Dynamic compression-shear testing machine damping self-adaptive control method and system

Abstract

The application relates to the technical field of dynamic compression shear test control, and discloses a dynamic compression shear testing machine damping self-adaptive control method and system. The method first identifies a current working condition steady-state damping parameter set online, predicts transient damping requirements of working condition switching, generates a transient damping compensation feedforward control sequence, injects the servo valve control loop, and then based on the transient damping residual error, real-time correction of the damping feedback gain parameter is completed to realize closed-loop self-adaptive regulation and control, effectively reduces transient loading error of working condition switching, shortens the transient process time, and improves the efficiency and test precision of multi-working condition continuous dynamic compression shear test.

Description

Technical Field

[0001] This invention relates to the field of dynamic compression-shear test control technology, and in particular to a damping adaptive control method and system for a dynamic compression-shear test machine. Background Technology

[0002] Dynamic compression-shear testing machines are widely used for testing the mechanical properties of seismic isolation and damping components such as seismic isolation bearings. They can perform continuous dynamic compression-shear tests under multiple working conditions with adjustable loading frequency, loading amplitude, and loading direction. During the test, an adaptive damping control algorithm is used to match the system damping, ensuring the tracking accuracy of loading force and displacement. This allows for the acquisition of mechanical performance parameters of the components under different working conditions, providing data support for the performance verification and engineering application of seismic isolation and damping components.

[0003] During multi-condition continuous dynamic compression-shear tests of seismic isolation bearings, after completing the condition switching operation, test engineers observed that the loading force and displacement tracking errors exceeded the allowable range. Subsequent steady-state loading required a considerable period of time to recover to the allowable accuracy range. The root cause of this problem lies in the fact that the damping matching logic of the existing damping adaptive control algorithm is based solely on the inherent system damping and test specimen damping parameters identified during the steady-state loading phase. It does not incorporate the transient fluctuation characteristics of the oil pressure in the hydraulic system cavity pipeline at the moment of condition switching, nor the nonlinear abrupt change parameters of the test specimen's damping under sudden load changes. Consequently, it cannot predict the damping requirements during the transient process. Furthermore, the existing damping adjustment closed loop uses only the loading error at the current moment as the sole feedback input, without associating it with preset subsequent condition loading parameters. The damping adjustment response lags behind the change in the total system damping characteristics caused by the condition switching. This problem can lead to unexpected damage accumulation inside the test piece, resulting in hidden systematic biases in the mechanical performance test results under subsequent steady-state conditions. Test personnel cannot identify these biases through conventional data verification, which can easily lead to discrepancies between the component performance verification results and actual service performance. At the same time, the fixed waiting time setting will also reduce the efficiency of test execution. Summary of the Invention

[0004] To address the above problems, the present invention provides the following solution: An adaptive damping control method for a dynamic compression-shear testing machine: S1: During the steady-state loading phase of the multi-condition continuous dynamic compression-shear test performed by the dynamic compression-shear testing machine, the hydraulic system pressure time-series data, actuator displacement time-series data, and loading force time-series data are collected synchronously. The hydraulic system pressure time-series data, actuator displacement time-series data, and loading force time-series data are recursively identified online to obtain the current working condition steady-state damping parameter set. Based on the current working condition steady-state damping parameter set and the preset subsequent working condition loading parameter table, the transient damping demand prediction vector for working condition switching is calculated. Further, step S1 includes: S11: Deploy pressure sensors on the oil supply lines of the vertical loading cylinder and the horizontal shear actuator of the dynamic compression-shear testing machine; deploy displacement sensors on the piston rod ends of the vertical loading cylinder and the horizontal shear actuator; deploy force sensors on the load ends of the vertical loading cylinder and the horizontal shear actuator; and synchronously collect hydraulic system pressure timing data, actuator displacement timing data, and loading force timing data at a preset sampling period during the steady-state loading phase. S12: Perform online recursive identification of hydraulic system pressure timing data, actuator displacement timing data, and loading force timing data to obtain the current working condition steady-state damping parameter set; Further, step S12 includes: S121: The actuator displacement timing data is used as the system excitation input, and the loading force timing data is used as the system response output. The recursive least squares method is used to identify the force-displacement hysteresis relationship of the test specimen online, and the equivalent stiffness and equivalent damping ratio of the test specimen are obtained. S122: Based on the pressure difference time series data of the two chambers of the vertical loading cylinder and the pressure difference time series data of the two chambers of the horizontal shear actuator in the hydraulic system pressure time series data, and combined with the velocity components in the corresponding directions in the actuator displacement time series data, calculate the hydraulic system vertical channel oil damping coefficient and hydraulic system horizontal channel oil damping coefficient. S123: Combine the equivalent stiffness of the test specimen, the equivalent damping ratio of the test specimen, the hydraulic damping coefficient of the vertical channel of the hydraulic system, and the hydraulic damping coefficient of the horizontal channel of the hydraulic system into a set of steady-state damping parameters for the current working condition; S13: Based on the current steady-state damping parameter set and the preset subsequent loading parameter table, calculate the transient damping demand prediction vector for the change of operating conditions; Further, step S13 includes: S131: Read the target loading frequency, target loading amplitude, and target loading direction of the subsequent working condition from the preset subsequent working condition loading parameter table; S132: Based on the equivalent stiffness and equivalent damping ratio of the test specimen in the current steady-state damping parameter set, and combined with the target loading frequency and target loading amplitude of the subsequent working conditions, calculate the load mutation amount, and look up the pre-calibrated test specimen damping nonlinear mutation characteristic table according to the load mutation amount to obtain the estimated increment of transient damping of the test specimen. S133: Based on the hydraulic system's vertical channel oil damping coefficient and horizontal channel oil damping coefficient in the current steady-state damping parameter set, combined with the target loading frequency and target loading amplitude of the subsequent working conditions, and based on the oil compressibility modulus and pipeline cavity volume parameters, calculate the estimated value of the transient fluctuation amplitude of the oil pressure at the moment of working condition switching, and convert the estimated value of the transient additional damping of the hydraulic system into the estimated value of the transient fluctuation amplitude of the oil pressure. S134: Combine the estimated incremental transient damping of the test specimen, the estimated transient additional damping of the hydraulic system, and the target loading frequency, target loading amplitude, and target loading direction of the subsequent working conditions into a predictive vector for transient damping demand during working condition switching. S2: Based on the transient damping demand prediction vector during the working condition switching, a transient damping compensation feedforward control sequence is generated; at the time of working condition switching execution, the transient damping compensation feedforward control sequence is injected into the servo valve control loop, and the transient loading force time series data and transient displacement time series data during the working condition switching process are collected simultaneously. The transient loading force time series data and transient displacement time series data are compared and calculated with the transient damping compensation feedforward control sequence to obtain the transient damping residual time series data. Further, step S2 includes: S21: Generate a transient damping compensation feedforward control sequence based on the transient damping demand prediction vector during operating condition switching; Further, step S21 includes: S211: Extract the estimated increment of transient damping of the test specimen and the estimated amount of transient additional damping of the hydraulic system from the transient damping demand prediction vector of the working condition switching, and sum the estimated increment of transient damping of the test specimen and the estimated amount of transient additional damping of the hydraulic system to obtain the estimated amount of transient additional damping. S212: Determine the transient compensation time window length based on the estimated total transient additional damping and the target loading frequency in the transient damping demand prediction vector for operating condition switching; specifically, the transient compensation time window length is a preset multiple of the period corresponding to the target loading frequency; S213: Within the transient compensation time window, with the servo valve control cycle as the time step, the estimated total transient additional damping is discretized and expanded in the time domain according to the exponential decay envelope to generate the transient damping compensation amount corresponding to each time step; the decay time constant of the exponential decay envelope is determined based on the larger value of the hydraulic system vertical channel oil damping coefficient and the hydraulic system horizontal channel oil damping coefficient. S214: Convert the transient damping compensation amount corresponding to each time step into the corresponding servo valve opening correction amount, and arrange them in time order to form a transient damping compensation feedforward control sequence. S22: At the moment of operation condition switching, the transient damping compensation feedforward control sequence is sequentially superimposed into the command signal of the servo valve control circuit according to the time step; the transient loading force timing data and transient displacement timing data of the vertical loading cylinder and the horizontal shear actuator are collected simultaneously during the operation condition switching process. S23: Compare and calculate the transient loading force time series data and transient displacement time series data with the transient damping compensation feedforward control sequence to obtain the transient damping residual time series data; Further, step S23 includes: S231: Based on the servo valve opening correction amount at each time step in the transient damping compensation feedforward control sequence, and combined with the servo valve flow gain coefficient, calculate the expected compensation force increment sequence at each time step. S232: Calculate the actual load force change at each time step based on the transient load force time series data, and subtract the expected compensation force increment and the target working condition command force increment at the corresponding time step of the transient damping compensation feedforward control sequence from the actual load force change to obtain the force residual value at each time step. S233: Calculate the actual displacement change at each time step based on the transient displacement time series data, and subtract the target working condition command displacement increment from the actual displacement change to obtain the displacement residual value at each time step. S234: Arrange the force residuals and displacement residuals at each time step in chronological order to form transient damping residual time series data; S3: Perform sliding window segmentation and feature extraction on the transient damping residual time series data to obtain the transient residual feature sequence; perform correlation calculation between the transient residual feature sequence and the current operating condition steady-state damping parameter set to generate a damping correction increment lookup table; perform real-time correction on the damping feedback gain parameter in the servo valve control loop based on the damping correction increment lookup table to obtain the corrected damping feedback gain parameter set. Further, step S3 includes: S31: Perform sliding window segmentation and feature extraction on the transient damped residual time series data to obtain the transient residual feature sequence; Further, step S31 includes: S311: Divide the transient damping residual time series data into segments according to a preset sliding window width and a preset sliding step size to obtain several transient residual data segments; the sliding window width is a preset integer multiple of the servo valve control cycle; S312: Calculate the root mean square value of force residual, root mean square value of displacement residual, peak-to-peak value of force residual, and maximum rate of change of displacement residual for each transient residual data segment. S313: Combine the root mean square value of force residual, root mean square value of displacement residual, peak-to-peak value of force residual, and maximum value of displacement residual change rate of each transient residual data segment into a transient residual feature vector, and arrange the transient residual feature vectors of all transient residual data segments in time order to form a transient residual feature sequence. S32: Perform correlation calculations between the transient residual feature sequence and the current steady-state damping parameter set to generate a damping correction increment lookup table; Further, step S32 includes: S321: For each transient residual eigenvector in the transient residual characteristic sequence, calculate the ratio of the root mean square value of the force residual to the equivalent damping ratio of the current steady-state damping parameter ensemble test piece to obtain the damping identification deviation ratio; calculate the ratio of the root mean square value of the displacement residual to the average value of the servo valve opening correction amount in the corresponding time period of the transient damping compensation feedforward control sequence to obtain the compensation response deviation ratio; S322: Determine the damping correction increment of the test piece based on the damping identification deviation ratio, and determine the damping correction increment of the hydraulic system based on the compensation response deviation ratio; specifically, the damping correction increment of the test piece is equal to the damping identification deviation ratio multiplied by the equivalent damping ratio of the test piece and then multiplied by the preset test piece correction gain coefficient, and the damping correction increment of the hydraulic system is equal to the compensation response deviation ratio multiplied by the sum of the hydraulic system vertical channel oil damping coefficient and the hydraulic system horizontal channel oil damping coefficient and then multiplied by the preset hydraulic correction gain coefficient; S323: Using the peak-to-peak value of the force residual and the maximum value of the rate of change of the displacement residual in the transient residual eigenvector as the lookup index, and the corresponding damping correction increment of the test piece and the damping correction increment of the hydraulic system as the lookup output, a damping correction increment lookup table is constructed in chronological order. S33: Based on the damping correction incremental lookup table, the damping feedback gain parameter in the servo valve control loop is corrected in real time to obtain the corrected damping feedback gain parameter group. Further, step S33 includes: S331: During each servo valve control cycle in the working condition switching process, obtain the peak-to-peak value of the force residual and the maximum value of the displacement residual change rate at the current moment, and perform bilinear interpolation lookup in the damping correction increment lookup table to obtain the damping correction increment of the test piece and the damping correction increment of the hydraulic system at the current moment. S332: The original test piece damping feedback gain parameters in the servo valve control circuit are superimposed with the test piece damping correction increment corresponding to the current moment; the original hydraulic system damping feedback gain parameters in the servo valve control circuit are superimposed with the hydraulic system damping correction increment corresponding to the current moment, to obtain the corrected damping feedback gain parameter set; the corrected damping feedback gain parameter set includes the corrected test piece damping feedback gain parameters and the corrected hydraulic system damping feedback gain parameters; S4: Based on the modified damping feedback gain parameter set, the damping closed-loop adaptive control of the servo valve control loop is performed, the servo valve output of the vertical loading cylinder and the horizontal shear actuator is adjusted in real time, and the modified damping feedback gain parameter set is written back to the current working condition steady-state damping parameter set after the transient process ends, thus completing the transient damping adaptive control of the multi-working-condition continuous dynamic compression and shear test. Further, step S4 includes: S41: Replace the damping feedback gain parameters of the corresponding channels in the servo valve control circuit with the modified test piece damping feedback gain parameters and the modified hydraulic system damping feedback gain parameters in the modified damping feedback gain parameter group; within each servo valve control cycle, calculate the test piece damping compensation force component based on the modified test piece damping feedback gain parameters, calculate the hydraulic system damping compensation force component based on the modified hydraulic system damping feedback gain parameters, and superimpose the test piece damping compensation force component and the hydraulic system damping compensation force component into the force command signal of the servo valve control circuit to adjust the servo valve output of the vertical loading cylinder and the horizontal shearing actuator in real time; S42: Determine whether the transient process has ended within each servo valve control cycle; Further, step S42 includes: S421: Calculate the loading force tracking error and displacement tracking error of the current servo valve control cycle; the loading force tracking error is the ratio of the absolute value of the deviation between the measured loading force value and the target working condition command force value to the target working condition command force value, and the displacement tracking error is the ratio of the absolute value of the deviation between the measured displacement value and the target working condition command displacement value to the target working condition command displacement value. S422: Determine whether the loading force tracking error is less than the preset force steady-state convergence threshold and whether the displacement tracking error is less than the preset displacement steady-state convergence threshold; if the loading force tracking error is less than the force steady-state convergence threshold and the displacement tracking error is less than the displacement steady-state convergence threshold, then increment the continuous count by one; if the continuous count reaches the preset steady-state judgment window length, then the transient judgment process ends; if either error does not meet the condition, then reset the continuous count to zero and continue to execute the damping closed-loop adaptive control in step S41. S43: After the transient process ends, write the corrected damping feedback gain parameter set back to the current steady-state damping parameter set; Further, step S43 includes: S431: Convert the modified test specimen damping feedback gain parameter in the modified damping feedback gain parameter group into the updated test specimen equivalent damping ratio, and convert the modified hydraulic system damping feedback gain parameter into the updated hydraulic system vertical channel oil damping coefficient and the updated hydraulic system horizontal channel oil damping coefficient. S432: Replace the equivalent damping ratio of the test specimen in the current steady-state damping parameter set with the updated equivalent damping ratio of the test specimen; replace the hydraulic system vertical channel oil damping coefficient in the current steady-state damping parameter set with the updated hydraulic system vertical channel oil damping coefficient; replace the hydraulic system horizontal channel oil damping coefficient in the current steady-state damping parameter set with the updated hydraulic system horizontal channel oil damping coefficient; keep the equivalent stiffness of the test specimen unchanged; complete the update of the current steady-state damping parameter set; and use the updated current steady-state damping parameter set as the input for calculating the transient damping demand prediction vector for the working condition switching in step S13 during subsequent working condition switching; thus completing the transient damping adaptive control of the multi-working-condition continuous dynamic compression-shear test.

