A vibration energy dissipation feedback control system for core wall liquefaction resistance

Abstract

This invention relates to the fields of seismic resistance in geotechnical engineering and construction technology in hydraulic engineering, and particularly to a vibration energy dissipation feedback control system for core wall anti-liquefaction. The system includes: a data acquisition and positioning module that acquires vibration response signals and location information in real time during the compaction process; a signal processing and feature extraction module that extracts characteristic parameters representing the dynamic state of the fill material; a real-time liquefaction risk diagnosis module that outputs a quantified liquefaction risk index using a pre-trained discriminant model; an intelligent decision-making and control command generation module that generates control commands based on this index; an actuator adaptive adjustment and targeted compaction module that drives the roller to adjust its working parameters and performs targeted enhanced compaction in high-risk areas; and a global iteration and visualization monitoring module that coordinates the entire process and generates a quality risk distribution map. This invention achieves real-time perception and adaptive closed-loop control of anti-liquefaction performance during construction, fundamentally improving the seismic stability of the core wall.

Description

Technical Field

[0001] This invention relates to the field of geotechnical engineering earthquake resistance and hydraulic engineering construction technology, and in particular to a vibration energy dissipation feedback control system for core wall anti-liquefaction. Background Technology

[0002] As a key structure in hydraulic engineering, the compaction quality of the core wall of an earth-rock dam directly affects the long-term safety and seismic stability of the dam body. Under seismic loads, liquefaction of saturated fill material is one of the main risks leading to shear failure of the core wall and even dam instability. Therefore, taking effective measures to enhance the liquefaction resistance of the fill material during the core wall construction stage is an important way to ensure the seismic safety of high earth-rock dams from the root.

[0003] Currently, in core wall construction, compaction quality control mainly relies on post-construction inspection of the physical state of the compacted fill material, such as by sampling and measuring static indicators like dry density and void ratio. However, this conventional method has a significant technical problem: it completely lacks the ability to perceive and control the potential liquefaction resistance of the fill material under dynamic loads in real time. During construction, it is impossible to immediately identify "hidden" high-risk points with weak liquefaction resistance within the compaction area; only speculative evaluations can be made through limited, discrete tests after the filling is completed. This results in "blind spots" in the construction process, making it difficult to proactively and accurately intervene and eliminate liquefaction risks during compaction. Consequently, the core wall structure still faces the potential risk of shear failure due to localized liquefaction when encountering strong earthquakes. Summary of the Invention

[0004] To overcome the above shortcomings, this invention provides a vibration energy dissipation feedback control system for core wall anti-liquefaction, which aims to improve the problem that existing compaction control methods cannot detect and actively suppress the potential liquefaction risk of filler in real time, thus causing the seismic safety of the structure to rely on post-event inspection and have potential hidden dangers at the source.

[0005] This invention provides the following technical solution: a vibration energy dissipation feedback control system for core wall anti-liquefaction, comprising:

[0006] The data acquisition and positioning module is used to acquire, in real time, the vibration response signal generated by the interaction between the vibratory roller and the filling material when the vibratory roller is compacting the core wall filling material of the earth-rock dam, and to obtain the real-time position information of the vibratory roller.

[0007] The signal processing and feature extraction module is used to perform time-frequency domain transformation processing on the vibration response signal and extract feature parameters from the transformation result to characterize the dynamic state of the packing.

[0008] The real-time liquefaction risk diagnosis module is used to input the feature parameters into the pre-established liquefaction risk discrimination model and calculate the liquefaction risk index of the filler at the current rolling position.

[0009] The intelligent decision-making and control instruction generation module is used to generate real-time control instructions for the operating parameters of the vibratory roller based on the liquefaction risk index.

[0010] The actuator adaptive adjustment and targeted compaction module is used to adaptively adjust the excitation frequency, excitation force amplitude and travel speed of the vibratory roller according to the control command, so as to perform targeted compaction at the current compaction position that matches the liquefaction risk index;

[0011] The global iteration and visualization monitoring module is used to control the system to repeatedly execute the functions of the data acquisition and positioning module and the actuator adaptive adjustment and targeted compaction module until the full compaction of the core wall filling area is completed. Based on the real-time location information and the liquefaction risk index, it generates and updates the compaction quality and liquefaction risk distribution map of the core wall filling area.

[0012] Preferably, the workflow of the data acquisition and positioning module includes:

[0013] The vibration acceleration signals perpendicular to the rolling surface, parallel to the rolling direction, and perpendicular to the rolling direction are collected by a triaxial acceleration sensor installed at the bearing seat of the vibratory roller.

