A dam micro-deformation monitoring method

Abstract

The application discloses a dam micro-deformation monitoring method, relates to the technical field of water conservancy engineering structure health monitoring, and comprises the following steps: installing a self-stabilizing platform connected with a fixed end and a universal joint on a dam; an inertial unit is used to perceive dam dynamics and drive an actuator to reversely compensate, so that a platform probe is stabilized relative to a space reference; a non-contact array is used to measure the vector of the displacement of the fixed end along with the dam; data locking is triggered when the displacement or energy characteristics exceed the limit; relative displacement and platform residual compensation are fused, vibration interference is removed through coordinate transformation, real micro-deformation in a dam geographic coordinate system is solved, and early warning is generated. The application effectively isolates high-frequency vibration interference in a physical layer through active self-stabilizing technology, and establishes a stable local measurement reference; a triggered acquisition mechanism greatly improves data value density; the method has high measurement precision, strong anti-interference capability, and can effectively identify small plastic damage of a dam.

Description

Technical Field

[0001] This invention relates to the field of structural health monitoring technology for water conservancy projects, and in particular to a method for monitoring micro-deformation of dam bodies. Background Technology

[0002] As a core component of water conservancy projects, the structural safety of dams directly impacts the safety of life and property downstream and the stability of the regional economy and society. During long-term operation, dam bodies are susceptible to minute deformations due to multiple factors such as water pressure, temperature changes, geological subsidence, and water erosion. If these deformations are not monitored and warned of in a timely manner, they may gradually develop into structural damage such as cracks and slippage, ultimately leading to safety accidents. Therefore, achieving high-precision, real-time, and stable monitoring of dam body micro-deformations is a key technical means to ensure the long-term safe operation of dams.

[0003] Existing dam micro-deformation monitoring technologies have many limitations and cannot meet the high-precision monitoring requirements under complex working conditions. Traditional contact monitoring methods require rigid connection to the dam body, which can easily disturb the dam structure during installation. Furthermore, long-term exposure to the outdoor environment makes them susceptible to factors such as temperature, humidity, and corrosion, leading to sensor aging and data drift, thus failing to achieve long-term stable micro-deformation measurements. While non-contact monitoring methods avoid the drawbacks of contact installation, they are significantly constrained by environmental conditions: GPS measurement accuracy is easily affected by satellite signal blockage and ionospheric interference, making it difficult to achieve nanometer-level micro-deformation monitoring requirements; total stations rely on manual operation or fixed reference points, are limited by weather and lighting conditions, and cannot effectively isolate the dam's own vibrations from environmental dynamic interference, resulting in a high proportion of noise in the measurement data and masking the true deformation signal.

[0004] Furthermore, existing monitoring technologies generally lack effective dynamic interference isolation and data processing mechanisms. During dam operation, the dam body is inevitably affected by dynamic loads such as water flow impact, wind load disturbance, and traffic vibration. Traditional monitoring systems lack dedicated attitude stabilization mechanisms, causing the measurement reference plane to drift with dam body swaying, resulting in displacement measurement results containing a large amount of invalid disturbance components. Simultaneously, the triggering logic of existing systems often relies on a single displacement threshold, making them susceptible to false triggering by environmental background noise and unable to accurately capture real deformation events. Data processing does not fully consider the impact of environmental factors and system errors, lacking multi-source data fusion and error correction mechanisms, leading to insufficient accuracy in monitoring results and failing to provide reliable data support for dam structural health assessment. These technical shortcomings make it difficult for existing monitoring systems to effectively distinguish between conventional elastic deformation and irreversible plastic damage in the dam body, resulting in low early warning sensitivity and reliability, thus hindering the development of dam safety monitoring technology. Summary of the Invention