[0005] A damping adaptive control system for a dynamic compression-shear testing machine, used to implement the aforementioned damping adaptive control method for a dynamic compression-shear testing machine, the system comprising: Parameter identification and prediction module: During the steady-state loading stage, the hydraulic system pressure time series data, actuator displacement time series data and loading force time series data are collected simultaneously, and the steady-state damping parameter set of the current working condition is obtained through online recursive identification; and combined with the preset subsequent working condition loading parameter table, the transient damping demand prediction vector for working condition switching is calculated. Feedforward compensation and residual extraction module: Based on the predicted vector of transient damping demand during the working condition switching, a transient damping compensation feedforward control sequence is generated; at the execution time of the working condition switching, the servo valve control loop is injected, and the transient loading force time series data and transient displacement time series data during the working condition switching process are collected synchronously. The residual damping effect that the feedforward compensation fails to cover is extracted by comparison and calculation to obtain the transient damping residual time series data. Feedback gain correction module: Extracts features from the transient damping residual time series data to obtain a transient residual feature sequence; performs correlation calculation between the transient residual feature sequence and the current operating condition steady-state damping parameter set to generate a damping correction increment lookup table; performs real-time correction of the damping feedback gain parameters in the servo valve control loop based on the damping correction increment lookup table to obtain a corrected damping feedback gain parameter set; Closed-loop control and parameter update module: Based on the corrected damping feedback gain parameters, the module performs damped closed-loop adaptive control on the servo valve control loop, and after the transient process ends, it writes back the corrected damping parameters to update the current steady-state damping parameter set.

[0006] Compared to existing technologies, the advantages of this invention are as follows: During the steady-state loading stage, this invention identifies the damping characteristics of the test specimen and the hydraulic system's oil damping characteristics online. Combining the nonlinear abrupt change characteristics of the test specimen's damping with the hydraulic system's water hammer effect model, it predicts the transient damping requirements during condition switching in advance, generating a feedforward compensation sequence that matches the transient damping decay law. This sequence is actively injected into the servo valve control loop at the moment of condition switching, eliminating the inherent response lag of traditional feedback control and significantly reducing the overshoot of loading force and displacement tracking errors during condition switching. This invention achieves millisecond-level real-time closed-loop correction of feedforward compensation deviations through transient damping residual feature extraction and the construction of a damping correction increment lookup table, ensuring control stability during transient processes. After the transient process ends, the corrected damping parameters are written back to the steady-state damping parameter set, forming a positive feedback mechanism for parameter iterative optimization. This progressively improves the accuracy of damping prediction and compensation as the multi-condition continuous test progresses, shortening the transient process convergence time without the need for a conservative fixed waiting time, thus improving test efficiency. This invention effectively suppresses unpreset alternating loads during the transition between operating conditions, reduces the accumulation of unexpected structural damage to test specimens, ensures the accuracy and repeatability of dynamic compression-shear mechanical performance test results for components such as seismic isolation bearings, and improves the overall quality of material and component mechanical performance testing. Attached Figure Description

[0007] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0008] Figure 1 This is a flowchart of a damping adaptive control method for a dynamic compression-shear testing machine according to the present invention; Figure 2 This is a diagram illustrating the sensor deployment scheme in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the damping separation and identification principle in an embodiment of the present invention; Figure 4 The table below shows the damping nonlinear mutation characteristics of the test specimen in an embodiment of the present invention. Figure 5 This is a diagram illustrating the exponentially decaying envelope generation feedforward control sequence in an embodiment of the present invention. Figure 6 This is a comparison chart of transient damping residual extraction in an embodiment of the present invention; Figure 7 In an embodiment of the present invention, a sliding window is used to extract transient residual feature maps in segments; Figure 8 This is a structural diagram of the damping correction increment lookup table in an embodiment of the present invention; Figure 9 This is a calculation diagram of the dual-channel damping compensation force components in an embodiment of the present invention; Figure 10 This is a functional block diagram of a damping adaptive control system for a dynamic compression-shear testing machine according to the present invention. Detailed Implementation

[0009] 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.

[0010] Example 1: Please see Figure 1 As shown, this embodiment provides a damping adaptive control method for a dynamic compression-shear testing machine, including: This embodiment provides a damping adaptive control method for a dynamic compression-shear testing machine, applied to scenarios where a large-tonnage electro-hydraulic servo dynamic compression-shear testing machine performs multi-condition continuous dynamic compression-shear tests on components such as seismic isolation bearings. The dynamic compression-shear testing machine includes a vertical loading cylinder, a horizontal shear actuator, a servo valve control circuit, and a measurement and control system. The vertical loading cylinder applies a vertical pressure load to the test piece, and the horizontal shear actuator applies a horizontal shear load to the test piece. The multi-condition continuous dynamic compression-shear test refers to sequentially executing multiple dynamic compression-shear loading conditions with different combinations of loading frequencies, loading amplitudes, and loading directions on the same test piece, without unloading the test piece between conditions, and continuously switching between them. The method includes the following steps: S1: During the steady-state loading phase of the multi-condition continuous dynamic compression-shear test performed by the dynamic compression-shear testing machine, the hydraulic system pressure time-series data, actuator displacement time-series data, and loading force time-series data are collected simultaneously. The hydraulic system pressure time-series data, actuator displacement time-series data, and loading force time-series data are recursively identified online to obtain the current working condition steady-state damping parameter set. Based on the current working condition steady-state damping parameter set and the preset subsequent working condition loading parameter table, the transient damping demand prediction vector for working condition switching is calculated.

[0011] Further, step S1 includes the following steps: S11: Deploy pressure sensors on the oil supply lines of the vertical loading cylinder and the horizontal shear actuator of the dynamic compression-shear testing machine, deploy displacement sensors at the piston rod ends of the vertical loading cylinder and the horizontal shear actuator, and deploy force sensors at the load ends of the vertical loading cylinder and the horizontal shear actuator. During the steady-state loading phase, synchronously collect hydraulic system pressure timing data, actuator displacement timing data, and loading force timing data at a preset sampling period.

[0012] The pressure sensors are deployed at the inlet of the upper chamber oil supply line of the vertical loading cylinder, the inlet of the lower chamber oil supply line of the vertical loading cylinder, the inlet of the left chamber oil supply line of the horizontal shear actuator, and the inlet of the right chamber oil supply line of the horizontal shear actuator, totaling four pressure sensors, for real-time measurement of the oil supply pressure of each chamber. The displacement sensors are deployed at the piston rod ends of the vertical loading cylinder and the horizontal shear actuator, totaling two displacement sensors. These are magnetostrictive displacement sensors, used for real-time measurement of the displacement of the piston rods of each actuator. The force sensors are deployed at the load ends of the vertical loading cylinder and the horizontal shear actuator, totaling two force sensors, used for real-time measurement of the vertical loading force and horizontal shear force applied to the test piece. The preset sampling period is determined based on the highest loading frequency of the dynamic compression-shear test, specifically less than one-twentieth of the period corresponding to the highest loading frequency, to meet the requirements of the Nyquist sampling theorem and retain detailed information during the dynamic loading process. For example, when the highest loading frequency is 5 Hz, the preset sampling period can be set to 10 milliseconds. The synchronous acquisition refers to the simultaneous data sampling and timestamp recording of the aforementioned four pressure sensors, two displacement sensors, and two force sensors using the same clock source as a reference. The purpose of synchronous acquisition is to ensure that the hydraulic system pressure timing data, actuator displacement timing data, and loading force timing data are strictly aligned on the time axis. When a pressure fluctuation event occurs during a change in operating conditions, each channel sensor can simultaneously capture the response characteristics of the pressure fluctuation in different physical channels, providing a time-consistent data foundation for subsequent online recursive identification. The hydraulic system pressure timing data includes the pressure timing data of the upper chamber of the vertical loading cylinder, the pressure timing data of the lower chamber of the vertical loading cylinder, the pressure timing data of the left chamber of the horizontal shear actuator, and the pressure timing data of the right chamber of the horizontal shear actuator. The actuator displacement timing data includes the displacement timing data of the vertical loading cylinder and the displacement timing data of the horizontal shear actuator. The loading force timing data includes the vertical loading force timing data and the horizontal shear force timing data. The steady-state loading stage refers to the stage where, under the current working conditions, both the loading force tracking error and the displacement tracking error have converged to the preset steady-state range, and the test piece is in a periodic and stable loading state.

[0013] See Figure 2 This is a diagram illustrating the sensor deployment scheme for the dynamic compression-shear testing machine provided in an embodiment of this application. Figure 2As shown in the figure, this diagram illustrates the core mechanical structure and spatial layout of multiple sensors in a large-tonnage electro-hydraulic servo dynamic compression-shear testing machine. The test piece (rubber vibration isolation support) is located in the center of the diagram, bearing the vertical pressure from the upper vertical loading cylinder and the horizontal shear force from the left horizontal shear actuator. In the vertical channel, the upper and lower chamber pressure sensors are deployed at the inlet of the oil supply line, the displacement sensor is fixed to the piston rod end, and the force sensor is set on the pressure plate between the cylinder and the test piece. In the horizontal channel, the left and right chamber pressure sensors are deployed at the corresponding pipeline inlets, and the deployment logic of the displacement and force sensors is the same as in the vertical channel. In the multi-condition continuous dynamic compression-shear test scenario, the instantaneous switching of conditions will induce water hammer pressure pulsation in the hydraulic pipeline, and at the same time, the molecular chain segment motion state of the viscoelastic material inside the test piece will also undergo nonlinear abrupt changes. Traditional testing machines only deploy pressure sensors in the cylinder cavity, which cannot capture the pressure loss along the pipeline and the propagation process of water hammer waves; they only deploy displacement sensors on the surface of the test piece, which cannot separate the superimposed effect of the deformation of the test piece itself and the deformation of the frame. This invention achieves physical decoupling observation of the hydraulic system's oil damping characteristics and the test specimen's viscoelastic damping characteristics by simultaneously deploying sensors at three key locations: the pipeline inlet, the piston rod end, and the load end. This provides a multi-source data foundation with strict time alignment and clear physical meaning for the online recursive identification of the current operating condition's steady-state damping parameter set in step S1.