[0014] The planar coordinates and elevation information of the vibratory roller are obtained by a global navigation satellite system receiver installed on the frame of the vibratory roller, which serves as the real-time location information.

[0015] The vibration acceleration signal is synchronized with the real-time location information to form a vibration response signal dataset with spatial location identification.

[0016] Preferably, the workflow of the signal processing and feature extraction module includes:

[0017] The vibration response signal is preprocessed with bandpass filtering and noise reduction.

[0018] The preprocessed vibration response signal is subjected to a short-time Fourier transform to obtain the time spectrum of the signal.

[0019] Calculate the main frequency energy ratio, high frequency dissipation coefficient, and energy uniformity from the time-frequency spectrum.

[0020] Preferably, the workflow of the real-time liquefaction risk diagnosis module includes:

[0021] The main frequency energy ratio, high frequency dissipation coefficient and energy uniformity are normalized to form a standardized feature vector.

[0022] The standardized feature vector is input into the pre-established liquefaction risk discrimination model;

[0023] Based on the standardized feature vector, the liquefaction risk discrimination model outputs a quantified liquefaction risk index.

[0024] The liquefaction risk discrimination model is obtained by machine learning training based on historical test data of the core wall packing. The historical test data includes the dominant frequency energy ratio, high frequency dissipation coefficient, energy uniformity and their corresponding liquefaction resistance calibration values ​​measured under different compaction conditions.

[0025] Preferably, the workflow of the intelligent decision-making and control instruction generation module includes:

[0026] Based on the preset risk level range where the liquefaction risk index is located, the target compaction mode is determined. The risk level range includes a low-risk range, a medium-risk range, and a high-risk range, which correspond to the standard compaction mode, the reinforced compaction mode, and the repair compaction mode, respectively.

[0027] Based on the target compaction mode, a reference command is generated for the excitation frequency setting value, excitation force amplitude setting value, and travel speed setting value of the vibratory roller through a preset fuzzy reasoning rule base.

[0028] Based on the real-time changing trend of the liquefaction risk index, the baseline command is adjusted proportionally, integrally, and derivatively to ultimately generate the real-time control command.

[0029] Preferably, the operating process of the actuator adaptive adjustment and targeted compaction module includes:

[0030] The controller of the vibratory roller receives the real-time control command;

[0031] The controller generates corresponding drive signals based on the excitation frequency setting, excitation force amplitude setting, and travel speed setting in the real-time control command.

[0032] The drive signal is sent to the frequency converter, excitation force regulating valve and travel motor of the vibratory roller to synchronously adjust the excitation frequency, excitation force amplitude and travel speed of the vibratory roller.

[0033] The current compaction position is compacted at least once using the adjusted working parameters to complete the targeted compaction.

[0034] Preferably, the workflow of the global iteration and visualization monitoring module includes:

[0035] During the continuous compaction operation of the vibratory roller on the core wall filling area, the functions of the data acquisition and positioning module, signal processing and feature extraction module, liquefaction risk real-time diagnosis module, intelligent decision-making and control command generation module, and actuator adaptive adjustment and targeted compaction module are cyclically called and executed.

[0036] The real-time location information and corresponding liquefaction risk index obtained from each loop execution are merged with historical data and stored in a preset regional database;

[0037] Based on the liquefaction risk index of all locations in the regional database, a spatial interpolation algorithm is used to generate and continuously update a compaction quality and liquefaction risk distribution map covering the entire core wall filling area.

[0038] The distribution map is visualized, and areas where the liquefaction risk index is consistently higher than a preset threshold are marked and alerted.

[0039] The present invention has the following beneficial effects:

[0040] 1. In this invention, by collecting vibration response signals and extracting dynamic characteristic parameters in real time during the compaction process, a liquefaction risk index related to the liquefaction resistance strength of the filler is directly constructed. This enables the real-time identification of areas with weak liquefaction resistance during the construction phase, significantly shifting the control point of liquefaction risk from after completion to the construction process, thus achieving a fundamental shift from passive acceptance to proactive defense.

[0041] 2. In this invention, based on the real-time diagnosed liquefaction risk, the optimal control command is automatically generated through an intelligent decision-making algorithm, and the road roller is driven to perform targeted enhanced compaction with variable frequency, variable amplitude, and variable speed. While ensuring overall quality, the uniformity and reliability of the compaction effect are significantly improved, and the integrity of the core wall structure is optimized at the microscale.