[0005] This invention proposes a method for monitoring micro-deformation of dam bodies to solve the problems mentioned in the prior art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method for monitoring micro-deformation of a dam body, comprising the following steps: S1. Construct and install a dynamic isolation monitoring system, dividing the system into a fixed-end module and a self-stabilizing platform module with universal joint suspension support; the former deploys a high-frequency inertial measurement unit, and the latter is equipped with an active damping actuator and a non-contact displacement detection array to establish an absolute spatial reference. S2. Perform system zero-position self-calibration and environmental parameter initialization. In the absence of dynamic load triggering, collect static relative position data between the fixed end module and the self-stabilizing platform module as the initial zero-bias matrix, and combine it with ambient temperature and humidity sensor data to perform reference compensation on the optical or electromagnetic characteristic parameters of the non-contact displacement detection array. S3. Real-time monitoring and calculation of the dynamic attitude change vector of the fixed end of the dam body. The three-dimensional angular velocity and acceleration components are collected by the high-frequency inertial measurement unit. The instantaneous attitude angle is output by the quaternion attitude calculation algorithm. The attitude data is used as a feedforward signal to input the self-stabilizing platform module controller. S4. Based on the dynamic load isolation principle, the controller of the self-stabilizing platform module drives the active damping actuator to generate a compensation torque with reverse matching amplitude according to the attitude change data. The vibration transmission is offset by the mechanical decoupling characteristics of the universal joint, so that the probe always maintains a constant attitude. S5. Using a non-contact displacement detection array of a self-stabilizing platform, a measurement beam is emitted and received towards a fixed target surface to continuously measure minute three-dimensional displacement vectors and obtain the original displacement data stream containing dam deformation information and residual noise. S6. Execute trigger-based data acquisition and judgment based on dynamic threshold, set composite trigger conditions, calculate the short-time energy characteristics of the original displacement data stream in real time, and lock the transient data of the current time window when the absolute value of the relative displacement or the signal energy characteristics exceed the threshold, and record the data synchronously. S7. Integrate multi-source data to calculate the actual deformation of the dam body. Combine the relative displacement vector recorded at the trigger time with the dynamic compensation residual of the self-stabilizing platform for vector superposition calculation. Eliminate systematic deviations caused by active stabilization control, output the true micro-deformation of the dam body monitoring points in the geographic coordinate system, and generate monitoring logs.

[0007] Furthermore, the active damping control strategy of the self-stabilizing platform is optimized: in step S4, an active disturbance rejection control algorithm is used to drive the active damping actuator. In order to overcome the influence of universal joint friction and nonlinear aerodynamic disturbances on the stability of the platform, a compensation mechanism based on an extended state observer is introduced; by observing the angular position deviation and angular velocity fluctuation of the self-stabilizing platform in real time, the precise reverse compensation torque that needs to be applied is calculated.

[0008] Furthermore, it also includes fine-tuning the self-stabilizing platform using a fluid dynamics model: In step S4, a mathematical model of the platform's dynamic equilibrium is established, and the real-time compensation driving torque required by the self-stabilizing platform is calculated using the following dynamic response equilibrium formula. : ; in, This represents the three-dimensional rotational inertia matrix of the self-stabilizing platform system. Represents the differential control gain coefficient; This represents the relative angular velocity deviation vector between the fixed end and the self-stabilizing platform; Represents the integral control gain coefficient; This represents the relative angular deviation vector between the fixed end and the self-stabilizing platform; Indicates the time variable of integration; This represents the viscous damping coefficient matrix caused by the universal joint bearing and air damping. This represents the absolute angular velocity vector of the self-stabilizing platform relative to inertial space; This represents the magnitude of the Coulomb frictional torque existing within the system; Represents a symbolic function.

[0009] Furthermore, the method also includes displacement measurement using multi-beam laser interferometry: In step S5, the non-contact displacement detection array employs a laser interferometry subsystem; this subsystem consists of three orthogonally arranged laser emitting probes and corresponding position-sensitive detectors or interference fringe receivers, which respectively measure the minute displacement components of the fixed end target surface in the X, Y, and Z axes; using the stability of the laser wavelength as a reference for length measurement, the relative displacement value with nanometer-level resolution is resolved by analyzing the phase difference between the reflected light wave and the reference light wave or the movement of the interference fringes; simultaneously, the system performs real-time correction of the air refractive index along the optical path.

