed. This pattern uses operating frequency as the independent variable and quantitative indicators of the morphological change trend as the dependent variable. These quantitative indicators include the mean rate of change of the curve slope, the increment of the number of inflection points on the curve, and the ratio of the curve's oscillation period. The hysteresis characteristic change pattern is a mathematical model that establishes a mapping relationship between operating frequency and hysteresis characteristic changes, providing a theoretical basis for subsequent influence model construction. The construction process first standardizes the three indicators of morphological change trend to eliminate the influence of dimensions. Then, using operating frequency as the independent variable, a regression model is established to fit the variation law of each indicator with frequency. The model form is dynamically selected according to the data characteristics and can be a polynomial function, exponential function, or piecewise function. The constructed hysteresis characteristic change pattern can predict hysteresis characteristic changes at any operating frequency, filling the gap in experimental sampling and realizing continuous modeling of hysteresis characteristics.

[0096] In this embodiment of the invention, the detailed implementation steps for constructing a dynamic influence model of the servo valve hysteresis characteristic change on positioning accuracy based on the deviation distribution between the hysteresis characteristic change pattern and the piston displacement of the hydraulic cylinder include:

[0097] The deviation between the actual value and the target value of the hydraulic cylinder piston displacement is recorded as the positioning deviation. Positioning deviation is a direct indicator for evaluating system performance, reflecting the positioning accuracy of the control system. The acquisition process involves using a high-precision displacement sensor to collect the actual position of the hydraulic cylinder piston in real time, comparing it with the target position set by the control system, and calculating the displacement difference. To improve measurement accuracy, a multiple sampling averaging method is used to eliminate random measurement errors. For dynamic operating conditions, the rate of change of the target value is also considered, and predictive compensation technology is used to eliminate the effect of time delay. The positioning deviation data is grouped and stored according to operating frequency, providing a basic dataset for subsequent time-frequency analysis.

[0098] Time-frequency analysis was performed on the positioning deviation to obtain its time-domain fluctuation characteristics and frequency-domain distribution characteristics at each operating frequency. Time-frequency analysis is a key method for deeply understanding the characteristics of positioning deviation, revealing its temporal and frequency characteristics through multi-dimensional analysis. The analysis process employed wavelet transform technology, which preserves time-domain information while extracting frequency-domain features, overcoming the limitations of traditional Fourier transform. The Morlet wavelet was chosen as the wavelet basis function due to its excellent time-frequency positioning capabilities, making it suitable for analyzing non-stationary signals. The analysis results consist of two parts: time-domain fluctuation characteristics describe the variation of the deviation over time, including statistical indicators such as peak value and variance; frequency-domain distribution characteristics reveal the frequency components of the deviation, reflecting the dynamic response characteristics of the system. This multi-dimensional analysis provides a comprehensive feature description for subsequent correlation assessment.

[0099] Based on the quantitative indicators of morphological change trends in the hysteresis characteristic variation pattern, the correlation coefficient between the quantitative indicators and the temporal fluctuation characteristics of the positioning deviation is calculated and denoted as the temporal influence weight. The temporal influence weight is a key parameter reflecting the statistical correlation between the quantitative hysteresis characteristics and temporal positioning accuracy. The calculation process uses the Pearson correlation coefficient method to assess the degree of linear correlation between each quantitative indicator of the morphological change trend and the temporal characteristics of the positioning deviation. The calculation formula is:

[0100] ;

[0101] in, For the first The time-domain influence weight of each quantitative indicator For the first The quantitative indicator in the first The value at each operating frequency For the first Positioning deviation time-domain characteristics at each operating frequency and These are the average values ​​of the corresponding sequences. This represents the number of sampling points for the operating frequency.

[0102] The larger the absolute value of the correlation coefficient, the more significant the impact of the indicator on positioning accuracy. The time-domain influence weight provides a key parameter for constructing the dynamic influence model, ensuring that the model accurately characterizes the time-domain characteristics.

[0103] Based on the quantitative indicators of the morphological change trend in the hysteresis characteristic variation pattern, the correlation coefficient between the quantitative indicators and the frequency domain distribution characteristics of the positioning error is calculated and denoted as the frequency domain influence weight. The frequency domain influence weight is a key parameter for quantifying the impact of hysteresis characteristics on frequency domain positioning accuracy, assessing the degree of influence of characteristic changes on the system's frequency response. The calculation method is similar to that of the time domain influence weight, using the Pearson correlation coefficient, but the analysis object is replaced by frequency domain distribution characteristics. The calculation process considers the multidimensional characteristics of the spectrum, calculating the correlation coefficient of each frequency component separately, and finally taking the weighted average as the overall evaluation. The frequency domain influence weight and the time domain influence weight together constitute a comprehensive influence assessment system, providing dual protection for the dynamic influence model.