[0042] 3. In this invention, by generating and updating the compaction quality and liquefaction risk distribution map of the entire dam surface, the construction process is fully digitized and visualized. The implicit quality status during construction is transformed into explicit, traceable, and analyzable spatial data, providing an irreplaceable data foundation for subsequent construction quality assessment, operational safety monitoring, and earthquake damage emergency assessment, greatly improving the precision and scientific level of project management. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of the architecture of a vibration energy dissipation feedback control system for core wall anti-liquefaction proposed in this invention. Detailed Implementation

[0044] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] In a first embodiment of the present invention, the present invention provides an excitation energy dissipation feedback control system for core wall anti-liquefaction, such as... Figure 1 As shown, it includes the following modules:

[0046] The data acquisition and positioning module is used to collect the vibration response signal generated by the interaction between the vibratory roller's excitation wheel and the filling material in real time when the vibratory roller is compacting the core wall filling material of the earth-rock dam, and to obtain the real-time position information of the vibratory roller.

[0047] Furthermore, the workflow of the data acquisition and positioning module includes:

[0048] The vibration acceleration signals perpendicular to the rolling surface, parallel to the rolling direction, and perpendicular to the rolling direction are collected by a triaxial accelerometer installed at the bearing seat of the vibratory roller.

[0049] The planar coordinates and elevation information of the vibratory roller are obtained by a global navigation satellite system receiver installed on the frame of the vibratory roller, which serves as its real-time location information;

[0050] The vibration acceleration signal is synchronized with the real-time location information to form a vibration response signal dataset with spatial location identification.

[0051] Specifically, to achieve comprehensive sensing of the three-dimensional dynamic response of the interaction between the vibratory roller and the packing material, an industrial-grade triaxial accelerometer is rigidly installed at the bearing housing of the vibratory roller. The sensor's three sensitive axes are oriented as follows: Z-axis: perpendicular to the compaction plane and upwards, used to collect vertical vibration acceleration information. X-axis: parallel to the direction of travel of the road roller, used to collect longitudinal vibration acceleration signals. The Y-axis is perpendicular to the XZ plane, i.e., the horizontal direction perpendicular to the direction of travel, and is used to acquire lateral vibration acceleration signals. The sensor sampling frequency is no less than 2000Hz to fully capture the high-frequency components within the operating frequency band of the excitation wheel. The acquired raw acceleration signal... This is the original vibration response signal;

[0052] To accurately correlate vibration data with spatial location, a high-precision real-time dynamic global navigation satellite system receiver is installed at the center of the vibratory roller's frame. This receiver operates continuously during compaction, outputting the planar coordinates and elevation information of the roller's compaction wheel center at an update frequency of no less than 10Hz. This information constitutes the real-time location information for the compaction operation. ,in, East coordinates, North coordinates, Elevation;

[0053] To ensure strict temporal correspondence between vibration signals and position information, a combination of hardware triggering and high-precision time synchronization is employed. The specific process is as follows: First, a unified high-precision time synchronization module is configured for both the triaxial accelerometer and the GNSS receiver to ensure internal clock synchronization. Second, in the controller of the data acquisition unit, a time synchronization module is used for each frame of GNSS position data. Stamp the precise timestamp Finally, based on this timestamp From the synchronously sampled acceleration signal sequence a(t), data segments within the corresponding time window are extracted. Through the above processing, the vibration acceleration signal a(t) is bound to its corresponding spatial location information P(t), forming a vibration response signal data record with spatial location identification. All such records generated during continuous operation together constitute the vibration response signal dataset for subsequent analysis. Each piece of data has a clear spatiotemporal attribute;

[0054] The acquisition and positioning module, through hardware integration and spatiotemporal synchronization processing, achieves precise and synchronous capture of the dynamic state and spatial location information of the compaction process. It transforms traditional discrete, a posteriori compaction quality inspection into a continuous, real-time spatialized process sensing system, providing a unique and reliable data foundation for subsequent location-based precise positioning, risk diagnosis, and targeted control. This ensures the accuracy and traceability of the input information for the entire intelligent feedback control closed loop.

[0055] The signal processing and feature extraction module is used to perform time-frequency domain transformation on the vibration response signal and extract feature parameters from the transformation results to characterize the dynamic state of the packing.