[0014] Specifically, in large-tonnage dynamic compression-shear testing machines, the oil supply lines for the vertical loading cylinder and the horizontal shear actuator are typically several meters to tens of meters long. During dynamic loading, the oil within these lines undergoes a periodic compression-release process. The elastic compression effect of the oil and the frictional effect along the pipeline together constitute the hydraulic damping characteristics of the hydraulic system. Pressure sensors are deployed at the inlet of the oil supply lines of each chamber, rather than inside the cylinder cavity, because the pressure at the inlet of the oil supply lines can simultaneously reflect the combined information of the servo valve outlet pressure, the pressure loss along the pipeline, and the pressure in the cylinder cavity. The pressure inside the cylinder cavity only reflects the local pressure state and cannot fully characterize the hydraulic damping characteristics of the pipeline system. Displacement sensors are deployed at the piston rod end, rather than on the surface of the test piece, because the piston rod displacement directly reflects the execution output of the servo valve control circuit. The displacement on the surface of the test piece also includes the superposition of the elastic deformation of the test piece itself and the deformation of the testing machine frame. Using the piston rod displacement as the system excitation input allows for a more accurate identification of the force-displacement hysteresis relationship of the test piece.

[0015] S12: Perform online recursive identification on hydraulic system pressure timing data, actuator displacement timing data, and loading force timing data to obtain the current steady-state damping parameter set.

[0016] Further, step S12 includes the following steps: S121: The actuator displacement timing data is used as the system excitation input, and the loading force timing data is used as the system response output. The recursive least squares method is used to identify the force-displacement hysteresis relationship of the test piece online, so as to obtain the equivalent stiffness and equivalent damping ratio of the test piece.

[0017] The force-displacement hysteresis relationship refers to the closed-loop curve relationship between the loading force and displacement of the test specimen during periodic dynamic loading. The area of ​​the closed-loop curve reflects the energy dissipated by the test specimen in each loading cycle, and the slope of the closed-loop curve reflects the equivalent stiffness of the test specimen. The online parameter identification process of the recursive least squares method is as follows: the force-displacement relationship of the test specimen is modeled as a linear viscoelastic model. The mechanical equation of the linear viscoelastic model is that the loading force equals the equivalent stiffness of the test specimen multiplied by the displacement plus the equivalent damping coefficient of the test specimen multiplied by the velocity, where the velocity is calculated by dividing the actuator displacement time series data by a preset sampling period using first-order difference. The displacement value and velocity value at each sampling time constitute the input vector, and the loading force value at the corresponding sampling time is used as the output scalar. The estimated values ​​of the equivalent stiffness and equivalent damping coefficient of the test specimen are updated recursively according to the sampling time. The specific recursive update process is as follows: At the nth sampling time, the predicted loading force value is calculated based on the input vector at the nth sampling time and the parameter estimate value at the previous sampling time. The difference between the predicted loading force value and the actual loading force value at the nth sampling time is used as the prediction error. The parameter estimate value is updated based on the prediction error and the recursive gain matrix. The recursive gain matrix is ​​calculated based on the covariance matrix at the previous sampling time and the input vector at the nth sampling time. The covariance matrix is ​​initialized at the first sampling time as an identity matrix multiplied by a preset large positive number. For example, the preset large positive number can be set to 10000 to ensure high sensitivity to new data in the initial stage. To prevent the covariance matrix from tending to zero during long-term recursion, which would cause the parameter estimate to lose its adaptive ability, a forgetting factor is introduced to correct the covariance matrix. The value of the forgetting factor is determined according to the loading frequency. The higher the loading frequency, the closer the forgetting factor is to 1 to maintain the stability of the parameter estimate. The lower the loading frequency, the further the forgetting factor is from 1 to enhance the tracking ability of slow parameter changes. For example, the forgetting factor can be set to 0.995. After identifying the equivalent stiffness and equivalent damping coefficient of the test specimen, the equivalent damping coefficient is divided by the square root of the product of twice the equivalent stiffness and the mass of the test specimen to obtain the equivalent damping ratio. The mass of the test specimen is a known quantity obtained before the test by weighing or referring to the test specimen parameter table.

[0018] For the vertical loading channel and the horizontal shear channel, the above-described recursive least squares identification process is executed independently. The vertical loading channel uses the vertical loading cylinder displacement time-series data as the system excitation input and the vertical loading force time-series data as the system response output. The horizontal shear channel uses the horizontal shear actuator displacement time-series data as the system excitation input and the horizontal shear force time-series data as the system response output. This yields the equivalent stiffness and equivalent damping ratio of the specimen in the vertical direction, and the equivalent stiffness and equivalent damping ratio of the specimen in the horizontal direction, respectively. For simplicity, the equivalent stiffness and equivalent damping ratio of the specimen mentioned in subsequent steps refer to parameter pairs that include both vertical and horizontal components.

[0019] S122: Based on the pressure difference time series data of the two chambers of the vertical loading cylinder and the pressure difference time series data of the two chambers of the horizontal shear actuator in the hydraulic system pressure time series data, and combined with the velocity components in the corresponding directions in the actuator displacement time series data, calculate the hydraulic system vertical channel oil damping coefficient and hydraulic system horizontal channel oil damping coefficient.

[0020] The time-series data of the pressure difference between the two chambers of the vertical loading cylinder is the time-series difference between the pressure time-series data of the upper chamber and the lower chamber of the vertical loading cylinder. The time-series data of the pressure difference between the two chambers of the horizontal shear actuator is the time-series difference between the pressure time-series data of the left chamber and the right chamber of the horizontal shear actuator. The velocity components in the corresponding directions of the actuator displacement time-series data are the vertical velocity time-series data obtained by dividing the vertical loading cylinder displacement time-series data by a preset sampling period after obtaining the first-order difference, and the horizontal velocity time-series data obtained by dividing the horizontal shear actuator displacement time-series data by a preset sampling period after obtaining the first-order difference. The calculation method for the hydraulic damping coefficient of the vertical channel of the hydraulic system is as follows: multiply the time-series data of the pressure difference between the two chambers of the vertical loading cylinder by the effective piston area of ​​the vertical loading cylinder to obtain the time-series data of the vertical hydraulic driving force; subtract the time-series data of the vertical loading force from the time-series data of the vertical hydraulic driving force to obtain the time-series data of the vertical hydraulic loss force; perform linear regression fitting on the time-series data of the vertical hydraulic loss force and the time-series data of the vertical velocity during multiple complete loading cycles in the steady-state loading stage, and the regression slope is the hydraulic damping coefficient of the vertical channel of the hydraulic system. The calculation method for the hydraulic damping coefficient of the horizontal channel of the hydraulic system is the same as that of the vertical channel, except that the corresponding data are replaced with the time-series data of the pressure difference between the two chambers of the horizontal shear actuator, the effective piston area of ​​the horizontal shear actuator, the time-series data of the horizontal shear force, and the time-series data of the horizontal velocity. The effective piston area of ​​the vertical loading cylinder and the effective piston area of ​​the horizontal shear actuator are known constants in the parameter table of the testing machine.

[0021] Specifically, the hydraulic damping characteristics of a hydraulic system originate from the viscous friction of the oil flowing in the pipeline and the elastic hysteresis effect of the oil during compression within the cylinder cavity. During the steady-state loading phase, the difference between the hydraulic driving force and the actual loading force borne by the test piece is primarily consumed by the hydraulic damping. By performing a linear regression fitting between the hydraulic loss force and the piston speed, the hydraulic damping characteristics of the hydraulic system can be separated from the complex coupled system and quantified into a damping coefficient. Data from multiple complete loading cycles is used for linear regression fitting instead of data from a single cycle because data from a single cycle may be affected by random factors such as loading waveform distortion and sensor measurement noise. Multi-cycle averaging can effectively reduce the impact of random errors on the damping coefficient identification results.

[0022] S123: Combine the equivalent stiffness of the test specimen, the equivalent damping ratio of the test specimen, the hydraulic damping coefficient of the vertical channel of the hydraulic system, and the hydraulic damping coefficient of the horizontal channel of the hydraulic system into a set of steady-state damping parameters for the current operating condition.

[0023] The data structure of the current operating condition steady-state damping parameter set is an ordered parameter group containing four parameter components, which are, in order, the equivalent stiffness of the test specimen, the equivalent damping ratio of the test specimen, the hydraulic damping coefficient of the vertical channel of the hydraulic system, and the hydraulic damping coefficient of the horizontal channel of the hydraulic system. Both the equivalent stiffness and the equivalent damping ratio of the test specimen include components in both the vertical and horizontal directions.

[0024] See Figure 3 This is a schematic diagram of the damping separation identification principle provided in the embodiments of this application. Figure 3As shown in the figure, the dashed lines separate the two coupled physical mechanisms of hydraulic system oil damping and test piece damping on the left and right sides. The dotted texture in the left box symbolizes the viscous friction and compressive elasticity effect of the oil flowing in the pipeline. The velocity arrow is input from the bottom, and the oil damping coefficient is output by fitting the linear relationship between hydraulic loss force and piston speed. The wavy texture in the right box symbolizes the energy dissipation process of the molecular chain segment motion inside the viscoelastic material. The loading force arrow is input from the bottom, and the area characteristics of the force-displacement hysteresis loop are identified by the recursive least squares method, and the equivalent damping ratio is output. In the traditional dynamic testing machine control algorithm, the total system damping is treated as a black box for compensation, ignoring the fundamentally different damping change mechanisms of the hydraulic system side and the test piece side: the former originates from the pipeline geometry and oil physical parameters, and is mainly dominated by the water hammer effect during working condition switching, with a fast change rate and an exponential decay law; the latter originates from the nonlinear characteristics of the material constitutive relationship, and is mainly dominated by the stress relaxation process during sudden load changes, with a large change amplitude and strong dependence on load history. By separating and identifying the damping parameters on both sides, this invention provides a clear quantitative baseline with a physical mechanism for estimating the transient damping increment of the test piece and the transient additional damping of the hydraulic system in step S13, thus avoiding the undercompensation or overcompensation problems caused by traditional algorithm compensation.

[0025] S13: Based on the current steady-state damping parameter set and the preset subsequent loading parameter table, calculate the transient damping demand prediction vector for the change of operating conditions.

[0026] Furthermore, step S13 includes the following steps: S131: Read the target loading frequency, target loading amplitude, and target loading direction of the subsequent working condition from the preset subsequent working condition loading parameter table.

[0027] The preset subsequent loading parameter table is a parameter table pre-compiled by the operators according to the test procedures and stored in the measurement and control system before the start of the test. This preset subsequent loading parameter table includes the loading frequency, loading amplitude, and loading direction of all conditions in this multi-condition continuous dynamic compression-shear test, with each condition arranged in a preset execution order. The target loading frequency is in Hertz, the target loading amplitude includes the vertical loading force amplitude and the horizontal shear displacement amplitude, and the target loading direction is the direction angle of the horizontal shear loading. When the steady-state loading phase of the current condition is about to end and preparations are made to switch to the subsequent condition, the measurement and control system reads the target loading frequency, target loading amplitude, and target loading direction corresponding to the next condition immediately adjacent to the current condition from the preset subsequent loading parameter table.

[0028] S132: Based on the equivalent stiffness and equivalent damping ratio of the test specimen in the current steady-state damping parameter set, and combined with the target loading frequency and target loading amplitude of the subsequent working conditions, calculate the load mutation amount. Based on the load mutation amount, look up the pre-calibrated test specimen damping nonlinear mutation characteristic table to obtain the estimated increment of transient damping of the test specimen.

[0029] The method for calculating the load mutation is as follows: Based on the target loading frequency and target loading amplitude of the subsequent working conditions, the expected steady-state loading force peak value of the subsequent working conditions is calculated using the equivalent stiffness of the test piece in the current working condition steady-state damping parameter set. The expected steady-state loading force peak value of the subsequent working conditions is subtracted from the actual steady-state loading force peak value of the current working condition, and the absolute value of the difference is taken as the load mutation. The actual steady-state loading force peak value of the current working condition is determined by the average peak value of the most recent several complete loading cycles in the loading force time series data of the steady-state loading stage. The number of several complete loading cycles is determined according to the stability of the loading waveform. For example, the average peak value of the most recent 5 complete loading cycles can be taken. The pre-calibrated test piece damping nonlinear mutation characteristic table is a lookup table established before the formal test of the test piece or in the historical test of the same model of test piece by applying step loads of different amplitudes to the test piece and recording the change rate of the force-displacement hysteresis loop area of ​​the test piece. The method for establishing the pre-calibrated test specimen damping nonlinear mutation characteristic table is as follows: Multiple step loads of different amplitudes are applied sequentially to the test specimen. After each step load application, the maximum deviation rate of the hysteresis loop area relative to the steady-state value is recorded during the transition process from the moment of load mutation to the point where the test specimen re-reaches a steady-state hysteresis loop. The step load amplitude is used as the input index, and the corresponding maximum deviation rate of the hysteresis loop area is converted into the equivalent damping ratio increment as the output value to construct a lookup table. Based on the load mutation amount calculated in step S132, a linear interpolation lookup is performed in the pre-calibrated test specimen damping nonlinear mutation characteristic table to obtain the estimated transient damping increment of the test specimen corresponding to the load mutation amount.