[0104] A dynamic influence model is constructed based on time-domain and frequency-domain influence weights. This model takes the operating frequency as input and the predicted positioning error as output. The predicted value is determined by the weighted sum of the time-domain and frequency-domain influence weights. The dynamic influence model is the final result of the hysteresis characteristic influence analysis, establishing a mapping relationship between operating frequency, hysteresis characteristics, and positioning accuracy. The construction process employs a multiple regression method, comprehensively considering the influencing factors in both the time and frequency domains. The mathematical expression of the model is as follows:

[0105] ;

[0106] in, For operating frequency Predicted positioning deviation value For operating frequency Next Predicted values ​​of a quantitative indicator, and The first The time-domain and frequency-domain influence weights of each indicator The total number of quantitative indicators, and For the corresponding weighting coefficients, and .

[0107] The weighting coefficients are dynamically adjusted based on system characteristics. For low-frequency operating conditions, the time-domain weight is higher; for high-frequency operating conditions, the frequency-domain weight is higher. This adaptive weighting mechanism ensures the model's prediction accuracy across the entire frequency band. The constructed dynamic impact model can not only predict positioning deviations at known frequencies but also predict positioning accuracy at arbitrary operating frequencies through interpolation and extrapolation, providing a scientific basis for compensation strategy design.

[0108] In this embodiment of the invention, based on a dynamic influence model and combined with the pressure fluctuation value of the hydraulic system, a frequency adaptive compensation strategy for the servo valve is designed. The detailed implementation steps for obtaining the compensation control signal of the servo valve at each operating frequency include:

[0109] Based on the dynamic influence model, the predicted positioning deviation value is calculated for each operating frequency. Positioning deviation prediction is the starting point for compensation strategy design, as the model estimates the expected deviation of the system under uncompensated conditions. The prediction process directly applies the dynamic influence model, inputting the target operating frequency to calculate the corresponding predicted positioning deviation value. To improve the reliability of the prediction, a confidence interval estimation method is used to calculate the possible fluctuation range of the predicted value, providing a more comprehensive reference for subsequent compensation design. The prediction results are sorted by frequency, forming a frequency response curve of the positioning deviation, which intuitively shows the trend of positioning accuracy changes at different frequencies, providing a basis for frequency partitioning of the compensation strategy.

[0110] Based on the pressure fluctuation values ​​of the hydraulic system, the periodic and random components of the pressure fluctuation values ​​at each operating frequency are calculated and denoted as pressure fluctuation characteristics. Pressure fluctuation characteristics are an important indicator of the system's dynamic performance, reflecting the stability of the hydraulic system and the impact of external disturbances. The calculation process first performs bandpass filtering on the pressure fluctuation data to extract signal components related to the operating frequency; then, through autocorrelation analysis and power spectral density estimation, the signal is decomposed into periodic and random components. The periodic components mainly originate from the system's own operating rhythm and harmonics, reflecting the system's internal dynamic characteristics; the random components originate from external disturbances and system noise, exhibiting irregular fluctuations. This decomposition method reveals the essential characteristics of pressure fluctuations, providing fundamental data for subsequent gain effect analysis.

[0111] Based on the characteristics of pressure fluctuations, the gain effect of pressure fluctuation values ​​on positioning deviation is extracted and denoted as the pressure gain factor. The pressure gain factor is a key parameter for quantifying the impact of pressure fluctuations on positioning accuracy, describing the amplification or suppression effect of deviation caused by pressure changes. The extraction process employs a system identification method to establish a transfer function model between pressure fluctuations and positioning deviation. The calculation formula is as follows:

[0112] ;

[0113] in, For operating frequency Pressure gain factor under, This represents the cross-power spectral density of pressure fluctuations and positioning deviations. This represents the self-power spectral density of the pressure fluctuation.

[0114] A gain factor greater than 1 indicates that pressure fluctuations amplify positioning errors, while a gain factor less than 1 indicates that they suppress errors. This indicator provides an important reference for the design of compensation strategies, especially under high-frequency operating conditions where the impact of pressure fluctuations is often more significant.

[0115] Based on the predicted positioning deviation and the pressure gain factor, a frequency adaptive compensation strategy is designed. This strategy includes: when the operating frequency is below a preset frequency threshold, a feedforward compensation signal is designed based on the predicted positioning deviation, with the amplitude of the feedforward compensation signal proportional to the predicted positioning deviation; when the operating frequency is above the preset frequency threshold, a feedback compensation signal is designed based on the pressure gain factor, with the phase of the feedback compensation signal out of phase with the periodic component of the pressure fluctuation value. The frequency adaptive compensation strategy is a core technology for improving positioning accuracy, employing differentiated compensation methods according to the system characteristics in different frequency ranges.