[0056] Furthermore, the workflow of the signal processing and feature extraction module includes:

[0057] Bandpass filtering and noise reduction preprocessing are performed on the vibration response signal;

[0058] The preprocessed vibration response signal is subjected to a short-time Fourier transform to obtain the time spectrum of the signal.

[0059] Calculate the main frequency energy ratio, high-frequency dissipation coefficient, and energy uniformity from the time-frequency spectrum.

[0060] Specifically, firstly, the raw vibration acceleration signal from the data acquisition module is preprocessed to eliminate interference and focus on the effective frequency band. A zero-phase digital bandpass filter with a passband frequency of 10Hz to 500Hz is used to filter the signal, covering the main components of the roller's excitation energy and soil response. Simultaneously, wavelet thresholding is used to smooth the signal to suppress high-frequency random noise.

[0061] The preprocessed vibration acceleration signal is analyzed using the short-time Fourier transform method to obtain the signal energy distribution in the two-dimensional plane of time and frequency, i.e., the time spectrum. Specifically, the acceleration signal sequence a[n] is multiplied by a sliding time window function w[n−m], where m is the sliding index of the time window. A discrete Fourier transform is then performed on each windowed signal segment to obtain its time spectrum S[m,k].

[0062] ;

[0063] Where k is the frequency index and N is the number of Fourier transform points. The window function w is a Hanning window, with a window length set to include 1024 sampling points and an overlap rate of 50%. The final result is the energy distribution of the signal in the time-frequency plane. Real-time spectrum;

[0064] From the time-frequency spectrum P[m,k], calculate three core characteristic parameters, specifically, the main frequency energy ratio. The calculation measures the ratio of the energy concentrated in the main frequency band to the total energy across the entire frequency band.

[0065] ;

[0066] Among them, the main frequency band The corresponding frequency index range is The frequency peak value of P[m,k] is determined by finding the frequency peak value and taking its ±5Hz bandwidth.

[0067] High-frequency dissipation coefficient It calculates the average decay rate of signal energy over time within the high-frequency band of 150Hz to 400Hz, to characterize the energy dissipation characteristics:

[0068] ;

[0069] in, M represents the frequency index range corresponding to the high-frequency band, M represents the total number of time windows, and Δt represents the time corresponding to the time window moving step.

[0070] Energy uniformity is calculated as the reciprocal of the coefficient of variation of the energy values ​​at all points in the time-frequency plane of the spectrum P[m,k], and is used to characterize the uniformity of energy distribution.

[0071] ;

[0072] in, Let be the mean of all elements in P[m,k]. Its standard deviation;

[0073] Finally, the output is a feature vector composed of these three parameters. ;

[0074] The signal processing and feature extraction module achieves in-depth perception and index transformation of the dynamic state of packing compaction by analyzing the time-frequency domain of vibration signals and extracting quantitative features. It can refine the raw, mixed time-domain vibration waveforms into spectral feature vectors with clear physical meaning. These features are directly related to key dynamic properties of the packing, such as stiffness, damping, and uniformity, thus providing high signal-to-noise ratio and quantifiable core input data for subsequent intelligent diagnosis of liquefaction risks.

[0075] The real-time liquefaction risk diagnosis module is used to input characteristic parameters into a pre-established liquefaction risk discrimination model to calculate the liquefaction risk index of the filler at the current compaction position.

[0076] Furthermore, the workflow of the real-time liquefaction risk diagnosis module includes:

[0077] The main frequency energy ratio, high frequency dissipation coefficient and energy uniformity are normalized to form a standardized feature vector;

[0078] The standardized feature vectors are input into a pre-established liquefaction risk discrimination model;

[0079] Based on standardized feature vectors, the liquefaction risk discrimination model outputs a quantified liquefaction risk index.

[0080] The liquefaction risk discrimination model is based on historical test data of the core wall packing and is trained by machine learning. The historical test data includes the dominant frequency energy ratio, high frequency dissipation coefficient, energy uniformity and their corresponding liquefaction resistance calibration values ​​measured under different compaction conditions.

[0081] Specifically, firstly, the feature vector output by the feature extraction module... Normalization is performed to eliminate the influence of differences in the dimensions and numerical ranges of various features on the model. The maximum-minimum normalization method is used.

[0082] ;

[0083] in, and These are the minimum and maximum value vectors of each feature component corresponding to the center wall filler in the training dataset across all samples. This process yields the standardized feature vector. Each of its component values ​​is located in the interval [0, 1].