[0010] Furthermore, it also includes adaptive triggering logic based on energy spectral density: In step S6, to avoid false triggering caused by environmental background noise, the system does not simply rely on a single displacement amplitude threshold, but introduces a signal frequency domain energy analysis mechanism; by performing a sliding window fast Fourier transform on the original displacement data stream, the energy spectral density characteristics within a specific frequency band are extracted; when a sudden jump in energy is detected within the frequency range of the dam deformation characteristics, and the jump amplitude exceeds a preset multiple of the background noise base energy, it is determined to be a valid deformation event and data locking is triggered; after triggering, the system automatically extends the recording time window forward and backward, and downsamples and compresses the data in the non-trigger period for storage.

[0011] Furthermore, the process includes spatial coordinate transformation and error correction of the measurement data: In step S7, the measured relative displacement vector needs to be mapped from the instrument coordinate system to the dam's structural coordinate system; a quaternion rotation matrix is ​​used to perform a spatial rotation transformation on the original displacement vector, while introducing a nonlinear deviation correction term caused by temperature drift and mechanical installation errors; the true deformation vector of the dam body in the geographic coordinate system is calculated using the following spatial coordinate calculation formula. : ; in, This represents the true deformation and displacement vector of the dam body monitoring points in the geographic coordinate system after correction; This represents the rotation transformation matrix from the instrument measurement coordinate system to the geographic coordinate system, which is determined by the system's mounting Euler angles. Decide; This represents the inverse operation of a matrix; This represents the original relative displacement vector measured by the displacement detection array; This represents the high-frequency random noise vector of the system estimated by the filtering algorithm; This represents the lever arm vector from the probe center to the center of rotation of the fixed end; This represents the vector cross product operation; This represents the small angular velocity vector remaining after compensation on the self-stabilizing platform; Indicates the sampling time interval.

[0012] Furthermore, a redundant measurement mechanism based on the capacitive sensing principle is also included: In step S5, a differential capacitive displacement sensor or eddy current sensor is arranged in the slit between the fixed end and the self-stabilizing platform; the sensor uses the change in capacitance caused by the change in distance between the two plates, or the change in impedance generated by eddy currents induced on the surface of the metal conductor, to invert minute displacements; in the data processing unit, the analog voltage signal acquired by the capacitive / inductive sensor is converted from analog to digital and then weighted and fused with the optical measurement data; when the optical path is blocked by dust or water mist, the system automatically increases the weight of the capacitive sensing data.

[0013] Furthermore, it also includes distributed clock synchronization and trigger network coordination: In step S6, when a single monitoring point meets the triggering conditions, the monitoring node sends a coordinated trigger broadcast signal to adjacent monitoring nodes deployed in other locations on the dam through an industrial wireless network or fiber optic ring network; all nodes that receive the broadcast signal will synchronously start the high-frequency data acquisition mode even if they have not reached the local trigger threshold, and use the IEEE 1588 precise time protocol to align the data timestamps of all nodes at the sub-microsecond level.

[0014] Furthermore, it also includes multidimensional correlation analysis and elimination of environmental impact factors: In step S7, the data processing module not only calculates the geometric deformation, but also simultaneously reads the temperature field distribution, reservoir water level, and atmospheric pressure data on the dam surface; it uses multiple regression analysis or machine learning algorithms to establish a benchmark correlation model between environmental quantities and deformation quantities; when calculating the final deformation quantity, it first uses this model to deduct the conventional periodic components caused by thermal expansion and contraction and hydraulic elastic deformation, and separates the non-correlated abnormal micro-deformation mutations.

[0015] Furthermore, it also includes trend prediction and health scoring based on historical data: After step S7, the system stores the calculated actual micro-deformation into the historical database and constructs a time series prediction model; this model uses autoregressive integral moving average or long short-term memory network to extrapolate the future deformation trend of the dam body; the system calculates the structural health score of the dam in real time according to the degree of deviation between the measured deformation and the predicted value; when the score is lower than the safety warning line or the predicted deformation rate shows an exponential growth trend, the system automatically generates a graded alarm signal and sends a diagnostic report to the dam safety monitoring center.