[0116] In the low-frequency region (operating frequency below a preset frequency threshold), the system exhibits quasi-static characteristics, with hysteresis dominating. A feedforward compensation method is employed. The design formula for the feedforward compensation signal is:

[0117] ;

[0118] in, This is a feedforward compensation signal, typically measured in milliamperes (mA), which is directly added to the control current of the servo valve. The feedforward compensation gain coefficient is dimensionless and used to adjust the compensation strength. Its value is usually determined experimentally or by system identification methods, and it typically ranges from 0.8 to 1.2. This is the predicted value of the positioning deviation.

[0119] Feedforward compensation directly addresses the expected deviation, with the amplitude of the compensation signal proportional to the predicted deviation value but in the opposite direction, thus achieving pre-emptive cancellation of the deviation. This method is highly effective in the low-frequency region, effectively compensating for static deviations caused by hysteresis.

[0120] In the high-frequency region (operating frequency greater than a preset frequency threshold), the system exhibits dynamic characteristics, with pressure fluctuations having a significant impact. A feedback compensation method is employed. The design formula for the feedback compensation signal is as follows:

[0121] ;

[0122] in, To provide feedback compensation signals, For frequency-dependent feedback compensation gain, It is the periodic component of pressure fluctuation.

[0123] Feedback compensation, based on real-time pressure measurement, generates a compensation signal that is inversely phase to the pressure fluctuation, reducing the impact of pressure fluctuations through the phase cancellation principle. Feedback gain. With pressure gain factor Related, usually set to ,in This is the adjustment coefficient. This design ensures the consistency of compensation effect at different frequencies and effectively suppresses high-frequency dynamic interference.

[0124] The feedforward compensation signal and the feedback compensation signal are weighted and fused to obtain the compensation control signal. The weights for the fusion are determined by the normalized value of the operating frequency. Weighted fusion is a key technology for achieving smooth transition across the entire frequency band, avoiding abrupt changes in the compensation strategy at critical frequencies. The fusion formula is:

[0125] ;

[0126] in, For the final compensation control signal, The weighting coefficients are frequency-dependent and satisfy... Weighting coefficients Determined by the normalized value of the operating frequency:

[0127] when hour, ;

[0128] when hour, ;

[0129] when hour, ;

[0130] in, and These are the low-frequency and high-frequency thresholds, typically... , , This is a preset frequency threshold.

[0131] This smooth transition mechanism ensures the continuity and stability of the compensation strategy across the entire frequency band, avoiding system oscillations that may be caused by traditional segmented compensation methods at critical frequencies.

[0132] In this embodiment of the invention, the detailed implementation steps for real-time correction of the input current signal based on the compensation control signal to obtain the corrected servo valve control signal include:

[0133] The compensation control signal is smoothed in the time domain to obtain a smoothed compensation control signal. The time-domain smoothing process employs an adaptive window length based on the operating frequency, with the adaptive window length inversely proportional to the operating frequency. Time-domain smoothing is a crucial step in ensuring the stability of the compensation signal, improving the reliability of the compensation effect by eliminating high-frequency noise and abrupt changes. The processing uses a moving average filtering method to perform a locally weighted average on the original compensation signal, resulting in a smooth compensation curve. The key technology lies in the adaptive adjustment of the window length, which is inversely proportional to the operating frequency. The calculation formula is as follows:

[0134] ;

[0135] in, For operating frequency The length of the window below, As the baseline window length, This is the reference frequency.

[0136] The principle behind this adaptive design is as follows: under low-frequency conditions, the system response is slower, so a longer smoothing window can be used to obtain a more stable compensation signal; under high-frequency conditions, the system response is faster, requiring a shorter window to preserve the dynamic response capability of the compensation signal. Adaptive smoothing ensures both the stability of the compensation signal and the preservation of necessary dynamic characteristics, providing high-quality compensation input for signal superposition.

[0137] The smoothed compensation control signal is superimposed on the input current signal to obtain the superimposed current signal. Signal superposition is the direct operation to achieve the compensation effect, integrating the designed compensation signal into the original control signal to form the corrected comprehensive control quantity. The superposition process uses a simple linear addition method, and the calculation formula is as follows:

[0138] ;

[0139] in, The superimposed current signal The original input current signal, This is the smoothed compensation control signal.

[0140] The superposition process takes into account the timing alignment of the signals. Through interpolation and synchronization, it ensures a precise time match between the compensation signal and the original signal, avoiding compensation failure caused by phase shift. The superimposed current signal integrates the original control intent and the compensation correction effect, providing input for subsequent limiting processing.