[0084] The liquefaction risk assessment model is built based on a machine learning algorithm. Its training relies on a pre-established historical test database for a specific core wall packing. The database construction process is as follows: in the laboratory, samples of the core wall packing under different combinations of moisture content and dry density are prepared. For each sample, on the one hand, using the same sensor setup and signal processing methods as in the field, its characteristic parameters under standard excitation are simulated and measured. On the other hand, the liquefaction resistance of the specimen was determined by dynamic triaxial testing, i.e., the cyclic stress ratio required to initiate initial liquefaction, and recorded as the calibration value. Collecting a large number of such samples is crucial. This constitutes the training dataset. A support vector machine regression algorithm is used to train this dataset, with the feature parameter F as input and the liquefaction resistance strength calibration value CRR as the prediction target. The trained model M then possesses the ability to predict liquefaction resistance strength from dynamic features.

[0085] During the compaction operation, the standardized feature vectors obtained in real time are... The input is fed into the pre-trained model M. The model performs forward computation and outputs a predicted cyclic stress ratio. To obtain more intuitive risk assessment indicators, the predicted... This is converted into a dimensionless liquefaction risk index, LRI. The conversion formula is defined as:

[0086] ;

[0087] in, The LRILRI index is the liquefaction resistance threshold determined according to the seismic design standard for dams. Theoretically, the range of this index is (−∞, 1]. A higher LRILRI value indicates a lower safety margin for liquefaction resistance of the current fill material under the preset seismic load, and a higher risk of liquefaction. This index serves as a direct basis for subsequent intelligent decision-making.

[0088] The real-time liquefaction risk diagnosis module solidifies the physical relationship between dynamic spectrum characteristics and liquefaction resistance into an executable intelligent model, enabling online and quantitative assessment of the liquefaction resistance of fill material during construction. It transforms the liquefaction resistance strength, traditionally obtained only through complex laboratory tests, into a liquefaction risk index that can be calculated in real-time on the compaction site. This establishes a mapping path from "vibration signals during construction" to "core indicators of seismic performance," providing crucial quantitative judgment for the immediate identification and intervention of liquefaction risks during the filling stage.

[0089] The intelligent decision-making and control command generation module is used to generate real-time control commands for the operating parameters of the vibratory roller based on the liquefaction risk index.

[0090] Furthermore, the workflow of the intelligent decision-making and control command generation module includes:

[0091] Based on the preset risk level range where the liquefaction risk index is located, the target compaction mode is determined. The risk level range includes low risk range, medium risk range and high risk range, which correspond to standard compaction mode, reinforced compaction mode and repair compaction mode, respectively.

[0092] Based on the target compaction mode, the reference commands for the excitation frequency setting value, excitation force amplitude setting value and travel speed setting value of the vibratory roller are generated through the preset fuzzy reasoning rule base.

[0093] Based on the real-time changing trend of the liquefaction risk index, the baseline command is adjusted proportionally, integrally, and derivatively to ultimately generate a real-time control command.

[0094] Specifically, firstly, based on preset risk level thresholds, the liquefaction risk index (LRI) is mapped to a macro-operation mode. Two thresholds are set. and ,and ,like If it is determined to be low risk, the target compaction pattern is determined to be the standard compaction pattern. If so, it is judged as medium risk, and the target crushing mode is determined to be an enhanced crushing mode. If so, it is judged as high risk, and the target compaction mode is determined to be repair compaction mode;

[0095] For the three compaction modes mentioned above, a fuzzy inference rule base is pre-defined to generate baseline instructions. The input to this rule base is the compaction mode, and the output is a verbalized setting of the roller's excitation frequency f, excitation force amplitude A, and travel speed v. Example rules are as follows: IF mode IS standard mode, THEN f IS medium, A IS medium, v IS high; IF mode IS enhanced mode, THEN f IS high, A IS high, v IS medium; IF mode IS repair mode, THEN f IS very high, A IS very high, v IS low. The centroid method is used for defuzzification, converting the above verbal outputs into precise baseline settings: excitation frequency setting value. Excitation force amplitude setting value and travel speed setting value ;

[0096] To address changes in packing condition and achieve stable control, a PID controller is introduced to dynamically fine-tune the baseline command. Taking the excitation frequency as an example, the adjustment process is as follows, defining the control error. ,in The target risk index value is 0 in low-risk mode, and the output of the PID controller is the adjustment amount of the excitation frequency. Calculated by the following formula:

[0097] ;

[0098] in, , and These are the proportional, integral, and derivative coefficients, which are determined through on-site debugging.