[0016] Compared with existing technologies, the beneficial effects of this invention are: The dynamic isolation monitoring system constructed using this method, through the modular design of the fixed end and the self-stabilizing platform, combined with multi-degree-of-freedom universal joints and active damping actuators, effectively isolates dynamic disturbances in the dam body. The self-stabilizing platform performs reverse motion compensation based on the attitude data from the fixed end, which can counteract high-frequency vibrations and environmental interference, maintain the constant attitude of the measurement reference plane, and physically eliminate the influence of invalid disturbances on the measurement. This provides a stable inertial reference for micro-deformation measurement and significantly improves the signal-to-noise ratio of the measurement data.

[0017] In terms of measurement accuracy, a non-contact displacement detection array combined with multi-beam laser interferometry is employed, fully utilizing the stability and high-resolution characteristics of laser wavelength to achieve three-dimensional displacement measurement with nanometer-level precision. Simultaneously, mechanisms such as real-time air refractive index correction and redundant capacitance / eddy current sensing effectively resist interference from environmental medium changes and harsh working conditions, ensuring the continuity and reliability of data acquisition and solving the problems of insufficient accuracy and weak anti-interference capability of traditional methods.

[0018] In terms of data processing and triggering mechanisms, an adaptive triggering logic based on energy spectral density is introduced. Frequency domain energy analysis distinguishes effective deformation from background noise, avoiding false triggering caused by a single threshold and ensuring complete capture of deformation events. Combined with spatial coordinate transformation, multi-source data fusion, and environmental factor elimination algorithms, the system can accurately separate the actual deformation of the dam from conventional periodic deformation and systematic errors, outputting the true micro-deformation in the geographic coordinate system, providing precise data support for structural health assessment.

[0019] Furthermore, the distributed clock synchronization and trigger network coordination mechanism enables the acquisition of spatiotemporal correlated data from multiple monitoring points, distinguishing between local cracks and overall structural displacement, thus providing the possibility for full-field deformation modal analysis of the dam. The trend prediction and health scoring functions based on historical data not only assess the dam's structural status in real time but also predict deformation trends in advance, generating tiered alarm signals. This significantly improves the timeliness and accuracy of early warnings, providing proactive prevention and control measures for the safe operation and maintenance of the dam.

[0020] Overall, this invention, through technological innovations such as dynamic isolation, high-precision measurement, intelligent triggering, and multi-source fusion, achieves high-precision, high-reliability, and real-time monitoring of dam body micro-deformation, effectively solving the problems of weak anti-interference capability, insufficient accuracy, and inaccurate early warning in existing technologies. This method is adaptable to complex outdoor working conditions, convenient to install and maintain, and capable of long-term stable operation. It provides comprehensive technical support for dam structural health assessment and safety early warning, and is of great significance for ensuring the safe operation of water conservancy projects and reducing accident risks, with broad application prospects. Attached Figure Description

[0021] Figure 1 This is a schematic block diagram of the dam micro-deformation monitoring method proposed in this invention; Figure 2 Comparison of micro-deformation signal extraction under strong vibration background; Figure 3 This is a schematic diagram of a trigger-based acquisition and judgment method based on energy spectral density. Figure 4 This is a trend chart of net deformation after removing the correlation of environmental factors. Detailed Implementation

[0022] 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.

[0023] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0024] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The invention will now be described in further detail with reference to the accompanying drawings.