[0141] The superimposed current signal is amplitude-limited to obtain the corrected servo valve control signal. The upper and lower limits of the amplitude limiting are determined by the rated operating range of the servo valve. Amplitude limiting is a safeguard to ensure the usability of the control signal and prevent system saturation or damage caused by over-compensation. The process employs a soft-limiting algorithm, gradually compressing the gain as it approaches the limit boundary to avoid discontinuities caused by hard truncation. The limiting formula is:

[0142] when hour, ;

[0143] when hour, ;

[0144] when hour, ;

[0145] in, This is the corrected servo valve control signal. and These are the maximum and minimum operating currents of the servo valve, respectively.

[0146] The amplitude limiting parameters are determined based on the servo valve's technical specifications, and a certain safety margin is typically set to prevent prolonged operation at extreme limits. Amplitude limiting ensures that the control signal remains within the servo valve's effective operating range, guaranteeing system safety and reliability, and is a necessary step in practical compensation methods.

[0147] In this embodiment of the invention, the step of curve fitting for the scatter distribution includes:

[0148] Outlier removal is performed on the scatter plot distribution. The criterion for outlier removal is that the difference between the local regression value and the mean of the scatter plot distribution is greater than a preset outlier threshold. Outlier removal is a preprocessing step to improve fitting accuracy. By identifying and removing data points that do not conform to the overall distribution pattern, it improves the representativeness and consistency of the dataset. The removal process first calculates the mean and standard deviation of all local regression values, and then sets the judgment criteria based on statistical principles. The judgment formula is:

[0149] ;

[0150] in, For the first Local lag values ​​for each time segment, The mean of all return values. Let the standard deviation of all return values ​​be . The threshold coefficient (usually 2.5-3) is used to identify outliers when the above conditions are met.

[0151] Outliers can arise from various causes, such as momentary sensor malfunctions, external interference, or transient system responses. These data points do not represent the true characteristics of the system, and removing them can significantly improve the accuracy and reliability of the fitted model. Outlier removal preserves the main features of the dataset, providing high-quality input data for subsequent piecewise fitting.

[0152] The scatter distribution after outlier removal is piecewise fitted, with the segmentation points determined by the inflection points of the rate of change of the input current signal amplitude. Piecewise fitting is an effective method for handling nonlinear data, improving overall fitting accuracy by decomposing complex relationships into multiple simpler ones. The fitting process first requires determining suitable segmentation points; this method innovatively uses the inflection points of the rate of change of the input current signal amplitude as the segmentation basis. The inflection point identification algorithm is based on curvature calculation, finding locations on the rate of change curve where curvature abruptly changes. These locations typically represent transition points in system characteristics and are ideal segmentation boundaries. After determining the segmentation points, the entire dataset is divided into multiple sub-intervals. The data distribution within each interval is relatively simple, suitable for fitting with low-order models, avoiding both underfitting and overfitting risks.

[0153] Within each segment, a spline interpolation-based curve fitting method is used to obtain piecewise fitted curves. Spline interpolation is a key technique for achieving high-precision fitting, and it achieves smoothness and accurate representation of the curve through piecewise polynomial functions. This method selects cubic spline interpolation, with each segment represented by a third-order polynomial, which has the characteristic of continuous second derivatives, ensuring the smoothness of the fitted curve. The fitting process adopts the least squares principle to solve for the coefficients of the spline function, minimizing the sum of squared errors between the fitted curve and the actual data points. This spline-based fitting method has the flexibility of polynomial fitting while avoiding the Runge phenomenon of higher-order polynomials, making it particularly suitable for complex nonlinear engineering problems such as hysteresis characteristics. The fitting result for each segment is a locally optimal curve, accurately expressing the variation law of hysteresis characteristics within that interval.

[0154] All piecewise fitted curves are stitched together to obtain the dynamic hysteresis characteristic curve. The smoothness of the stitching is constrained by the consistency of the endpoint slopes of adjacent piecewise fitted curves. Curve stitching is the final step in constructing the complete characteristic curve. By reasonably connecting the various piecewise curves, a continuous model covering the entire range is formed. The key to the stitching process is to ensure a smooth transition at the connection points and avoid abrupt changes and discontinuities. This method uses endpoint slope consistency constraints, requiring that adjacent curve segments not only have equal function values ​​at the connection points, but also equal first derivatives (slopes), ensuring the first-order continuity of the curves. This is achieved by introducing boundary conditions during the fitting stage, adding the endpoint slopes of adjacent segments as constraints to the optimization objective, and adjusting the spline function parameters to achieve smooth stitching. The resulting dynamic hysteresis characteristic curve is a continuous, smooth, and accurate mathematical model that comprehensively describes the hysteresis characteristics of the servo valve at the current operating frequency, laying the foundation for subsequent extraction of hysteresis characteristic change patterns.