[0099] Ultimately, the excitation frequency setpoint in the real-time control command... for:

[0100] ;

[0101] Similarly, the excitation force amplitude setpoint can be calculated. and travel speed setting value Therefore, the module outputs complete real-time control commands. ;

[0102] The intelligent decision-making and control command generation module, through a three-level decision-making architecture of "risk classification - fuzzy inference - PID fine-tuning," transforms a quantitative liquefaction risk index into precise and stable control commands for the road roller that adapt to different working conditions. This enables adaptive and refined control strategies. It ensures the rationality of compaction energy input under different risk levels through macroscopic pattern matching, and ensures a rapid and stable response to changes in the packing state through dynamic fine-tuning. This results in subsequent targeted compaction actions that are both strategically specific and precisely executed.

[0103] The actuator adaptive adjustment and targeted compaction module is used to adaptively adjust the excitation frequency, excitation force amplitude and travel speed of the vibratory roller according to the control command, so as to perform targeted compaction at the current compaction position that matches the liquefaction risk index.

[0104] Furthermore, the operating procedure of the actuator adaptive adjustment and targeted compaction module includes:

[0105] The controller of the vibratory roller receives real-time control commands;

[0106] The controller generates corresponding drive signals based on the excitation frequency setting, excitation force amplitude setting, and travel speed setting in the real-time control command.

[0107] The drive signal is sent to the frequency converter, excitation force regulating valve and travel motor of the vibratory roller to synchronously adjust the excitation frequency, excitation force amplitude and travel speed of the vibratory roller.

[0108] Perform at least one additional compaction pass on the current compaction location using the adjusted working parameters to complete the targeted compaction.

[0109] Specifically, the vibratory roller's onboard programmable logic controller receives real-time control commands sent by the intelligent decision-making module. The controller converts command values ​​into electrical signals that can directly drive the actuator based on its internally stored calibration parameters. Specifically, for the excitation frequency setpoint... The controller generates the corresponding analog voltage signal based on the preset f-V mapping relationship. The excitation force amplitude setpoint is sent to the frequency converter that drives the vibration motor. The controller generates the corresponding analog current signal based on the preset A-I mapping relationship. The proportional regulating valve in the hydraulic circuit controlling the eccentric block sends the setpoint for the travel speed. The controller generates the corresponding pulse width modulation signal based on the preset v-PWM mapping relationship. The signal is sent to the driving motor that drives the walking mechanism;

[0110] The aforementioned drive signals are simultaneously sent to each actuator. The frequency converter then adjusts the output based on the voltage signal. Adjusting the power supply frequency to the vibration motor changes the excitation frequency. The proportional control valve adjusts according to the current signal. Adjusting the hydraulic oil flow and pressure changes the centrifugal force of the vibrating eccentric block, i.e., the amplitude of the excitation force. The speed controller of the travel motor operates based on the pulse width modulation signal. The system adjusts the motor speed, thereby changing the roller's travel speed, ensuring that the adjustments of the three parameters are completed synchronously within a short period of time. The roller then operates with a new combination of operating parameters designed to deliver higher or more optimized compaction energy to identified high-risk areas.

[0111] The roller controller records the GNSS coordinates corresponding to the current high-risk location. Using adjusted operating parameters, the roller performs at least one additional pass-through compaction on a square area centered at these coordinates and with a preset side length. If the area is completely covered after one pass, one additional compaction pass is completed; if multiple passes are required for coverage, multiple compaction passes are performed until the entire target area has been fully applied at least once by the new compaction parameters. This process constitutes targeted compaction matching the liquefaction risk index. After completion, the system returns to the regular compaction process or waits to receive the next control command.

[0112] The actuator adaptive adjustment and targeted compaction module achieves a reliable closed loop from intelligent decision-making to physical execution through precise command-drive signal conversion and multi-actuator collaborative control. It can translate the optimized control strategy from the upper level into the actual actions of the road roller without delay or distortion. Through the coordinated adjustment of frequency conversion, amplitude conversion, and speed conversion, it applies enhanced energy input in a predetermined mode to high-risk areas, thus tightly linking "risk identification" and "risk management" in time and space, ensuring the practical implementation of liquefaction risk prevention and control measures.

[0113] The global iteration and visualization monitoring module is used to control the system to repeatedly execute the functions of the data acquisition and positioning module and the actuator adaptive adjustment and targeted compaction module until the full compaction of the core wall filling area is completed. Based on real-time location information and liquefaction risk index, it generates and updates the compaction quality and liquefaction risk distribution map of the core wall filling area.