[0030] See Figure 4 This is a table showing the damping nonlinear mutation characteristics of the test specimen provided in the embodiments of this application. For example... Figure 4As shown, the table uses an 8-row × 8-column two-dimensional grid structure. The horizontal axis represents the increasing amplitude of the step load from low to high, and the vertical axis represents the increasing rate of deviation of the hysteresis loop area from low to high. The table internally uses lattice density to visually represent the magnitude of the equivalent damping ratio increment—the denser the lattice, the larger the corresponding estimated increment of transient damping. The legend in the lower right corner clarifies the mapping relationship of "sparse lattice - low increment", "medium lattice - medium increment", and "dense lattice - high increment". In test specimens made of viscoelastic materials such as rubber seismic isolation bearings, the damping characteristics have significant load amplitude dependence and rate correlation. When the load changes abruptly due to a change in operating conditions, the molecular chain segments inside the material cannot instantly complete the transition from the old equilibrium state to the new equilibrium state. Instead, there is a transition process in which the hysteresis loop area first increases sharply and then gradually decreases. Existing control algorithms only adjust the damping parameters identified in the steady-state phase, treating this nonlinear abrupt change effect as system noise and ignoring it. This results in a damping matching vacuum period of tens of milliseconds during the change of operating conditions, allowing the experimental specimen to experience non-preset alternating load amplitudes exceeding 30% of the design value. This invention, by pre-calibrating the characteristic table, transforms discrete test data into a rapidly queryable two-dimensional mapping relationship. This enables step S13 to predict the transient damping change trend on the experimental specimen side based on the load abrupt change before the change of operating conditions, providing a quantitative prediction of the damping change on the experimental specimen side for generating the feedforward compensation signal in step S2. This advances the starting point of damping compensation from after error detection to before the change of operating conditions.

[0031] Specifically, test specimens such as seismic isolation bearings are composed of viscoelastic materials such as rubber and lead cores. The damping characteristics of viscoelastic materials exhibit significant load amplitude and frequency dependence. When the load changes abruptly, the molecular chain segments within the material cannot instantly transition from one steady state to another, causing a nonlinear abrupt change in the equivalent damping of the test specimen at the moment of load change. This typically manifests as a rapid increase in damping followed by a gradual decrease to the new steady-state value during the transient process. Existing adaptive damping control algorithms only adjust the damping parameters identified in the steady-state phase, failing to incorporate the nonlinear abrupt change effect of load change on the test specimen's damping. Therefore, damping matching deviations occur at the moment of load condition switching. By pre-calibrating the nonlinear abrupt change characteristic table of the test specimen's damping and querying the estimated transient damping increment of the test specimen before load condition switching, the nonlinear abrupt change effect of the test specimen's damping can be incorporated into the transient damping demand prediction, providing a damping prediction basis on the test specimen side for the subsequent generation of feedforward compensation signals.

[0032] S133: Based on the hydraulic system's vertical channel oil damping coefficient and horizontal channel oil damping coefficient in the current steady-state damping parameter set, combined with the target loading frequency and target loading amplitude of the subsequent operating conditions, and based on the oil compressibility modulus and pipeline cavity volume parameters, calculate the estimated value of the transient fluctuation amplitude of the oil pressure at the moment of operating condition switching, and convert the estimated value of the transient additional damping of the hydraulic system into the estimated value of the transient fluctuation amplitude of the oil pressure.

[0033] The compressive elastic modulus of the hydraulic oil is the bulk elastic modulus of the hydraulic oil at the working temperature and pressure. This modulus is obtained from the hydraulic oil product parameter table or through online identification of the hydraulic system. The pipeline cavity volume parameters include the total cavity volume of the vertical loading cylinder supply pipeline and the total cavity volume of the horizontal shear actuator supply pipeline. These parameters are obtained from the testing machine equipment parameter table. The calculation method for the estimated transient fluctuation amplitude of the hydraulic pressure is as follows: At the instant of the operating condition switch, the servo valve control signal abruptly changes from the loading command of the current operating condition to the loading command of the subsequent operating condition. The change in the servo valve opening causes a sudden change in the flow rate of the hydraulic oil entering the cylinder cavity, resulting in pressure fluctuations within the pipeline cavity. Based on the basic model of the water hammer effect, the estimated transient fluctuation amplitude of the hydraulic pressure is equal to the hydraulic oil compressive elastic modulus divided by the pipeline cavity volume parameters, multiplied by the change in servo valve flow rate before and after the operating condition switch, multiplied by the quotient of the equivalent length of the pipeline and the propagation speed of the pressure wave in the hydraulic oil. The change in servo valve flow rate before and after the operating condition switch is calculated by multiplying the difference between the average servo valve opening during the current steady-state loading phase and the estimated average servo valve opening required for the subsequent operating condition by the servo valve flow gain coefficient. The estimated average servo valve opening required for the subsequent operating condition is estimated based on the target loading frequency, target loading amplitude, and hydraulic system oil damping coefficient in the current steady-state damping parameter set for the subsequent operating condition. The equivalent length of the pipeline and the propagation speed of the pressure wave in the oil are obtained from the equipment parameter table and hydraulic oil parameter table of the testing machine. The estimated transient wave amplitude is obtained by multiplying the estimated transient pressure transient fluctuation amplitude by the effective piston area of ​​the corresponding cylinder. The estimated transient wave amplitude is then divided by the estimated peak value of the piston speed at the moment of the operating condition switch to obtain the estimated transient additional damping of the hydraulic system. The estimated peak value of the piston speed is calculated based on the target loading frequency and target loading amplitude of the subsequent operating condition and is equal to twice pi multiplied by the target loading frequency multiplied by the target loading amplitude.

[0034] Specifically, the oil supply lines of large-tonnage dynamic compression-shear testing machines are typically long with a large volume of oil. When the operating conditions change rapidly, causing a rapid change in the servo valve opening, the inertial and compression effects of the oil in the lines can cause water hammer. This water hammer effect leads to pressure pulsations within the lines, which manifest as additional dynamic pressure within the cylinder cavity. This additional dynamic pressure acts on the piston, generating an additional force that hinders piston movement, equivalent to transient additional damping in the hydraulic system. Existing damping control algorithms do not consider the transient additional damping caused by the water hammer effect during operating condition changes. This results in insufficient damping compensation force in the servo valve control circuit after the operating condition change to overcome the actual change in total system damping, causing errors in loading force and displacement tracking to exceed tolerances. By predicting the transient fluctuation amplitude of oil pressure at the moment of working condition switching based on the water hammer effect model and converting it into the estimated transient additional damping of the hydraulic system, the damping change on the hydraulic system side can be predicted before the working condition switch. Together with the damping change on the test specimen side obtained in step S132, it constitutes a complete prediction of transient damping demand.

[0035] S134: Combine the estimated incremental transient damping of the test specimen, the estimated transient additional damping of the hydraulic system, and the target loading frequency, target loading amplitude, and target loading direction of the subsequent working conditions into a predictive vector for transient damping demand during working condition switching.

[0036] The data structure of the transient damping demand prediction vector for the working condition switching is an ordered vector containing five components, which, in order, are: the estimated transient damping increment of the test specimen, the estimated transient additional damping of the hydraulic system, the target loading frequency, the target loading amplitude, and the target loading direction. The transient damping demand prediction vector for the working condition switching encapsulates the estimated transient damping changes on both the test specimen and hydraulic system sides, along with the target loading parameters for subsequent working conditions, into a unified data structure, serving as the sole input for generating the transient damping compensation feedforward control sequence in step S2.

[0037] Step S1 achieves real-time quantitative separation of the damping characteristics of the test specimen and the hydraulic system by performing online recursive identification of the hydraulic system pressure time-series data, actuator displacement time-series data, and loading force time-series data during the steady-state loading phase. This yields the current steady-state damping parameter set as a baseline reference for subsequent transient damping prediction. Based on the steady-state damping parameter identification, Step S1 further combines a pre-set subsequent loading parameter table to proactively predict the damping change at the moment of condition switching from two dimensions: the nonlinear abrupt change in test specimen damping and the water hammer effect in the hydraulic system. This generates a transient damping demand prediction vector for condition switching, which includes the transient damping estimates for both the test specimen and the hydraulic system, as well as the target parameters for subsequent conditions. Compared to existing damping control algorithms that rely solely on post-event feedback adjustments based on current loading errors, step S1 introduces a damping demand prediction mechanism before the operating condition switch. This advances the starting point for damping control from the error detection time after the operating condition switch to the parameter estimation time before the switch, providing a time lead and damping quantification basis for generating the feedforward compensation signal in step S2. Step S1 employs recursive least squares for online identification rather than offline identification. This allows for continuous tracking of the slow drift of the test specimen's damping parameters during long-term loading due to material fatigue and softening without interrupting the test process. This ensures that the current steady-state damping parameter set always reflects the latest state of the test specimen and hydraulic system, avoiding transient damping prediction errors caused by outdated parameters. In step S1, the damping of the test specimen and the hydraulic system oil damping are identified separately rather than just the total system damping. This is because the damping of the test specimen and the hydraulic system oil damping have different change mechanisms at the moment of switching operating conditions. The abrupt change in the damping of the test specimen originates from the nonlinear constitutive response of the viscoelastic material, while the abrupt change in the hydraulic system oil damping originates from the water hammer effect in the pipeline. The change amplitude, change rate and decay law of the two are different. Separate identification is necessary to provide accurate baseline parameters for subsequent transient damping prediction by channel and mechanism.

[0038] For example, assuming the current operating condition is a steady-state loading stage with a horizontal shear amplitude of ±50 mm at 0.5 Hz, step S121 identifies the equivalent stiffness of the horizontal test piece as 3000 kN / mm and the equivalent damping ratio as 0.12. Step S122 identifies the hydraulic system's horizontal channel oil damping coefficient as 45 kN / mm. The target loading frequency for the subsequent operating condition is 2.0 Hz, and the target loading amplitude is ±100 mm. In step S132, the expected peak steady-state loading force for the subsequent operating condition is 3000 multiplied by 100, which equals 300000 kN. The actual peak steady-state loading force for the current operating condition is 3000 multiplied by 50, which equals 150000 kN. The load mutation is 300000 minus 150000, which equals 150000 kN. In the pre-calibrated test specimen damping nonlinear mutation characteristic table, the estimated transient damping increment corresponding to a load mutation of 150,000 kN is 0.035. In step S133, assuming the pipeline cavity volume is 0.02 cubic meters, the oil compressibility modulus is 1700 MPa, the equivalent pipeline length is 8 meters, the pressure wave propagation speed is 1200 m / s, the servo valve flow gain coefficient is 120 liters / min / mA, and the servo valve opening difference before and after the operating condition switch is 3.5 mA, then the flow rate change is 420 liters / min, and the estimated value of the transient fluctuation amplitude of the oil pressure is approximately 2.38 MPa calculated based on the water hammer model. Multiplying the estimated value of the transient fluctuation amplitude of the oil pressure of 2.38 MPa by the effective area of ​​the horizontal shear actuator piston and dividing it by the estimated peak value of the piston speed, the estimated value of the transient additional damping of the hydraulic system is obtained. The final transient damping demand prediction vector for the working condition switching is an ordered vector that includes the estimated incremental transient damping of the test specimen (0.035), the estimated transient additional damping of the hydraulic system, the target loading frequency (2.0 Hz), the target loading amplitude (±100 mm), and the target loading direction.

[0039] S2: Based on the transient damping demand prediction vector during operating condition switching, a transient damping compensation feedforward control sequence is generated; at the time of operating condition switching execution, the transient damping compensation feedforward control sequence is injected into the servo valve control loop, and the transient loading force time series data and transient displacement time series data during the operating condition switching process are collected simultaneously. The transient loading force time series data and transient displacement time series data are compared and calculated with the transient damping compensation feedforward control sequence to obtain the transient damping residual time series data.

[0040] Further, step S2 includes the following steps: S21: Generate a transient damping compensation feedforward control sequence based on the transient damping demand prediction vector during operating condition switching.

[0041] Further, step S21 includes the following steps: S211: Extract the estimated increment of transient damping of the test specimen and the estimated amount of transient additional damping of the hydraulic system from the transient damping demand prediction vector of the working condition switching, and sum the estimated increment of transient damping of the test specimen and the estimated amount of transient additional damping of the hydraulic system to obtain the estimated amount of transient additional damping.

[0042] The transient total additional damping estimate characterizes the superimposed total damping changes on the test specimen side and the hydraulic system side at the moment of operating condition switching. The physical meaning of the transient total additional damping estimate is the equivalent coefficient of the total damping force that the servo valve control circuit needs to overcome at the moment of operating condition switching.

[0043] S212: Determine the length of the transient compensation time window based on the estimated total transient additional damping and the target loading frequency in the transient damping demand prediction vector for operating condition switching.