[0025] Reference Figures 1 to 4 A method for monitoring micro-deformation of a dam body, comprising the following steps: Step S1: Construct and install a dynamic isolation monitoring system. The monitoring system is divided into a fixed end module that is rigidly connected to the measuring points on the dam body and a self-stabilizing platform module that is suspended and supported by a multi-degree-of-freedom universal joint. A high-frequency inertial measurement unit is deployed in the fixed end module to capture the transient vibration and tilt attitude of the dam body surface. At the same time, an active damping actuator and a non-contact displacement detection array are configured on the self-stabilizing platform module to establish an absolute spatial reference based on the geocentric inertial coordinate system. Step S2: Perform system zero-point self-calibration and environmental parameter initialization. Under the condition of no dynamic load triggering, collect the static relative position data between the fixed end module and the self-stabilizing platform module as the initial zero bias matrix, and combine the ambient temperature and humidity sensor data to perform benchmark compensation on the optical or electromagnetic characteristic parameters of the non-contact displacement detection array to establish the static equilibrium operating point of the monitoring system. Step S3: Monitor and calculate the dynamic attitude change vector of the fixed end of the dam in real time. The angular velocity component and linear acceleration component of the dam in three-dimensional space are continuously collected by the high-frequency inertial measurement unit in the fixed end module. The instantaneous attitude angle of the fixed end relative to the absolute space reference is output in real time using the quaternion attitude calculation algorithm. The attitude data is then used as a feedforward signal to be input to the controller of the self-stabilizing platform module. Step S4: Based on the dynamic load isolation principle, reverse motion compensation of the self-stabilizing platform is performed. The controller of the self-stabilizing platform module drives the active damping actuator to generate a compensation torque that is opposite to the direction of the dam disturbance and matches the amplitude according to the received attitude change data. The mechanical decoupling characteristics of the universal joint cancel out the transmission of external vibration energy, forcing the probe carried by the self-stabilizing platform module to always maintain a constant attitude relative to the preset spatial reference, forming an inertial stable measurement field independent of the dam sway. Step S5: Perform precise measurement of the relative micro-displacement between the fixed end and the self-stabilizing platform. Using a non-contact displacement detection array installed on the self-stabilizing platform, a measurement beam is emitted towards the target surface of the fixed end and the reflected signal is received. The tiny three-dimensional displacement vector generated by the fixed end relative to the inertial stabilizing probe as the dam body deforms is continuously measured at a time resolution of microseconds. The original displacement data stream containing the real deformation information of the dam body and residual vibration noise is obtained. Step S6: Execute trigger-based data acquisition judgment based on dynamic threshold, set a composite trigger condition including displacement amplitude threshold and energy density threshold, calculate the short-time energy characteristics of the original displacement data stream in real time, and immediately lock the transient data in the current time window when the monitored relative displacement absolute value or signal energy characteristics exceed the preset trigger threshold, and simultaneously record the compensation execution amount of the self-stabilizing platform and the attitude angle data of the fixed end at this time. Step S7: Integrate multi-source data to calculate the actual deformation of the dam body. Combine the relative displacement vector recorded at the trigger time with the dynamic compensation residual of the self-stabilizing platform for vector superposition calculation. Eliminate systematic deviations caused by active stabilization control, output the true micro-deformation of the dam body monitoring points in the geographic coordinate system after coordinate system transformation, and generate a monitoring log containing deformation trend and structural health status assessment.

[0026] This invention also includes optimizing the active damping control strategy of the self-stabilizing platform: In step S4, an active disturbance rejection control algorithm is used to drive the active damping actuator. In order to overcome the influence of universal joint friction and nonlinear aerodynamic interference on the stability of the platform, a compensation mechanism based on an extended state observer is introduced. By observing the angular position deviation and angular velocity fluctuation of the self-stabilizing platform in real time, the required precise reverse compensation torque is calculated to ensure that the platform can maintain arcsecond-level attitude stability accuracy when the dam body experiences high-frequency vibration, thereby ensuring that the reference plane for displacement measurement does not drift. This control process generates a corresponding electromagnetic driving force by adjusting the output current of the actuator, so that the angular acceleration of the platform relative to the inertial space approaches zero, thereby achieving physical isolation from the high-frequency dynamic load of the dam body.