[0114] Furthermore, the workflow of the global iteration and visualization monitoring module includes:

[0115] During the continuous compaction operation of the vibratory roller on the core wall filling area, the functions of the data acquisition and positioning module, signal processing and feature extraction module, liquefaction risk real-time diagnosis module, intelligent decision-making and control command generation module, and actuator adaptive adjustment and targeted compaction module are cyclically called and executed.

[0116] The real-time location information and corresponding liquefaction risk index obtained from each loop execution are merged with historical data and stored in a preset regional database;

[0117] Based on the liquefaction risk index of all locations in the regional database, a spatial interpolation algorithm is used to generate and continuously update a compaction quality and liquefaction risk distribution map covering the entire core wall filling area.

[0118] The distribution map is visualized, and areas where the liquefaction risk index consistently exceeds a preset threshold are marked and alerted.

[0119] Specifically, the global iteration and visualization monitoring module controls the core processing flow in a loop. During the compaction operation, triggered by each preset sampling interval or fixed time period as the advance of the compactor, the following module functions are sequentially called and executed: the data acquisition and positioning module acquires new data; the signal processing and feature extraction module calculates feature parameters; the liquefaction risk real-time diagnosis module outputs LRI; the intelligent decision-making and control instruction generation module judges and generates instructions; and the actuator adaptive adjustment and targeted compaction module executes the action. This loop continues until the compaction operation covers the entire core wall filling area.

[0120] In each loop, the obtained real-time location information and its corresponding liquefaction risk index To form a data record All records It is stored in real time in a pre-defined spatial database, namely a regional database. This database not only stores the current data, but also retains all relevant records of historical rolling passes, forming a complete time-space dataset;

[0121] To create a continuous spatial distribution view, a liquefaction risk index isosurface map of the entire operating area is generated using the Kriging spatial interpolation algorithm, based on all LRI data points in the regional database at a specific time. Kriging interpolation is an optimal unbiased estimation method for any unsampled points. The estimated value Known sample values Linear combination:

[0122] ;

[0123] Among them, the weighting coefficient Solving the Kriging equations reveals that this system minimizes the variance of the estimation error and satisfies the unbiased condition. This isosurface map is then rendered using a color gradient to obtain a real-time distribution map of compaction quality and liquefaction risk. This map is dynamically updated as new data is added.

[0124] The generated distribution map is displayed in real time on the touchscreen display in the cockpit and on the large screen in the project command center. The module also sets a liquefaction risk index alarm threshold. The system continuously monitors the LRI values ​​of each region in the distribution map. If the LRI value of a certain region is higher than the target value for multiple consecutive sampling periods, the system will take action. If this occurs, the area will be highlighted as a flashing red block on the distribution map, and an audible and visual alarm signal will be triggered simultaneously to remind operators and managers to pay close attention.

[0125] The global iteration and visualization monitoring module achieves holistic digital management and control of the core wall filling quality by cyclically executing intelligent control loops and integrating spatiotemporal data from the entire site. It integrates and elevates discrete, instantaneous point-based control processes into a continuous, global "digital twin" construction scenario. This not only ensures the continuity and self-consistency of intelligent compaction actions but also generates a traceable and assessable global quality status map. This provides construction management with a macro-level decision-making view and precise risk warnings, achieving a leap from single-point intelligence to site-wide wisdom.

[0126] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A vibration energy dissipation feedback control system for core wall anti-liquefaction, characterized in that, include: The data acquisition and positioning module is used to acquire, in real time, the vibration response signal generated by the interaction between the vibratory roller and the filling material when the vibratory roller is compacting the core wall filling material of the earth-rock dam, and to obtain the real-time position information of the vibratory roller. The signal processing and feature extraction module is used to perform time-frequency domain transformation processing on the vibration response signal and extract feature parameters from the transformation result to characterize the dynamic state of the packing. The real-time liquefaction risk diagnosis module is used to input the feature parameters into the pre-established liquefaction risk discrimination model and calculate the liquefaction risk index of the filler at the current rolling position. The intelligent decision-making and control instruction generation module is used to generate real-time control instructions for the operating parameters of the vibratory roller based on the liquefaction risk index. The actuator adaptive adjustment and targeted compaction module is used to adaptively adjust the excitation frequency, excitation force amplitude and travel speed of the vibratory roller according to the control command, so as to perform targeted compaction at the current compaction position that matches the liquefaction risk index; The global iteration and visualization monitoring module is used to control the system to repeatedly execute the functions of the data acquisition and positioning module and the actuator adaptive adjustment and targeted compaction module until the full compaction of the core wall filling area is completed. Based on the real-time location information and the liquefaction risk index, it generates and updates the compaction quality and liquefaction risk distribution map of the core wall filling area.