[0155] In this embodiment of the invention, the method for obtaining time-domain fluctuation characteristics includes:

[0156] Wavelet decomposition is performed on the positioning deviation to obtain its wavelet coefficients at different time scales. Wavelet decomposition is a powerful tool for multi-resolution analysis, capable of simultaneously capturing the time and frequency characteristics of a signal, and is particularly suitable for analyzing non-stationary signals. The decomposition process uses Discrete Wavelet Transform (DWT) to decompose the positioning deviation signal into wavelet coefficients at multiple time scales. The mathematical expression is as follows:

[0157] ;

[0158] in, For the first Scale No. Wavelet coefficients of position, This is a positioning deviation signal. The scale parameter is Translation parameters are wavelet basis functions, This indicates the total number of sampling points or the signal length of the positioning deviation signal. The sampling point index of the time-domain signal is a number from 0 to... Integers.

[0159] The wavelet basis function is chosen to be the Daubechies wavelet (usually db4 or db6), whose good orthogonality and tight support properties are suitable for engineering signal processing. The number of decomposition levels is determined according to the signal characteristics, usually 3-5 levels, covering the complete spectrum from high-frequency details to low-frequency approximations. The wavelet decomposition result is an expression of the positioning deviation signal at multiple time scales, providing a rich source of information for wave feature extraction.

[0160] The energy proportion of wavelet coefficients at each time scale is calculated and denoted as the time-domain fluctuation energy distribution. The energy proportion is a key indicator for quantifying the importance of each scale, reflecting the energy distribution of positioning errors across different frequency bands. The calculation process first obtains the energy of the wavelet coefficients at each scale:

[0161] ;

[0162] Then calculate the energy percentage:

[0163] ;

[0164] in, For the first Energy at scale Its energy percentage.

[0165] The energy distribution reveals the spectral structure of the positioning deviation, with high-energy-proportion scales corresponding to the system's main vibration modes or disturbance sources. This analysis provides a focus for subsequent feature extraction, indicating the time scales that require particular attention.

[0166] Based on the temporal fluctuation energy distribution, the temporal fluctuation characteristics of positioning deviation are extracted. These characteristics include the peak time position of wavelet coefficients, the variance of the energy proportion of wavelet coefficients, and the periodic component of wavelet coefficients. The temporal fluctuation characteristics are a quantitative expression of the temporal properties of positioning deviation, describing the temporal pattern of the deviation through several key indicators. The extraction process is based on wavelet decomposition results, focusing on three aspects: the peak time position reflects the instantaneous characteristics of the deviation, determined by detecting the local maximum value of the wavelet coefficients and its corresponding time point; the variance of the energy proportion characterizes the concentration of energy distribution; a larger variance indicates a more uneven energy distribution and a more unstable system response; and the periodic component describes the repetitive pattern of the deviation, determined through autocorrelation analysis of the wavelet coefficients. These three indicators characterize the temporal properties of positioning deviation from different perspectives, providing a quantitative basis for assessing the impact of hysteresis characteristics. The temporal fluctuation characteristics, together with the frequency domain distribution characteristics, constitute a comprehensive feature description of positioning deviation, providing a reliable data foundation for constructing dynamic impact models.

[0167] In this embodiment of the invention, the method for obtaining pressure fluctuation characteristics includes:

[0168] Fourier transforms are performed on the pressure fluctuation values ​​to obtain their spectral components. Fourier transform is a fundamental tool in spectrum analysis, converting time-domain signals into frequency-domain representations and revealing the signal's frequency composition. The transform process employs the Fast Fourier Transform (FFT) algorithm to improve computational efficiency. To enhance the accuracy of spectrum analysis, window function techniques (typically the Hanning window) are used to reduce spectral leakage; simultaneously, a zero-filling method is employed to improve frequency resolution and facilitate accurate identification of spectral peaks. The spectrum analysis results include amplitude and phase spectra, comprehensively describing the frequency characteristics of the pressure fluctuations and providing a basis for extracting periodic components.

[0169] Based on the spectral components, the periodic components of the pressure fluctuation values ​​are extracted. The frequency range of these periodic components is determined by the harmonic components of the operating frequency. The periodic components represent the regularity of the pressure fluctuations and are typically related to the system's operating rhythm and mechanical characteristics. The extraction process is based on spectral analysis results, focusing on spectral peaks related to the operating frequency. The frequency range is determined to be the fundamental frequency and harmonic components of the operating frequency, typically including 1-5 harmonics, with the specific range dynamically adjusted according to system characteristics. The extraction method employs bandpass filtering technology, designing multiple bandpass filters to extract each harmonic component separately, which are then superimposed to form the complete periodic component. The filter design uses a finite impulse response (FIR) structure to ensure linear phase characteristics and avoid waveform distortion caused by phase distortion. The extracted periodic components reflect the main regularity of the pressure fluctuations, providing crucial information for subsequent analysis.