[0027] This invention also includes fine-tuning control of the self-stabilizing platform using a fluid dynamics model: In step S4, to further eliminate the influence of environmental wind loads or micro-airflow on the suspended self-stabilizing platform, a mathematical model of the platform's dynamic equilibrium is established. Precise control commands are output by calculating the relationship between the platform's moment of inertia and the required angular acceleration. A control law is set so that the platform can quickly return to its horizontal zero position when subjected to external disturbances. This process involves comprehensive calculations of the platform's moment of inertia tensor, damping coefficient tensor, and nonlinear disturbance terms. The real-time compensation driving torque required by the self-stabilizing platform is calculated using the following dynamic response equilibrium formula. : ; in, The three-dimensional rotational inertia matrix represents the self-stabilizing platform system and is used to describe the inertial impedance characteristics of the platform's mass distribution in response to rotation. This represents the differential control gain coefficient, used to adjust the response sensitivity to the rate of change of angle; This represents the relative angular velocity deviation vector between the fixed end and the self-stabilizing platform; This represents the integral control gain coefficient, used to eliminate steady-state accumulated error; This represents the relative angular deviation vector between the fixed end and the self-stabilizing platform; Indicates the time variable of integration; This represents the viscous damping coefficient matrix caused by the universal joint bearing and air damping. This represents the absolute angular velocity vector of the self-stabilizing platform relative to inertial space; This represents the magnitude of the Coulomb frictional torque existing within the system; The sign function is used to determine the direction of frictional force.

[0028] This invention also includes displacement measurement using multi-beam laser interferometry: In step S5, the non-contact displacement detection array employs a laser interferometry subsystem; this subsystem consists of three orthogonally arranged laser emitting probes and corresponding position-sensitive detectors (PSDs) or interference fringe receivers, which respectively measure the minute displacement components of the fixed end target surface in the X, Y, and Z axes; using the stability of the laser wavelength as a reference for length measurement, the relative displacement value with nanometer-level resolution is resolved by analyzing the phase difference between the reflected light wave and the reference light wave or the movement of the interference fringes; simultaneously, the system performs real-time correction of the air refractive index along the optical path, with the correction factor generated based on the on-site measured air pressure and temperature gradient data to eliminate the influence of environmental medium changes on optical path measurement.

[0029] This invention also includes an adaptive triggering logic based on energy spectral density: In step S6, to avoid false triggering caused by environmental background noise (such as water flow impact and traffic vibration), the system does not simply rely on a single displacement amplitude threshold, but introduces a signal frequency domain energy analysis mechanism; by performing a sliding window Fast Fourier Transform (FFT) on the original displacement data stream, the energy spectral density features within a specific frequency band are extracted; when a sudden jump in energy is detected within the frequency range of the dam deformation characteristics, and the jump amplitude exceeds a preset multiple of the background noise base energy, it is determined to be a valid deformation event and data locking is triggered; after triggering, the system automatically extends the recording time window forward and backward to ensure complete capture of the start and end states of the deformation process, and downsamples and compresses the data during non-trigger periods for storage as a background reference.

[0030] This invention also includes spatial coordinate transformation and error correction of the measurement data: In step S7, considering that the self-stabilizing platform may still have a small residual attitude tilt angle even after compensation, the measured relative displacement vector needs to be accurately mapped from the instrument coordinate system to the dam's structural coordinate system; a quaternion rotation matrix is ​​used to perform a spatial rotation transformation on the original displacement vector, while introducing a nonlinear deviation correction term caused by temperature drift and mechanical installation errors; the true deformation vector of the dam body in the geographic coordinate system is calculated using the following spatial coordinate calculation formula. : ; in, This represents the true deformation and displacement vector of the dam body monitoring points in the geographic coordinate system after correction; This represents the rotation transformation matrix from the instrument measurement coordinate system to the geographic coordinate system, which is determined by the system's mounting Euler angles. Decide; This represents the inverse operation of a matrix; This represents the original relative displacement vector measured by the displacement detection array; This represents the high-frequency random noise vector of the system estimated by the filtering algorithm; This represents the lever arm vector from the probe center to the center of rotation of the fixed end; This represents the vector cross product operation; This represents the small angular velocity vector remaining after compensation on the self-stabilizing platform; This represents the sampling time interval, used to convert the velocity component integral into a displacement correction.