2. The excitation energy dissipation feedback control system for core wall anti-liquefaction according to claim 1, characterized in that, The workflow of the data acquisition and positioning module includes: The vibration acceleration signals perpendicular to the rolling surface, parallel to the rolling direction, and perpendicular to the rolling direction are collected by a triaxial acceleration sensor installed at the bearing seat of the vibratory roller. The planar coordinates and elevation information of the vibratory roller are obtained by a global navigation satellite system receiver installed on the frame of the vibratory roller, which serves as the real-time location information. The vibration acceleration signal is synchronized with the real-time location information to form a vibration response signal dataset with spatial location identification.

3. The excitation energy dissipation feedback control system for core wall anti-liquefaction according to claim 1, characterized in that, The workflow of the signal processing and feature extraction module includes: The vibration response signal is preprocessed with bandpass filtering and noise reduction. The preprocessed vibration response signal is subjected to a short-time Fourier transform to obtain the time spectrum of the signal. Calculate the main frequency energy ratio, high frequency dissipation coefficient, and energy uniformity from the time-frequency spectrum.

4. The excitation energy dissipation feedback control system for core wall anti-liquefaction according to claim 3, characterized in that, The workflow of the real-time liquefaction risk diagnosis module includes: The main frequency energy ratio, high frequency dissipation coefficient and energy uniformity are normalized to form a standardized feature vector. The standardized feature vector is input into the pre-established liquefaction risk discrimination model; Based on the standardized feature vector, the liquefaction risk discrimination model outputs a quantified liquefaction risk index. The liquefaction risk discrimination model is obtained by machine learning training based on historical test data of the core wall packing. The historical test data includes the dominant frequency energy ratio, high frequency dissipation coefficient, energy uniformity and their corresponding liquefaction resistance calibration values ​​measured under different compaction conditions.

5. The excitation energy dissipation feedback control system for core wall anti-liquefaction according to claim 4, characterized in that, The workflow of the intelligent decision-making and control command generation module includes: Based on the preset risk level range where the liquefaction risk index is located, the target compaction mode is determined. The risk level range includes a low-risk range, a medium-risk range, and a high-risk range, which correspond to the standard compaction mode, the reinforced compaction mode, and the repair compaction mode, respectively. Based on the target compaction mode, a reference command is generated for the excitation frequency setting value, excitation force amplitude setting value, and travel speed setting value of the vibratory roller through a preset fuzzy reasoning rule base. Based on the real-time changing trend of the liquefaction risk index, the baseline command is adjusted proportionally, integrally, and derivatively to ultimately generate the real-time control command.

6. The excitation energy dissipation feedback control system for core wall anti-liquefaction according to claim 5, characterized in that, The operating process of the actuator adaptive adjustment and targeted compaction module includes: The controller of the vibratory roller receives the real-time control command; The controller generates corresponding drive signals based on the excitation frequency setting, excitation force amplitude setting, and travel speed setting in the real-time control command. The drive signal is sent to the frequency converter, excitation force regulating valve and travel motor of the vibratory roller to synchronously adjust the excitation frequency, excitation force amplitude and travel speed of the vibratory roller. The current compaction position is compacted at least once using the adjusted working parameters to complete the targeted compaction.

7. The excitation energy dissipation feedback control system for core wall anti-liquefaction according to claim 1, characterized in that, The workflow of the global iteration and visualization monitoring module includes: During the continuous compaction operation of the vibratory roller on the core wall filling area, the functions of the data acquisition and positioning module, signal processing and feature extraction module, liquefaction risk real-time diagnosis module, intelligent decision-making and control command generation module, and actuator adaptive adjustment and targeted compaction module are cyclically called and executed. The real-time location information and corresponding liquefaction risk index obtained from each loop execution are merged with historical data and stored in a preset regional database; Based on the liquefaction risk index of all locations in the regional database, a spatial interpolation algorithm is used to generate and continuously update a compaction quality and liquefaction risk distribution map covering the entire core wall filling area. The distribution map is visualized, and areas where the liquefaction risk index is consistently higher than a preset threshold are marked and alerted.

Images