[0170] Subtracting the periodic component from the pressure fluctuation value yields the random component. The random component represents the irregular portion of the pressure fluctuation, typically originating from system noise, external disturbances, and random perturbations. The acquisition process employs a simple time-domain subtraction, subtracting the extracted periodic component from the original pressure fluctuation value to obtain the remaining random portion. To improve the reliability of the random component analysis, statistical tests are used to verify its randomness and ensure effective separation. The characteristics of the random component are usually described by statistical parameters such as variance, skewness, and kurtosis, which reflect the intensity and distribution characteristics of the random fluctuations. The separation of the periodic and random components provides a foundation for an accurate description of the pressure fluctuation characteristics and helps identify the influence of fluctuations from different sources.

[0171] The amplitude of the periodic component and the variance of the random component are calculated and denoted as the pressure fluctuation characteristic. The pressure fluctuation characteristic is a quantitative expression of the fluctuation properties, describing the main characteristics of the fluctuation through two key parameters. In the calculation, the root mean square (RMS) value is used for the amplitude of the periodic component, comprehensively reflecting the intensity of the periodic fluctuation; the variance of the random component is directly calculated, characterizing the dispersion of random fluctuations. These two parameters together constitute the pressure fluctuation characteristic, providing a quantitative basis for subsequent gain effect analysis. A large amplitude of the periodic component indicates that the system has significant periodic disturbances, which may require targeted suppression measures; a large variance of the random component indicates that the system is subject to significant random disturbances, which may require enhanced robustness design. The accurate calculation of the pressure fluctuation characteristic provides reliable data support for the design of frequency adaptive compensation strategies.

[0172] The above describes a high-precision positioning control method for a servo electro-hydraulic actuator according to an embodiment of this application. The following describes a high-precision positioning control system for a servo electro-hydraulic actuator according to an embodiment of this application. Please refer to [link to relevant documentation]. Figure 2 One embodiment of a high-precision positioning control system for a servo electro-hydraulic actuator in this application includes:

[0173] The parameter acquisition module is used to acquire the real-time operating parameters of the servo electro-hydraulic actuator at different operating frequencies. The real-time operating parameters include the valve core displacement of the servo valve, the input current signal of the servo valve, the piston displacement of the hydraulic cylinder, and the pressure fluctuation value of the hydraulic system.

[0174] The dynamic hysteresis modeling module is used to construct the dynamic hysteresis characteristic curve of the servo valve at each operating frequency based on the time-series correspondence between the valve core displacement and the input current signal in the real-time operating parameters.

[0175] The hysteresis change analysis module is used to extract the hysteresis characteristic change pattern of the servo valve at different operating frequencies based on the curve shape differences in adjacent operating frequency ranges in the dynamic hysteresis characteristic curve.

[0176] The influence model building module is used to construct a dynamic influence model of the servo valve hysteresis characteristic change on positioning accuracy based on the deviation distribution of the hysteresis characteristic change pattern and the piston displacement of the hydraulic cylinder.

[0177] The compensation strategy design module is used to design a frequency adaptive compensation strategy for the servo valve based on the dynamic influence model and the pressure fluctuation value of the hydraulic system, and to obtain the compensation control signal of the servo valve at each working frequency.

[0178] The control signal correction module is used to correct the input current signal in real time according to the compensation control signal, and obtain the corrected servo valve control signal.

[0179] The high-precision positioning and execution module is used to drive the servo electro-hydraulic actuator based on the corrected servo valve control signal, control the piston displacement of the hydraulic cylinder, and achieve high-precision positioning.

[0180] The modules are connected via wired and / or wireless means to enable data transmission between them.

[0181] This invention achieves high-precision positioning control of a servo electro-hydraulic actuator through servo valve hysteresis characteristic analysis, positioning deviation prediction, frequency adaptive compensation strategy design, and real-time control signal correction. The adaptive optimization feature of this invention can adjust control parameters in real time according to the system operating frequency and servo valve characteristics, significantly improving positioning accuracy and reducing system oscillation.