[0031] This invention also includes a redundant measurement mechanism based on the principle of capacitive sensing: In step S5, in order to improve the reliability and anti-interference capability of the monitoring data, in addition to optical measurement methods, a differential capacitive displacement sensor or eddy current sensor is arranged in the slit between the fixed end and the self-stabilizing platform; the sensor uses the change in capacitance caused by the change in distance between the two plates, or the change in impedance generated by the eddy current induced on the surface of the metal conductor, to invert the small displacement; in the data processing unit, the analog voltage signal obtained by the capacitive / inductive sensor is converted from analog to digital and then weighted and fused with the optical measurement data; when the optical path is blocked by dust or water mist, the system automatically increases the weight of the capacitive sensing data to ensure that the continuous micro-deformation monitoring capability can still be maintained under harsh working conditions.

[0032] This invention also includes distributed clock synchronization and trigger network coordination: In step S6, when a single monitoring point meets the triggering conditions, the monitoring node sends a coordinated trigger broadcast signal to adjacent monitoring nodes deployed in other locations on the dam via an industrial wireless network or fiber optic ring network; all nodes that receive the broadcast signal will be forced to synchronously start the high-frequency data acquisition mode even if they have not reached the local trigger threshold, and will use the IEEE 1588 Precise Time Protocol (PTP) to align the data timestamps of all nodes at the sub-microsecond level; this coordination mechanism is used to capture the overall modal wave propagation process of dam deformation, thereby distinguishing between the expansion of local surface cracks and the overall structural displacement of the dam, and providing spatiotemporal correlation data of the entire field for dam safety assessment.

[0033] This invention also includes multidimensional correlation analysis and elimination of environmental impact factors: In step S7, the data processing module not only calculates the geometric deformation, but also simultaneously reads the temperature field distribution, reservoir water level, and atmospheric pressure data of the dam surface; a benchmark correlation model between environmental quantities and deformation is established using multiple regression analysis or machine learning algorithms; when calculating the final deformation, the model is first used to deduct the conventional periodic components caused by thermal expansion and contraction and hydraulic elastic deformation, and to separate the non-correlated abnormal micro-deformation abrupt changes; this method can identify the irreversible plastic damage accumulation hidden under the diurnal temperature variation deformation cycle, and has a higher signal-to-noise ratio and early warning accuracy for capturing potential structural cracks or sliding trends in the dam.

[0034] This invention also includes trend prediction and health scoring based on historical data: After step S7, the system stores the calculated actual micro-deformation in a historical database and constructs a time series prediction model; this model uses autoregressive integral moving average (ARIMA) or long short-term memory network (LSTM) to extrapolate the future deformation trend of the dam body; the system calculates the structural health score of the dam in real time based on the deviation between the measured deformation and the predicted value, combined with the allowable deformation limit in the dam design specifications; when the score is lower than the safety warning line or the predicted deformation rate shows an exponential growth trend, the system automatically generates a graded alarm signal and sends a detailed diagnostic report containing deformation vector diagrams, acceleration spectrum and predicted curves to the dam safety monitoring center through a remote communication module.

[0035] The following two examples further illustrate specific embodiments of the present invention: Example 1: This example details a precision monitoring system for micro-deformation of a high-head arch dam in the flood discharge vibration zone and its operation method. This example primarily addresses the challenge of extracting micron-level displacements under strong vibration background noise. By constructing a high-precision mechanically self-stabilizing platform and a multi-beam laser interferometry unit, it achieves real-time capture of the actual deformation of key parts of the dam under dynamic loads.