[0044] The transient compensation time window length is a preset multiple of the period corresponding to the target loading frequency. This preset multiple is dynamically determined based on the ratio of the estimated transient total additional damping to the sum of the equivalent damping coefficients converted from the damping parameters in the current steady-state damping parameter set. Specifically, the determination method is as follows: first, multiply the equivalent damping ratio of the test piece in the current steady-state damping parameter set by twice the square root of the product of the test piece's equivalent stiffness and its mass, converting it to the test piece's equivalent damping coefficient. Then, calculate the ratio of the estimated transient total additional damping to the sum of the test piece's equivalent damping coefficient, the hydraulic system's vertical channel oil damping coefficient, and the hydraulic system's horizontal channel oil damping coefficient. Multiply this ratio by a reference multiple to obtain the preset multiple. The reference multiple is determined based on the response bandwidth of the testing machine's servo valve control circuit. A wider servo valve response bandwidth results in a smaller reference multiple, and a narrower servo valve response bandwidth results in a larger reference multiple. For example, the reference multiple can be set to 3. When the ratio is 0.3, the preset multiple is 0.3 multiplied by 3, which equals 0.9, rounded down to one target loading cycle; when the ratio is 1.5, the preset multiple is 1.5 multiplied by 3, which equals 4.5, rounded down to five target loading cycles. The transient compensation time window length is equal to the preset multiple multiplied by the cycle corresponding to the target loading frequency.

[0045] S213: Within the transient compensation time window, with the servo valve control cycle as the time step, the estimated total transient additional damping is discretized in the time domain according to the exponential decay envelope to generate the transient damping compensation amount corresponding to each time step.

[0046] The servo valve control cycle is the minimum time interval for updating control commands in the servo valve control loop. The servo valve control cycle is obtained from the configuration parameters of the testing machine's measurement and control system. For example, the servo valve control cycle can be 1 millisecond. The mathematical expression for the exponentially decaying envelope is: [The mathematical expression for the exponentially decaying envelope is missing from the original text]. Transient damping compensation amount corresponding to each time step Equal to the transient total additional damping estimate Multiply by the natural constant The power of, where the power is negative With servo valve control cycle The product divided by the decay time constant ,Right now: ; Where k is the time step number, and the value of k ranges from 0 to the integer part of the quotient of the transient compensation time window length divided by the servo valve control cycle. The decay time constant of the exponential decay envelope is determined based on the larger value of the hydraulic system's vertical channel oil damping coefficient and the hydraulic system's horizontal channel oil damping coefficient. The specific determination method is as follows: take the larger value of the hydraulic system's vertical channel oil damping coefficient and the hydraulic system's horizontal channel oil damping coefficient, divide the larger value by the square of the piston's effective area to obtain the hydraulic resistance, calculate the quotient of the corresponding pipeline cavity volume parameter divided by the hydraulic fluid's compressive elastic modulus to obtain the hydraulic capacity, and use the product of the hydraulic resistance and the hydraulic capacity as the decay time constant. The pipeline corresponding to the larger value is the pipeline with a slower pressure fluctuation decay rate. Determining the decay time constant of the exponential decay envelope based on the pressure fluctuation decay rate of the pipeline can ensure that the decay rate of the transient damping compensation amount matches the decay rate of the actual pressure fluctuation, avoiding insufficient compensation due to excessively fast decay or overcompensation due to excessively slow decay.

[0047] The time-domain expansion of the estimated transient total additional damping using an exponentially decaying envelope is employed because both the pressure pulsations caused by water hammer in the pipeline and the damping relaxation process after a sudden load change in the viscoelastic material exhibit exponential decay characteristics. The pressure pulsations in the pipeline decay exponentially with time under the influence of pipeline wall friction and oil viscosity, and the stress relaxation process of the viscoelastic material after a load step also approximately follows an exponential decay law. Therefore, using the exponentially decaying envelope as the time distribution function of the transient damping compensation can match the temporal evolution law of actual transient damping changes at the physical mechanism level.

[0048] See Figure 5 This is a feedforward control sequence diagram for generating the exponentially decaying envelope, provided in an embodiment of this application. For example... Figure 5As shown, the vertical axis represents the transient damping compensation amount, and the horizontal axis represents the time evolution from the moment of operating condition switching. The smooth curve in the figure shows the envelope shape of the estimated total transient additional damping decaying exponentially over time, gradually approaching zero from the initial peak until the end of the transient compensation time window. The dense vertical short lines below the curve represent discretized sampling with a servo valve control cycle (e.g., 1 millisecond) as the time step. Each short line corresponds to the transient damping compensation amount that needs to be superimposed on the control command within one servo valve control cycle. In large-tonnage hydraulic systems, the pipeline water hammer pressure pulsation and the stress relaxation process of viscoelastic materials caused by operating condition switching both exhibit exponential decay characteristics: the water hammer pressure decays exponentially under the action of pipeline wall friction, and the stress relaxation of viscoelastic materials also approximately follows the exponential decay of the Maxwell model. This invention uses an exponential decay envelope instead of a rectangular pulse or linear ramp envelope because the former matches the time evolution law of actual transient damping changes at the physical mechanism level, and can avoid secondary impacts caused by premature removal of compensation force or energy waste caused by maintaining compensation force for too long. By discretizing the estimated total transient additional damping into a time series according to the exponentially decaying envelope, the feedforward control sequence generated in step S21 can accurately apply a compensation force that matches the transient damping change at the current moment in each servo valve control cycle after the operating condition switch, thus realizing "precise tracking of the time profile" of damping compensation during the transient process.

[0049] S214: Convert the transient damping compensation amount corresponding to each time step into the corresponding servo valve opening correction amount, and arrange them in time order to form a transient damping compensation feedforward control sequence.

[0050] The method for converting the transient damping compensation amount into the servo valve opening correction amount is as follows: Multiply the transient damping compensation amount by the estimated piston speed for the corresponding time step to obtain the transient damping compensation force; divide the transient damping compensation force by the product of the servo valve flow gain coefficient and the effective area of ​​the cylinder piston to obtain the servo valve opening correction amount. The estimated piston speed for the corresponding time step is calculated using a sinusoidal motion model based on the target loading frequency and target loading amplitude of the subsequent operating conditions. The servo valve flow gain coefficient is obtained from the testing machine's equipment parameter table. The data structure of the transient damping compensation feedforward control sequence is an array of servo valve opening correction amounts arranged by time step number. The array length is equal to the quotient of the transient compensation time window length divided by the servo valve control period, rounded down, plus 1.

[0051] S22: At the moment of operation condition switching, the transient damping compensation feedforward control sequence is sequentially superimposed into the command signal of the servo valve control circuit according to the time step; the transient loading force timing data and transient displacement timing data of the vertical loading cylinder and the horizontal shear actuator are collected simultaneously during the operation condition switching process.

[0052] The working condition switching execution time is the moment when the measurement and control system switches the loading command signal of the current working condition to the loading command signal of the subsequent working condition. Starting at the working condition switching execution time, within each servo valve control cycle, the measurement and control system superimposes the servo valve opening correction amount corresponding to the current time step number in the transient damping compensation feedforward control sequence onto the original command signal of the servo valve control loop. The superposition method is to algebraically add the servo valve opening correction amount to the servo valve opening command in the original command signal. Synchronous acquisition refers to acquiring transient loading force timing data and transient displacement timing data of the vertical loading cylinder and horizontal shear actuator using the same preset sampling period and synchronous clock source as in step S11, while simultaneously superimposing the transient damping compensation feedforward control sequence. The transient loading force timing data includes vertical loading force timing data and horizontal shear force timing data during the working condition switching process. The transient displacement timing data includes vertical loading cylinder displacement timing data and horizontal shear actuator displacement timing data during the working condition switching process.

[0053] S23: Compare and calculate the transient loading force time series data and transient displacement time series data with the transient damping compensation feedforward control sequence to obtain the transient damping residual time series data.

[0054] Further, step S23 includes the following steps: S231: Based on the servo valve opening correction amount at each time step in the transient damping compensation feedforward control sequence, and combined with the servo valve flow gain coefficient, calculate the expected compensation force increment sequence at each time step.

[0055] The method for calculating the expected compensation force increment at each time step is as follows: multiply the servo valve opening correction amount at the corresponding time step by the servo valve flow gain coefficient and the effective area of ​​the cylinder piston to obtain the expected compensation force increment at the corresponding time step. Arrange the expected compensation force increments of all time steps in chronological order to form the expected compensation force increment sequence.

[0056] S232: Calculate the actual load force change at each time step based on the transient load force time series data, and subtract the expected compensation force increment and the target working condition command force increment at the corresponding time step from the actual load force change to obtain the force residual value at each time step.

[0057] The actual change in loading force at each time step is the difference between the loading force value at the current time step and the loading force value at the previous time step in the transient loading force time series data. The target operating condition command force increment is the difference between the command force value at the current time step and the command force value at the previous time step in the subsequent operating condition loading command signal. The physical meaning of the force residual value at each time step is: the residual force change after deducting the force change expected by feedforward compensation and the force change required by the target operating condition command from the actual change in loading force. The residual force change reflects the damping change effect that the transient damping compensation feedforward control sequence cannot fully cover.

[0058] S233: Calculate the actual displacement change at each time step based on the transient displacement time series data, and subtract the target working condition command displacement increment from the actual displacement change to obtain the displacement residual value at each time step.

[0059] The actual displacement change at each time step is the difference between the displacement value of the current time step and the displacement value of the previous time step in the transient displacement time sequence data. The target operating condition command displacement increment is the difference between the command displacement value of the current time step and the command displacement value of the previous time step in the subsequent operating condition loading command signal.

[0060] S234: Arrange the force residuals and displacement residuals at each time step in chronological order to form transient damping residual time series data.

[0061] The data structure of the transient damping residual time series data is a two-column time series array. The first column is a sequence of force residual values ​​arranged by time step, and the second column is a sequence of displacement residual values ​​arranged by time step. The two columns of data correspond one-to-one on the time axis.

[0062] See Figure 6 This is a comparison chart of transient damping residual extraction provided in the embodiments of this application. Figure 6As shown in the figure, four horizontal time bands are arranged from top to bottom, intuitively illustrating the calculation logic of residual extraction. The topmost expected compensation force increment sequence shows the force change waveform expected to be generated by the feedforward control sequence, exhibiting an exponential decay characteristic; the second target condition command force increment shows the force change waveform required by the subsequent condition loading command, exhibiting a sinusoidal increasing trend; the third actual loading force change shows the force change waveform actually generated by the system under the dual action of feedforward compensation and target command, its shape superimposed with the characteristics of the first two and containing residual fluctuations; the force residual value at the bottom, indicated by the dashed arrow, is obtained by subtracting the first two from the third, with the waveform amplitude significantly reduced and gradually approaching zero. The comparison calculation marked in brackets on the right side of the figure emphasizes the operational relationship between the four time bands. During the transient process of condition switching, even with the use of prediction-based feedforward compensation, the damping response of the actual system may still deviate from the expected value due to factors such as simplification of the prediction model, nonlinear coupling effects, and sensor measurement delays. Traditional control algorithms ignore such deviations, causing the prediction error of feedforward compensation to accumulate into a systematic tracking error in the subsequent steady-state stage. This invention extracts the residual damping effect not covered by feedforward compensation by comparing the actual acquired transient loading force and displacement data with the expected effect of the feedforward sequence step by step, forming transient damping residual time series data. This residual data quantifies the deviation between the prediction model and the actual system, providing a precise error signal for closed-loop correction in step S3 to construct the damping correction increment lookup table. This upgrades the damping adaptive control from open-loop control that relies solely on feedforward prediction to composite control that combines feedforward and closed-loop correction.

[0063] Step S2 first generates a transient damping compensation feedforward control sequence based on the transient damping demand prediction vector output from Step S1. At the moment of operating condition switching, this sequence is actively injected into the servo valve control loop to compensate for the estimated transient damping change in advance. Then, by comparing the actual acquired transient loading force and transient displacement time-series data with the expected effect of the transient damping compensation feedforward control sequence, the residual damping effect not covered by the feedforward compensation is extracted, forming transient damping residual time-series data. Step S2 employs a strategy combining feedforward compensation and residual extraction. Based on the prediction results from Step S1, the feedforward compensation actively applies compensation force during operating condition switching, eliminating the need to wait for error accumulation before passive feedback adjustment, thus eliminating the inherent response lag of traditional feedback control in the time dimension. The transient damping residual time-series data quantifies the prediction deviation of the feedforward compensation, providing an accurate error signal for the closed-loop correction in Step S3. If the feedforward prediction in step S1 is relied upon without the residual extraction in step S2, the prediction deviation cannot be detected and corrected. The accuracy of the feedforward compensation depends entirely on the accuracy of the test piece damping nonlinear mutation characteristic table and the water hammer effect model in step S1. However, non-ideal factors in the actual working condition switching process may cause prediction deviation, so step S3 needs to make real-time correction based on the transient damping residual time series data.

[0064] S3: Perform sliding window segmentation and feature extraction on the transient damping residual time series data to obtain the transient residual feature sequence; perform correlation calculation between the transient residual feature sequence and the current operating condition steady-state damping parameter set to generate a damping correction increment lookup table; perform real-time correction on the damping feedback gain parameter in the servo valve control loop based on the damping correction increment lookup table to obtain the corrected damping feedback gain parameter set.

[0065] Furthermore, step S3 includes the following steps: S31: Perform sliding window segmentation and feature extraction on the transient damped residual time series data to obtain the transient residual feature sequence.