[0182] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0183] It should be noted that all formulas in this manual are calculated by removing dimensions and taking their numerical values. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0184] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A high-precision positioning control method for a servo electro-hydraulic actuator, characterized in that, include: The real-time operating parameters of the servo electro-hydraulic actuator at different operating frequencies are obtained. The real-time operating parameters include the valve core displacement of the servo valve, the input current signal of the servo valve, the piston displacement of the hydraulic cylinder, and the pressure fluctuation value of the hydraulic system. Based on the time-series correspondence between the valve core displacement of the servo valve and the input current signal in the real-time operating parameters, a dynamic hysteresis characteristic curve of the servo valve at each operating frequency is constructed. Based on the differences in curve shape within adjacent operating frequency ranges in the dynamic hysteresis characteristic curve, the hysteresis characteristic variation pattern of the servo valve at different operating frequencies is extracted. Based on the deviation distribution between the hysteresis characteristic change pattern and the piston displacement of the hydraulic cylinder, a dynamic influence model of the servo valve hysteresis characteristic change on the positioning accuracy is constructed. Based on the dynamic influence model and the pressure fluctuation value of the hydraulic system, a frequency adaptive compensation strategy for the servo valve is designed to obtain the compensation control signal of the servo valve at each working frequency. The input current signal is corrected in real time according to the compensation control signal to obtain the corrected servo valve control signal; The servo electro-hydraulic actuator is driven based on the modified servo valve control signal to control the piston displacement of the hydraulic cylinder, thereby achieving high-precision positioning.

2. The high-precision positioning control method for a servo electro-hydraulic actuator according to claim 1, characterized in that, The step of constructing the dynamic hysteresis characteristic curve of the servo valve at each operating frequency includes: The valve core displacement of the servo valve in the real-time operating parameters is time-aligned with the input current signal to obtain time data pairs of valve core displacement and input current signal at each operating frequency; The timing data pairs are segmented, and the timing data pairs at each operating frequency are divided into multiple time segments, each time segment corresponding to one action cycle of the servo valve. Within each time segment, the displacement difference between the positive and negative change ranges of the valve core displacement is calculated and recorded as the local hysteresis value for each time segment. Based on the local hysteresis value of each time segment and the amplitude change rate of the input current signal, a scatter distribution of the local hysteresis value as a function of the amplitude change rate is constructed. Curve fitting is performed on the scattered point distribution to obtain the dynamic hysteresis characteristic curve of the servo valve at each operating frequency, wherein the horizontal axis of the dynamic hysteresis characteristic curve is the amplitude change rate of the input current signal, and the vertical axis is the local hysteresis value.

3. The high-precision positioning control method for a servo electro-hydraulic actuator according to claim 1, characterized in that, The step of extracting the hysteresis characteristic variation pattern of the servo valve at different operating frequencies includes: The dynamic hysteresis characteristic curve is decomposed in the frequency domain to obtain the spectral components of the dynamic hysteresis characteristic curve at each operating frequency. Calculate the amplitude and phase differences of the spectral components of the dynamic hysteresis characteristic curve within adjacent operating frequency intervals, and denote them as spectral variation characteristics; Based on the spectral variation characteristics, the morphological change trend of the dynamic hysteresis characteristic curve in adjacent operating frequency intervals is extracted. The morphological change trend includes the curve slope change rate, the number of curve inflection points, and the curve oscillation period. Based on the morphological change trend, the retracement characteristic change pattern is constructed, wherein the retracement characteristic change pattern uses the operating frequency as the independent variable and the quantitative index of the morphological change trend as the dependent variable. The quantitative index includes the mean of the curve slope change rate, the increment of the number of curve inflection points, and the ratio of the curve oscillation period.

4. The high-precision positioning control method for a servo electro-hydraulic actuator according to claim 3, characterized in that, The steps for constructing a dynamic impact model of the servo valve hysteresis characteristic variation on positioning accuracy include: The deviation between the actual value and the target value of the piston displacement of the hydraulic cylinder is obtained and recorded as the positioning deviation; Time-frequency analysis is performed on the positioning deviation to obtain the time-domain fluctuation characteristics and frequency-domain distribution characteristics of the positioning deviation at each operating frequency; Based on the quantitative index of the morphological change trend in the hysteresis characteristic change pattern, calculate the correlation coefficient between the quantitative index and the temporal fluctuation characteristics of the positioning deviation, and denote it as the temporal influence weight. Based on the quantitative index of the morphological change trend in the hysteresis characteristic change mode, calculate the correlation coefficient between the quantitative index and the frequency domain distribution characteristics of the positioning deviation, and denot it as the frequency domain influence weight. Based on the time-domain influence weight and the frequency-domain influence weight, the dynamic influence model is constructed, wherein the dynamic influence model takes the operating frequency as input and the predicted value of the positioning deviation as output.