[0066] Further, step S31 includes the following steps: S311: Divide the transient damped residual time series data into segments according to the preset sliding window width and preset sliding step size to obtain several transient residual data segments.

[0067] The width of the sliding window is a preset integer multiple of the servo valve control cycle. This preset integer multiple is determined based on the target loading frequency, specifically one-tenth of the quotient of the target loading frequency corresponding to the cycle divided by the servo valve control cycle, rounded down. This ensures that each sliding window contains enough sampling points for statistical feature calculation, while the window width does not exceed one-tenth of the loading cycle to guarantee resolution for detailed changes in the transient process. For example, when the target loading frequency is 2 Hz and the servo valve control cycle is 1 millisecond, the target loading frequency corresponds to a cycle of 500 milliseconds. The sliding window width is 500 divided by 1 divided by 10, rounded down, equal to 50 servo valve control cycles, i.e., 50 milliseconds. The preset sliding step size is half the width of the sliding window to achieve 50% overlap between adjacent transient residual data segments, improving the temporal resolution of the feature sequence.

[0068] S312: Calculate the root mean square value of force residual, root mean square value of displacement residual, peak-to-peak value of force residual, and maximum rate of change of displacement residual for each transient residual data segment.

[0069] The method for calculating the root mean square value of the force residual is as follows: square, average, and then take the square root of the force residual values ​​for all time steps in the transient residual data segment. The method for calculating the root mean square value of the displacement residual is as follows: square, average, and then take the square root of the displacement residual values ​​for all time steps in the transient residual data segment. The method for calculating the peak-to-peak value of the force residual is as follows: subtract the minimum value from the maximum value of the force residual values ​​for all time steps in the transient residual data segment. The method for calculating the maximum value of the displacement residual change rate is as follows: calculate the difference between the displacement residual values ​​for adjacent time steps in the transient residual data segment, divide by the servo valve control cycle to obtain the displacement residual change rate, and take the maximum absolute value of all displacement residual change rates.

[0070] S313: Combine the root mean square value of force residual, root mean square value of displacement residual, peak-to-peak value of force residual, and maximum value of displacement residual change rate of each transient residual data segment into a transient residual feature vector, and arrange the transient residual feature vectors of all transient residual data segments in chronological order to form a transient residual feature sequence.

[0071] The transient residual feature vector is an ordered vector containing four components, which are, in order, the root mean square value of the force residual, the root mean square value of the displacement residual, the peak-to-peak value of the force residual, and the maximum rate of change of the displacement residual. The transient residual feature sequence is an array of transient residual feature vectors arranged in chronological order, with the array length equal to the total number of transient residual data segments.

[0072] See Figure 7 This is the transient residual feature map extracted piecewise using a sliding window, as provided in the embodiments of this application. Figure 7As shown, the high-frequency fluctuation curve at the bottom of the figure represents the original waveform of the transient damping residual time series data. Five inverted U-shaped arcs overlay it represent five sliding window segments, with adjacent windows overlapping by 50% (the sliding step size is half the window width). Each window segment is indicated by a dashed arrow pointing upwards to a rectangle containing four feature icons arranged from top to bottom: a solid black circle representing the root mean square value of the force residual, a white circle with a black border representing the root mean square value of the displacement residual, an I-shaped vertical line representing the peak-to-peak value of the force residual, and a broken line peak representing the maximum rate of change of the displacement residual. All feature boxes are arranged horizontally in chronological order to form the transient residual feature sequence at the top. The legend area on the right explains the physical meaning of the four feature icons in detail. During the transient process, the residual value of a single sampling point is significantly affected by accidental factors such as sensor noise and numerical calculation truncation errors. Directly using these values ​​for damping correction would introduce high-frequency oscillations. This invention segments continuous time-series data into several time segments using a sliding window. Statistical features are calculated within each segment, suppressing measurement noise while preserving the evolutionary trend of the transient process. The use of a 50% overlap segmentation method, rather than a contiguous segmentation, improves the temporal resolution of the feature sequence, ensuring that abrupt changes are not missed even if they fall precisely at the window boundary. Of the four extracted features, the root mean square value reflects the energy level of the residual, the peak-to-peak value reflects the extreme fluctuation amplitude, and the maximum rate of change reflects the dynamic response speed. The combination of these four features constitutes a comprehensive, multi-dimensional characterization of the transient residual. Step S32 performs correlation calculations between these feature sequences and the current steady-state damping parameter set to generate a damping correction increment lookup table, providing the foundation for the real-time parameter correction in step S33, from time-series data to correction values.

[0073] S32: Perform correlation calculations between the transient residual feature sequence and the current steady-state damping parameter set to generate a damping correction increment lookup table.

[0074] Further, step S32 includes the following steps: S321: For each transient residual eigenvector in the transient residual characteristic sequence, calculate the ratio of the root mean square value of the force residual to the equivalent damping ratio of the current steady-state damping parameter ensemble test piece to obtain the damping identification deviation ratio; calculate the ratio of the root mean square value of the displacement residual to the average value of the servo valve opening correction amount in the corresponding time period of the transient damping compensation feedforward control sequence to obtain the compensation response deviation ratio.

[0075] The time period corresponding to the transient damping compensation feedforward control sequence refers to the time range covered by the transient residual data segment corresponding to the current transient residual feature vector on the time axis. The average value of the servo valve opening correction is the arithmetic mean of the servo valve opening correction values ​​for each time step within the time range in the transient damping compensation feedforward control sequence. The physical meaning of the damping identification deviation ratio is: the normalized deviation of the force residual relative to the equivalent damping ratio of the test piece. The larger the damping identification deviation ratio, the larger the prediction deviation of the transient damping of the test piece in step S1. The physical meaning of the compensation response deviation ratio is: the normalized deviation of the displacement residual relative to the feedforward compensation strength. The larger the compensation response deviation ratio, the larger the deviation between the actual response and the expected response of the hydraulic system to the feedforward compensation signal.

[0076] S322: Determine the damping correction increment of the test piece based on the damping identification deviation ratio, and determine the damping correction increment of the hydraulic system based on the compensation response deviation ratio.

[0077] The damping correction increment of the test specimen is equal to the damping identification deviation ratio multiplied by the equivalent damping ratio of the test specimen, and then multiplied by a preset test specimen correction gain coefficient. The preset test specimen correction gain coefficient is determined based on the test specimen material type and the calibration accuracy of the test specimen damping nonlinear mutation characteristic table. Higher calibration accuracy results in a smaller test specimen correction gain coefficient, and lower calibration accuracy results in a larger test specimen correction gain coefficient. For example, for lead-core rubber bearing test specimens, the test specimen correction gain coefficient can be set to 0.6. The hydraulic system damping correction increment is equal to the compensation response deviation ratio multiplied by the sum of the hydraulic system vertical channel oil damping coefficient and the hydraulic system horizontal channel oil damping coefficient, and then multiplied by a preset hydraulic correction gain coefficient. The preset hydraulic correction gain coefficient is determined based on the complexity of the hydraulic system piping and the simplification of the water hammer effect model. More complex piping or a higher degree of model simplification results in a larger hydraulic correction gain coefficient. For example, the hydraulic correction gain coefficient can be set to 0.5.

[0078] S323: Using the peak-to-peak value of the force residual and the maximum value of the rate of change of the displacement residual in the transient residual eigenvector as the lookup index, and the corresponding damping correction increment of the test piece and the damping correction increment of the hydraulic system as the lookup output, a damping correction increment lookup table is constructed in chronological order.

[0079] The damping correction increment lookup table has a two-dimensional data structure. The row index is the peak-to-peak value of the force residual, and the column index is the maximum rate of change of the displacement residual. Each cell in the table stores the corresponding damping correction increment for the test specimen and the damping correction increment for the hydraulic system. The row and column indices of the damping correction increment lookup table are determined by the actual values ​​of the peak-to-peak value of the force residual and the maximum rate of change of the displacement residual among all transient residual feature vectors in the transient residual feature sequence. The peak-to-peak value of the force residual and the maximum rate of change of the displacement residual are selected as the lookup index instead of the root mean square value of the force residual and the root mean square value of the displacement residual because they reflect the extreme fluctuation degree and the fastest rate of change of the damping residual during the transient process, which has a more direct impact on the stability of the servo valve control loop. Using extreme features as the index allows for matching the damping correction amount at the most dangerous operating point.

[0080] See Figure 8 This is a structural diagram of the damping correction increment lookup table provided in the embodiments of this application. For example... Figure 8 As shown, the table uses a two-dimensional grid structure. The peak-to-peak value of the force residual on the horizontal axis is divided into five levels from left to right: low, lower, medium, higher, and high. The maximum value of the displacement residual change rate on the vertical axis is divided into four levels from bottom to top: low, lower, medium, and high. This forms 5 columns × 4 rows, totaling 20 cells. Each cell contains two squares: the upper dark square represents the damping correction increment of the test specimen, and the lower light square represents the damping correction increment of the hydraulic system. The size of the squares increases with the row and column levels. The square in the upper right cell (high force residual peak-to-peak value + high displacement residual change rate) is the largest, and the square in the lower left cell (low force residual peak-to-peak value + low displacement residual change rate) is the smallest. The arrow in the lower right corner of the figure indicates that the larger the square, the larger the value. The legend area on the right clearly explains the physical meaning of the dark and light squares. During the transient process of switching operating conditions, the peak-to-peak value of the force residual reflects the degree of extreme fluctuation of the loading force caused by the damping prediction deviation of the test specimen, and the maximum value of the displacement residual change rate reflects the deviation of the hydraulic system's response speed to the feedforward compensation signal. These two extreme features were chosen as lookup indices instead of root mean square values ​​because they pose the most direct threat to the stability of the servo valve control loop. Extreme fluctuations may trigger the saturation limiting of the servo valve, and rapid changes may induce high-frequency resonance in the control loop. This invention constructs a two-dimensional lookup table to pre-calculate and store the complex nonlinear mapping relationship between the discrete transient residual feature vector and the damping correction increment. This allows step S33 to obtain the correction amount by performing only one bilinear interpolation lookup within each 1-millisecond servo valve control cycle, with the calculation delay controlled at the microsecond level, meeting the stringent timing requirements of real-time control. Simultaneously, the correction increments for both the test piece and the hydraulic system are output separately, ensuring that subsequent parameter corrections can be applied precisely by channel and mechanism, avoiding inter-channel interference caused by a one-size-fits-all correction.

[0081] S33: Based on the damping correction increment lookup table, the damping feedback gain parameters in the servo valve control loop are corrected in real time to obtain the corrected damping feedback gain parameter set.

[0082] Further, step S33 includes the following steps: S331: During each servo valve control cycle in the working condition switching process, obtain the peak-to-peak value of the force residual and the maximum value of the displacement residual change rate at the current moment, and perform bilinear interpolation lookup in the damping correction increment lookup table to obtain the damping correction increment of the test piece and the damping correction increment of the hydraulic system at the current moment.

[0083] The peak-to-peak value of the force residual and the maximum value of the displacement residual change rate at the current moment are provided by the feature calculation results containing the latest transient residual data segment at the current moment. The bilinear interpolation lookup method is as follows: In the damping correction increment lookup table, find the two adjacent row index values ​​that are closest to the peak-to-peak value of the force residual at the current moment in the dimension of the peak-to-peak value of the force residual, and find the two adjacent column index values ​​that are closest to the maximum value of the displacement residual change rate at the current moment in the dimension of the maximum value of the displacement residual change rate. Using the damping correction increment of the test specimen and the damping correction increment of the hydraulic system stored in the four nearest neighbor cells as the interpolation base points, bilinear interpolation is performed according to the position ratio of the peak-to-peak value of the force residual and the maximum value of the displacement residual change rate at the current moment between the adjacent index values ​​to obtain the damping correction increment of the test specimen and the damping correction increment of the hydraulic system corresponding to the current moment.

[0084] S332: Add the original test piece damping feedback gain parameter in the servo valve control circuit to the test piece damping correction increment corresponding to the current moment, and add the original hydraulic system damping feedback gain parameter in the servo valve control circuit to the hydraulic system damping correction increment corresponding to the current moment to obtain the corrected damping feedback gain parameter group.

[0085] The original test piece damping feedback gain parameters and hydraulic system damping feedback gain parameters in the servo valve control circuit are the damping feedback gain parameters used by the servo valve control circuit before the operating condition switch. These damping feedback gain parameters are determined by the damping parameters corresponding to the current steady-state damping parameter set before the operating condition switch. The corrected damping feedback gain parameter set includes the corrected test piece damping feedback gain parameters and the corrected hydraulic system damping feedback gain parameters. The corrected test piece damping feedback gain parameter is equal to the original test piece damping feedback gain parameter plus the test piece damping correction increment corresponding to the current moment. The corrected hydraulic system damping feedback gain parameter is equal to the original hydraulic system damping feedback gain parameter plus the hydraulic system damping correction increment corresponding to the current moment.

[0086] Step S3 performs sliding window segmentation and feature extraction on the transient damping residual time-series data output from Step S2, transforming the continuous time-series residual data into a transient residual feature sequence reflecting the evolution trend of the damping residual during the transient process. By associating the transient residual feature sequence with the current steady-state damping parameter set, a mapping relationship from residual features to damping correction increments is established, generating a damping correction increment lookup table. Finally, based on this lookup table, the damping feedback gain parameter in the servo valve control loop is corrected in real time during each servo valve control cycle. Step S3 constructs a closed-loop path from residual detection to parameter correction, compensating for the residual damping mismatch problem caused by prediction errors in the feedforward compensation in Step S2. The reason for using sliding window segmentation instead of calculating features point-by-point in Step S3 is that the residual value point-by-point is greatly affected by sensor measurement noise, making it difficult to reflect the trend change of the damping residual. The statistical features within the sliding window can capture the overall trend of the damping residual while suppressing measurement noise. The reason for generating a damping correction increment lookup table in step S3 instead of directly using the residual characteristics as the feedback signal is that the lookup table pre-calculates and stores the correspondence between discrete residual characteristics and damping correction increments. Only one bilinear interpolation lookup is needed in each servo valve control cycle to obtain the correction amount, resulting in minimal computational latency and meeting the millisecond-level real-time requirements of the servo valve control cycle. If an online iterative optimization algorithm were used instead of the lookup table in step S3, multiple iterative calculations would be required in each control cycle, leading to a computational latency far exceeding the servo valve control cycle and failing to meet the real-time correction requirements. In step S3, the damping correction increment of the test specimen and the damping correction increment of the hydraulic system are calculated separately and superimposed on their respective feedback gain parameters, instead of being combined into a single total damping correction increment. This is because the damping feedback gain parameters of the test specimen and the hydraulic system act on different compensation force calculation channels in the servo valve control loop. The damping feedback gain parameter of the test specimen is used to calculate the compensation force component that matches the force-displacement hysteresis characteristics of the test specimen, while the damping feedback gain parameter of the hydraulic system is used to calculate the compensation force component that matches the pressure fluctuation characteristics of the hydraulic pipeline. The compensation force calculation methods for the two channels are different, and they need to be corrected separately.

[0087] S4: Based on the modified damping feedback gain parameter set, the damping closed-loop adaptive control of the servo valve control loop is performed, the servo valve output of the vertical loading cylinder and the horizontal shear actuator is adjusted in real time, and the modified damping feedback gain parameter set is written back to the current working condition steady-state damping parameter set after the transient process ends, thus completing the transient damping adaptive control of the multi-working-condition continuous dynamic compression and shear test.

[0088] Further, step S4 includes the following steps: S41: Replace the damping feedback gain parameters of the corresponding channels in the servo valve control circuit with the modified test piece damping feedback gain parameters and the modified hydraulic system damping feedback gain parameters in the modified damping feedback gain parameter group; within each servo valve control cycle, calculate the test piece damping compensation force component based on the modified test piece damping feedback gain parameters, calculate the hydraulic system damping compensation force component based on the modified hydraulic system damping feedback gain parameters, and superimpose the test piece damping compensation force component and the hydraulic system damping compensation force component into the force command signal of the servo valve control circuit to adjust the servo valve output of the vertical loading cylinder and the horizontal shear actuator in real time.

[0089] The calculation method for the damping compensation force component of the test specimen is as follows: multiply the corrected damping feedback gain parameter of the test specimen by the difference between the measured value of the applied force and the target working condition command force value within the current servo valve control cycle to obtain the damping compensation force component of the test specimen. The calculation method for the damping compensation force component of the hydraulic system is as follows: multiply the corrected damping feedback gain parameter of the hydraulic system by the difference between the measured value of the displacement and the target working condition command displacement value within the current servo valve control cycle, and then multiply by the equivalent stiffness of the test specimen in the current working condition steady-state damping parameter set to obtain the damping compensation force component of the hydraulic system. The damping compensation force component of the test specimen and the damping compensation force component of the hydraulic system are added together and superimposed on the force command signal of the servo valve control loop. The superposition method is to convert the sum of the damping compensation force component of the test specimen and the damping compensation force component of the hydraulic system into a servo valve opening compensation amount and then algebraically add it to the servo valve opening command corresponding to the force command signal. The conversion method for the servo valve opening compensation is as follows: divide the sum of the damping compensation force component of the test piece and the damping compensation force component of the hydraulic system by the product of the servo valve flow gain coefficient and the effective area of ​​the cylinder piston.

[0090] See Figure 9 This is a calculation diagram of the dual-channel damping compensation force components provided in an embodiment of this application. Figure 9As shown in the figure, the test specimen channel on the left and the hydraulic system channel on the right demonstrate the independent calculation logic of the two compensation forces from top to bottom. The left channel takes the loading force tracking error as input, multiplies it by the corrected test specimen damping feedback gain parameter (a wavy texture box, symbolizing the nonlinear characteristics of viscoelastic materials), and outputs the test specimen damping compensation force component. The right channel combines the outputs of displacement tracking error and equivalent stiffness (a spring icon box), multiplies it by the corrected hydraulic system damping feedback gain parameter (a dot-matrix texture box, symbolizing the discrete resistance characteristics of oil flow), and outputs the hydraulic system damping compensation force component. The two compensation forces converge at the bottom via a plus circle and are finally superimposed and output to the servo valve opening command. The core value of this dual-channel separate calculation architecture lies in matching the physical response mechanisms of the test specimen damping and the hydraulic system damping: the test specimen damping is mainly manifested as the phase difference between the applied force and the displacement, so the area deviation of the hysteresis loop can be directly corrected by using the force error as input; the hydraulic system damping is mainly manifested as the pressure loss caused by the piston speed, so the displacement error needs to be converted into velocity (by multiplying by the combined effect of stiffness and frequency) before the damping force can be accurately calculated. Traditional control algorithms combine the damping on both sides into a single feedback gain, ignoring the difference in their response mechanisms, resulting in compensation mismatch when the test specimen stiffness changes over time or the hydraulic oil temperature fluctuates. This invention, by calculating separately in each channel and superimposing the results in the final stage, ensures that the compensation force on each side acts precisely on the corresponding physical mechanism, avoiding mutual interference between channels, and providing a divide-and-conquer, combined-use algorithm architecture to support the high-precision damping closed-loop adaptive control in step S4.

[0091] S42: Determine whether the transient process has ended within each servo valve control cycle.

[0092] Further, step S42 includes the following steps: S421: Calculate the loading force tracking error and displacement tracking error for the current servo valve control cycle.

[0093] The loading force tracking error is the ratio of the absolute value of the deviation between the measured loading force and the target working condition command force value to the peak value of the target working condition command force amplitude. The displacement tracking error is the ratio of the absolute value of the deviation between the measured displacement and the target working condition command displacement value to the peak value of the target working condition command displacement amplitude. The peak values ​​of the target working condition command force amplitude and displacement amplitude are respectively taken from the peak values ​​of the loading force amplitude and displacement amplitude of the corresponding working condition in the subsequent working condition loading parameter table. Using the peak amplitude as the normalization denominator instead of the instantaneous command value is to avoid the division by zero problem when the command signal passes through zero point during dynamic loading, and at the same time to ensure that the physical meaning of the tracking error is the error ratio relative to the full-scale amplitude. The measured loading force and displacement values ​​are provided by the sampled values ​​at the current moment in the transient loading force time series data and transient displacement time series data synchronously collected in step S22. The target working condition command force value and target working condition command displacement value are determined by the value of the loading command signal corresponding to the subsequent working condition in the preset subsequent working condition loading parameter table at the current moment.

[0094] S422: Determine whether the loading force tracking error is less than the preset force steady-state convergence threshold and whether the displacement tracking error is less than the preset displacement steady-state convergence threshold.

[0095] The preset force steady-state convergence threshold is determined according to the test procedure's requirements for loading force accuracy. For example, the force steady-state convergence threshold can be set to 0.01, meaning the loading force tracking error is less than 1%. The preset displacement steady-state convergence threshold is determined according to the test procedure's requirements for displacement accuracy. For example, the displacement steady-state convergence threshold can be set to 0.005, meaning the displacement tracking error is less than 0.5%. If the loading force tracking error is less than the force steady-state convergence threshold and the displacement tracking error is less than the displacement steady-state convergence threshold, the continuous count is incremented by 1. If the continuous count reaches the preset steady-state judgment window length, the transient process is judged to end. The preset steady-state judgment window length is determined according to the target loading frequency, specifically, it is the integer half of the quotient of the target loading frequency corresponding to the period divided by the servo valve control period, to ensure that the loading force tracking error and displacement tracking error continuously meet the convergence condition for at least half a loading period before being judged as steady state, avoiding misjudgment due to brief error fluctuations during the transient process. For example, when the target loading frequency is 2 Hz and the servo valve control cycle is 1 millisecond, the steady-state determination window length is 500 divided by 1 divided by 2 rounded up to equal 250 servo valve control cycles. If either the loading force tracking error is not less than the force steady-state convergence threshold or the displacement tracking error is not less than the displacement steady-state convergence threshold, the continuous count will be reset to zero, and the damping closed-loop adaptive control in step S41 will continue to be executed.

[0096] S43: After the transient process ends, write the corrected damping feedback gain parameter set back to the current steady-state damping parameter set.

[0097] Further, step S43 includes the following steps: S431: Convert the modified test specimen damping feedback gain parameter in the modified damping feedback gain parameter group into the updated test specimen equivalent damping ratio, and convert the modified hydraulic system damping feedback gain parameter into the updated hydraulic system vertical channel oil damping coefficient and the updated hydraulic system horizontal channel oil damping coefficient.

[0098] The method for converting the modified test piece damping feedback gain parameter into the updated equivalent damping ratio of the test piece is as follows: the value of the modified test piece damping feedback gain parameter is equal to the updated equivalent damping ratio of the test piece multiplied by the square root of the product of twice the equivalent stiffness of the test piece and the mass of the test piece. Therefore, the updated equivalent damping ratio of the test piece is equal to the modified test piece damping feedback gain parameter divided by the square root of the product of twice the equivalent stiffness of the test piece and the mass of the test piece. The method for converting the modified hydraulic system damping feedback gain parameter into the updated hydraulic system vertical channel oil damping coefficient and the updated hydraulic system horizontal channel oil damping coefficient is as follows: based on the ratio of the original vertical channel oil damping coefficient and horizontal channel oil damping coefficient in the current steady-state damping parameter set as the allocation weight, the modified hydraulic system damping feedback gain parameter is allocated to the vertical channel and horizontal channel according to the ratio, thereby obtaining the updated hydraulic system vertical channel oil damping coefficient and the updated hydraulic system horizontal channel oil damping coefficient, respectively. The specific calculation method is as follows: The updated hydraulic system vertical channel oil damping coefficient is equal to the corrected hydraulic system damping feedback gain parameter multiplied by the hydraulic system vertical channel oil damping coefficient divided by the sum of the hydraulic system vertical channel oil damping coefficient and the hydraulic system horizontal channel oil damping coefficient. The updated hydraulic system horizontal channel oil damping coefficient is equal to the corrected hydraulic system damping feedback gain parameter multiplied by the hydraulic system horizontal channel oil damping coefficient divided by the sum of the hydraulic system vertical channel oil damping coefficient and the hydraulic system horizontal channel oil damping coefficient.

[0099] S432: Replace the equivalent damping ratio of the test specimen in the current steady-state damping parameter set with the updated equivalent damping ratio of the test specimen; replace the hydraulic system vertical channel oil damping coefficient in the current steady-state damping parameter set with the updated hydraulic system vertical channel oil damping coefficient; replace the hydraulic system horizontal channel oil damping coefficient in the current steady-state damping parameter set with the updated hydraulic system horizontal channel oil damping coefficient; keep the equivalent stiffness of the test specimen unchanged; complete the update of the current steady-state damping parameter set; and use the updated current steady-state damping parameter set as the input for calculating the transient damping demand prediction vector for the working condition switching in step S13 during subsequent working condition switching; thus completing the transient damping adaptive control of the multi-working-condition continuous dynamic compression-shear test.

[0100] The reason for keeping the equivalent stiffness of the test specimen constant is that the equivalent stiffness of the test specimen is mainly determined by the geometric dimensions of the test specimen and the elastic modulus of the material. During a single multi-condition continuous dynamic compression and shear test, the geometric dimensions of the test specimen do not change, and the elastic modulus of the material changes very little within the elastic range. Therefore, the equivalent stiffness of the test specimen can be regarded as constant before and after the transient process.