5. A high-precision positioning control method for a servo electro-hydraulic actuator according to claim 4, characterized in that, The steps for designing the frequency adaptive compensation strategy for the servo valve include: Based on the dynamic influence model, predict the value of the positioning deviation at each operating frequency; Based on the pressure fluctuation value of the hydraulic system, calculate the periodic and random components of the pressure fluctuation value at each working frequency, and denot them as pressure fluctuation characteristics; Based on the pressure fluctuation characteristics, the gain effect of the pressure fluctuation value on the positioning deviation is extracted and denoted as the pressure gain factor. Based on the predicted value of the positioning deviation and the pressure gain factor, the frequency adaptive compensation strategy is designed, wherein the frequency adaptive compensation strategy includes: When the operating frequency is lower than a preset frequency threshold, a feedforward compensation signal is designed based on the predicted value of the positioning deviation, and the amplitude of the feedforward compensation signal is proportional to the predicted value of the positioning deviation. When the operating frequency is greater than a preset frequency threshold, a feedback compensation signal is designed based on the pressure gain factor, and the phase of the feedback compensation signal is out of phase with the periodic component of the pressure fluctuation value. The feedforward compensation signal and the feedback compensation signal are weighted and fused to obtain the compensation control signal.

6. The high-precision positioning control method for a servo electro-hydraulic actuator according to claim 1, characterized in that, The step of real-time correction of the input current signal based on the compensation control signal includes: The compensation control signal is smoothed in the time domain to obtain the smoothed compensation control signal; The smoothed compensation control signal is superimposed on the input current signal to obtain the superimposed current signal; The amplitude of the superimposed current signal is limited to obtain the corrected servo valve control signal.

7. A high-precision positioning control method for a servo electro-hydraulic actuator according to claim 2, characterized in that, The step of curve fitting the scatter distribution includes: Outlier removal is performed on the scatter distribution. The criteria for determining outliers are: the difference between the local hysteresis value and the mean of the scatter distribution is greater than a preset outlier threshold. The distribution of scattered points after outliers is fitted piecewise, wherein the segment points of the piecewise fitting are determined by the inflection points of the amplitude change rate of the input current signal. Within each segment, a spline interpolation-based curve fitting method is used to obtain the segmented fitting curve; All the segmented fitted curves are spliced ​​together to obtain the dynamic hysteresis characteristic curve.

8. A high-precision positioning control method for a servo electro-hydraulic actuator according to claim 4, characterized in that, The method for obtaining the time-domain fluctuation characteristics includes: The positioning deviation is decomposed by wavelet to obtain the wavelet coefficients of the positioning deviation at different time scales; Calculate the energy proportion of the wavelet coefficients at each time scale, and denote it as the time-domain fluctuation energy distribution; Based on the time-domain fluctuation energy distribution, the time-domain fluctuation characteristics of the positioning deviation are extracted, wherein the time-domain fluctuation characteristics include the peak time position of the wavelet coefficients, the variance of the energy proportion of the wavelet coefficients, and the periodic components of the wavelet coefficients.

9. A high-precision positioning control method for a servo electro-hydraulic actuator according to claim 5, characterized in that, The method for obtaining the pressure fluctuation characteristics includes: Perform a Fourier transform on the pressure fluctuation value to obtain the spectral components of the pressure fluctuation value; Based on the spectral components, the periodic components of the pressure fluctuation value are extracted, wherein the frequency range of the periodic components is determined by the harmonic components of the operating frequency. Subtract the periodic component from the pressure fluctuation value to obtain the random component of the pressure fluctuation value; The amplitude of the periodic component and the variance of the random component are calculated and denoted as the pressure fluctuation characteristic.

10. A high-precision positioning control system for a servo electro-hydraulic actuator, used to implement the high-precision positioning control method for a servo electro-hydraulic actuator as described in any one of claims 1 to 9, characterized in that, include: The parameter acquisition module is used to acquire the real-time operating parameters of the servo electro-hydraulic actuator at different operating frequencies. The real-time operating parameters include the valve core displacement of the servo valve, the input current signal of the servo valve, the piston displacement of the hydraulic cylinder, and the pressure fluctuation value of the hydraulic system. The dynamic hysteresis modeling module is used to construct the dynamic hysteresis characteristic curve of the servo valve at each operating frequency based on the time-series correspondence between the valve core displacement of the servo valve and the input current signal in the real-time operating parameters. The hysteresis variation analysis module is used to extract the hysteresis characteristic variation pattern of the servo valve at different operating frequencies based on the curve shape differences in adjacent operating frequency ranges of the dynamic hysteresis characteristic curve. The influence model construction module is used to construct a dynamic influence model of the servo valve hysteresis characteristic change on the positioning accuracy based on the deviation distribution between the hysteresis characteristic change pattern and the piston displacement of the hydraulic cylinder. The compensation strategy design module is used to design a frequency adaptive compensation strategy for the servo valve based on the dynamic influence model and the pressure fluctuation value of the hydraulic system, and to obtain the compensation control signal of the servo valve at each working frequency. The control signal correction module is used to correct the input current signal in real time according to the compensation control signal to obtain the corrected servo valve control signal; A high-precision positioning execution module is used to drive the servo electro-hydraulic actuator based on the corrected servo valve control signal, control the piston displacement of the hydraulic cylinder, and achieve high-precision positioning.

Images