[0101] Step S4 applies the corrected damping feedback gain parameter set output from step S3 to the real-time closed-loop control of the servo valve control circuit. By calculating the damping compensation force component of the test piece and the damping compensation force component of the hydraulic system in each servo valve control cycle and superimposing them onto the force command signal, real-time adjustment of the servo valve output of the vertical loading cylinder and the horizontal shear actuator is achieved. The transient process termination determination in step S4 adopts a continuous counting mechanism, requiring that the loading force tracking error and displacement tracking error simultaneously and continuously meet the convergence condition and reach the preset steady-state determination window length before determining the transient process termination. This avoids premature exit from damping adaptive control due to a single accidental error fallback. After the transient process ends, step S4 writes the corrected damping feedback gain parameter set back to the current steady-state damping parameter set, ensuring that the damping correction results accumulated during the transient process are not lost. Instead, they serve as the update baseline for calculating the transient damping demand prediction vector for the subsequent working condition switch in step S13, forming a parameter iteration closed loop from step S1 to step S4 and back to step S1. In multi-condition continuous dynamic compression-shear tests, as the conditions change successively, the updated parameters written back after each transient process continuously correct the accuracy of the steady-state damping parameter set for the current condition. This makes the damping demand prediction in step S1 during subsequent condition changes increasingly accurate, the feedforward compensation deviation in step S2 smaller, and the residual amount that needs correction in step S3 less. The duration of the transient process shortens successively, and the peak loading error decreases successively, achieving a positive feedback effect where the damping adaptive control accuracy increases with the test progress. This positive feedback effect fundamentally solves the inefficient practice of using a fixed waiting time for transient processes during condition changes in the industry. Because the convergence speed of the transient process continuously accelerates with the continuous optimization and updating of damping parameters, there is no need to set a conservative fixed waiting time to ensure data validity. At the same time, since the loading force and displacement tracking errors during the transient process are suppressed by both feedforward compensation and closed-loop correction, the amplitude of the non-preset alternating load borne by the test piece at the moment of condition change is significantly reduced, reducing the accumulation of unexpected damage to the internal structure of the test piece and ensuring the accuracy and repeatability of the mechanical performance test results under subsequent steady-state conditions.

[0102] Example 2: This embodiment, based on Embodiment 1, provides a damping adaptive control system for a dynamic compression-shear testing machine, such as... Figure 10 As shown, it includes: Parameter identification and prediction module: During the steady-state loading stage, the hydraulic system pressure time series data, actuator displacement time series data and loading force time series data are collected simultaneously, and the steady-state damping parameter set of the current working condition is obtained through online recursive identification; and combined with the preset subsequent working condition loading parameter table, the transient damping demand prediction vector for working condition switching is calculated. Feedforward compensation and residual extraction module: Based on the predicted vector of transient damping demand during the working condition switching, a transient damping compensation feedforward control sequence is generated; at the execution time of the working condition switching, the servo valve control loop is injected, and the transient loading force time series data and transient displacement time series data during the working condition switching process are collected synchronously. The residual damping effect that the feedforward compensation fails to cover is extracted by comparison and calculation to obtain the transient damping residual time series data. Feedback gain correction module: Extracts features from the transient damping residual time series data to obtain a transient residual feature sequence; performs correlation calculation between the transient residual feature sequence and the current operating condition steady-state damping parameter set to generate a damping correction increment lookup table; performs real-time correction of the damping feedback gain parameters in the servo valve control loop based on the damping correction increment lookup table to obtain a corrected damping feedback gain parameter set; Closed-loop control and parameter update module: Based on the corrected damping feedback gain parameters, the module performs damped closed-loop adaptive control on the servo valve control loop, and after the transient process ends, it writes back the corrected damping parameters to update the current steady-state damping parameter set.

Claims

1. A damping adaptive control method for a dynamic compression-shear testing machine, characterized in that, The method includes: S1: During the steady-state loading phase, the hydraulic system pressure time-series data, actuator displacement time-series data, and loading force time-series data are collected synchronously to obtain the current working condition steady-state damping parameter set through online recursive identification; and combined with the preset subsequent working condition loading parameter table, the transient damping demand prediction vector for working condition switching is calculated. S2: Based on the predicted vector of transient damping demand during the working condition switching, a transient damping compensation feedforward control sequence is generated; the servo valve control loop is injected at the time of working condition switching execution, and the transient loading force time series data and transient displacement time series data during the working condition switching process are collected synchronously. The residual damping effect that the feedforward compensation fails to cover is extracted by comparison and calculation to obtain the transient damping residual time series data. S3: Extract features from the transient damping residual time series data to obtain a transient residual feature sequence; perform correlation calculation between the transient residual feature sequence and the current operating condition steady-state damping parameter set to generate a damping correction increment lookup table; perform real-time correction of the damping feedback gain parameters in the servo valve control loop based on the damping correction increment lookup table to obtain the corrected damping feedback gain parameter set. S4: Perform damped closed-loop adaptive control on the servo valve control loop based on the corrected damping feedback gain parameters, and write back the corrected damping parameters to update the current steady-state damping parameter set after the transient process ends.

2. The damping adaptive control method for a dynamic compression-shear testing machine according to claim 1, characterized in that, The calculation vector for predicting transient damping demand during operating condition switching includes: During the steady-state loading phase of the dynamic compression-shear testing machine, hydraulic system pressure time-series data, actuator displacement time-series data, and loading force time-series data are collected synchronously at a preset sampling period. Online recursive identification is performed on the hydraulic system pressure timing data, the actuator displacement timing data, and the loading force timing data to obtain the current working condition steady-state damping parameter set; Based on the current steady-state damping parameter set and the preset subsequent loading parameter table, the transient damping demand prediction vector for the change of operating conditions is calculated.

3. The damping adaptive control method for a dynamic compression-shear testing machine according to claim 2, characterized in that, The obtained steady-state damping parameter set for the current operating condition includes: The actuator displacement timing data is used as the system excitation input, and the loading force timing data is used as the system response output. The recursive least squares method is used to identify the force-displacement hysteresis relationship of the test specimen online, and the equivalent stiffness and equivalent damping ratio of the test specimen are obtained. Based on the pressure difference time series data of the two chambers of the vertical loading cylinder and the pressure difference time series data of the two chambers of the horizontal shear actuator in the hydraulic system pressure time series data, and combined with the velocity components in the corresponding directions in the actuator displacement time series data, the hydraulic damping coefficient of the vertical channel and the hydraulic damping coefficient of the horizontal channel of the hydraulic system are calculated. The equivalent stiffness of the test specimen, the equivalent damping ratio of the test specimen, the hydraulic damping coefficient of the vertical channel of the hydraulic system, and the hydraulic damping coefficient of the horizontal channel of the hydraulic system are combined to form the steady-state damping parameter set for the current operating condition.

4. The damping adaptive control method for a dynamic compression-shear testing machine according to claim 2, characterized in that, The obtained transient damping demand prediction vector for operating condition switching includes: Read the target loading frequency, target loading amplitude, and target loading direction of the subsequent working conditions from the preset subsequent working condition loading parameter table; Based on the equivalent stiffness and equivalent damping ratio of the test specimen in the current steady-state damping parameter set, and combined with the target loading frequency and target loading amplitude of the subsequent working conditions, the load mutation amount is calculated. Based on the load mutation amount, the pre-calibrated test specimen damping nonlinear mutation characteristic table is consulted to obtain the estimated increment of transient damping of the test specimen. Based on the hydraulic system vertical channel oil damping coefficient and hydraulic system horizontal channel oil damping coefficient in the current steady-state damping parameter set, combined with the target loading frequency and target loading amplitude of the subsequent working conditions, the estimated value of the transient fluctuation amplitude of oil pressure at the moment of working condition switching is calculated, and the estimated value of transient additional damping of hydraulic system is obtained. The transient damping prediction increment of the test piece, the transient additional damping prediction of the hydraulic system, and the target loading frequency, target loading amplitude, and target loading direction of the subsequent working conditions are combined into a transient damping demand prediction vector for working condition switching.

5. The damping adaptive control method for a dynamic compression-shear testing machine according to claim 1, characterized in that, The obtained transient damping residual time-series data includes: Based on the servo valve opening correction amount at each time step in the transient damping compensation feedforward control sequence, calculate the expected compensation force increment sequence at each time step. Calculate the actual change in loading force at each time step, and subtract the expected compensation force increment and the target working condition command force increment at the corresponding time step from the actual change in loading force to obtain the force residual value at each time step. Calculate the actual displacement change at each time step, and subtract the target working condition command displacement increment from the actual displacement change to obtain the displacement residual value at each time step. The force residuals and displacement residuals at each time step are arranged in chronological order to form the transient damping residual time series data.

6. The damping adaptive control method for a dynamic compression-shear testing machine according to claim 5, characterized in that, The step of generating the transient damping compensated feedforward control sequence includes: Extract the estimated incremental transient damping of the test specimen and the estimated transient additional damping of the hydraulic system from the transient damping demand prediction vector during the switching of working conditions, and sum them to obtain the estimated total transient additional damping. The transient compensation time window length is determined based on the estimated total transient additional damping and the target loading frequency in the predicted vector of transient damping demand during operating condition switching. Within the transient compensation time window, with the servo valve control cycle as the time step, the estimated total transient additional damping is discretized in the time domain according to the exponential decay envelope to generate the transient damping compensation amount corresponding to each time step. The transient damping compensation amount corresponding to each time step is converted into the corresponding servo valve opening correction amount, and arranged in time order to form a transient damping compensation feedforward control sequence.

7. The damping adaptive control method for a dynamic compression-shear testing machine according to claim 5, characterized in that, The steps for obtaining the transient damping residual time series data include: Based on the servo valve opening correction amount at each time step in the transient damping compensation feedforward control sequence, calculate the expected compensation force increment sequence at each time step. Calculate the actual load force change at each time step based on the transient load force time series data, and subtract the expected compensation force increment and the target working condition command force increment at the corresponding time step from the actual load force change to obtain the force residual value at each time step. The actual displacement change at each time step is calculated based on the transient displacement time series data. The displacement residual value at each time step is obtained by subtracting the target working condition command displacement increment from the actual displacement change. The force residuals and displacement residuals at each time step are arranged in chronological order to form the transient damping residual time series data.

8. The damping adaptive control method for a dynamic compression-shear testing machine according to claim 1, characterized in that, The step of generating the damping correction increment lookup table includes: For each transient residual eigenvector in the transient residual characteristic sequence, the ratio of the root mean square value of the force residual to the equivalent damping ratio of the current steady-state damping parameter ensemble test piece is calculated to obtain the damping identification deviation ratio. The ratio of the root mean square value of the displacement residual to the average value of the servo valve opening correction amount in the corresponding time period of the transient damping compensation feedforward control sequence is calculated to obtain the compensation response deviation ratio. The damping correction increment of the test piece is determined based on the damping identification deviation ratio, and the damping correction increment of the hydraulic system is determined based on the compensation response deviation ratio. Using the peak-to-peak value of the force residual and the maximum rate of change of the displacement residual in the transient residual eigenvector as lookup indices, and the corresponding damping correction increment of the test specimen and the damping correction increment of the hydraulic system as lookup outputs, a damping correction increment lookup table is constructed.

9. The damping adaptive control method for a dynamic compression-shear testing machine according to claim 1, characterized in that, The write-back to the current operating condition steady-state damping parameter set includes: Replace the damping feedback gain parameters of the corresponding channels in the servo valve control circuit with the modified test piece damping feedback gain parameters and the modified hydraulic system damping feedback gain parameters in the modified damping feedback gain parameter group, and adjust the servo valve output of the vertical loading cylinder and the horizontal shear actuator in real time. Determine whether the transient process has ended within each servo valve control cycle; After the transient process ends, the corrected damping feedback gain parameter set is written back to the current steady-state damping parameter set.

10. The damping adaptive control method for a dynamic compression-shear testing machine according to claim 9, characterized in that, The steps for determining whether the transient process has ended include: Calculate the loading force tracking error and displacement tracking error for the current servo valve control cycle; If the loading force tracking error is less than the preset force steady-state convergence threshold and the displacement tracking error is less than the preset displacement steady-state convergence threshold, the continuous count is incremented by 1. If the continuous count reaches the preset steady-state determination window length, the transient determination process ends; otherwise, the continuous count is reset to zero.

11. A damping adaptive control system for a dynamic compression-shear testing machine, used to implement the damping adaptive control method for a dynamic compression-shear testing machine according to any one of claims 1-10, characterized in that, The system includes: Parameter identification and prediction module: During the steady-state loading stage, the hydraulic system pressure time series data, actuator displacement time series data and loading force time series data are collected simultaneously, and the steady-state damping parameter set of the current working condition is obtained through online recursive identification; and combined with the preset subsequent working condition loading parameter table, the transient damping demand prediction vector for working condition switching is calculated. Feedforward compensation and residual extraction module: Based on the predicted vector of transient damping demand during the working condition switching, a transient damping compensation feedforward control sequence is generated; at the execution time of the working condition switching, the servo valve control loop is injected, and the transient loading force time series data and transient displacement time series data during the working condition switching process are collected synchronously. The residual damping effect that the feedforward compensation fails to cover is extracted by comparison and calculation to obtain the transient damping residual time series data. Feedback gain correction module: Extracts features from the transient damping residual time series data to obtain a transient residual feature sequence; performs correlation calculation between the transient residual feature sequence and the current operating condition steady-state damping parameter set to generate a damping correction increment lookup table; performs real-time correction of the damping feedback gain parameters in the servo valve control loop based on the damping correction increment lookup table to obtain a corrected damping feedback gain parameter set; Closed-loop control and parameter update module: Based on the corrected damping feedback gain parameters, the module performs damped closed-loop adaptive control on the servo valve control loop, and after the transient process ends, it writes back the corrected damping parameters to update the current steady-state damping parameter set.

Images