[0036] During the hardware construction and installation phase, dynamic isolation monitoring devices are first installed on both sides of key structural joints at the selected dam crest or waist. This device mainly consists of two parts: a fixed-end module and a suspended self-stabilizing platform module. The fixed-end module is made of high-strength, corrosion-resistant alloy material and is rigidly connected to the dam's measuring point surface via multi-point anchoring, ensuring it can synchronously move with the dam body without attenuation, including any minute translations, rotations, and high-frequency vibrations. At the core of the fixed-end module, a high-frequency inertial measurement unit is integrated. This unit includes a three-axis fiber optic gyroscope and a three-axis microelectromechanical system accelerometer, with a sampling frequency set at the kilohertz level, specifically designed to capture transient vibration vectors and tilt attitude change rates on the dam surface.

[0037] The self-stabilizing platform module is suspended and supported within the internal cavity of the fixed-end module by a set of precision three-degree-of-freedom air bearings or magnetic levitation universal joints, achieving flexible decoupling in the mechanical structure. A non-contact active damping actuator is configured on the self-stabilizing platform, consisting of multiple voice coil motors or torque motors, capable of outputting precise control torque along three orthogonal axes. Furthermore, the self-stabilizing platform carries the core measurement component, namely a non-contact displacement detection array. In this embodiment, the array employs an optical system based on the principle of laser interferometry, containing three orthogonally arranged frequency-stabilized laser emission probes. The corresponding reflecting target mirror is mounted on the inner wall of the fixed-end module. To ensure the absolute reference of the measurement, the system performs zero-point self-calibration during the initialization phase, collecting relative position data in a static state as the initial zero-bias matrix, and combining this with data from the built-in high-precision temperature and humidity sensors to establish a compensation model for the refractive index of the environmental medium, thereby establishing the system's static equilibrium operating point.

[0038] During monitoring and operation, the system enters the real-time dynamic isolation and measurement phase. The inertial measurement unit within the fixed-end module senses the angular velocity and linear acceleration components of the dam body in real time due to flood discharge or crustal micro-motion. This motion data is transmitted to the central controller of the self-stabilizing platform via a high-speed bus. The controller internally runs an attitude calculation algorithm based on active disturbance rejection control theory and an active damping control strategy. To overcome the impact of potential minor frictional forces in the universal joint bearings and nonlinear aerodynamic disturbances on platform stability, the controller introduces an extended state observer mechanism. This observer estimates and compensates for the total disturbance within the system in real time, calculating the counter-compensation torque required to maintain the platform absolutely stationary relative to the geocentric inertial coordinate system. Subsequently, the actuator controls the execution mechanism to generate the corresponding electromagnetic driving force, ensuring that the angular acceleration of the self-stabilizing platform remains close to zero even when the dam body experiences severe shaking, thus maintaining a constant attitude inertial measurement reference plane in a violent vibration environment.

[0039] On this stable reference plane, the multi-beam laser interferometry subsystem begins operation. Three laser probes emit measurement beams towards the fixed-end target surface, which sways with the dam. By analyzing the phase difference or interference fringe shift between the reflected and reference beams, the system resolves the minute displacement vector of the fixed end relative to the self-stabilizing platform at a nanosecond-level response speed. Since the self-stabilizing platform remains stationary, this measurement essentially reflects the instantaneous displacement of the measuring point on the dam relative to inertial space. To eliminate air disturbance errors in the laser beam path, the system uses real-time acquired air pressure and temperature gradient data to correct the refractive index of the optical path, ensuring the physical accuracy of the displacement data.

[0040] Subsequently, the system executes a trigger-based data acquisition and judgment based on dynamic thresholds. The processor monitors the raw displacement data stream in real time and sets composite trigger conditions that include displacement amplitude and energy density. Only when the absolute value of the monitored relative displacement or the signal energy characteristics exceed the preset safety threshold will the system lock the transient data within the current time window. This mechanism effectively filters out irrelevant interference such as vehicle traffic. The data acquired during the triggering process not only includes the relative displacement but also simultaneously records the dynamic compensation residual of the self-stabilizing platform and the attitude angle data of the fixed end.

[0041] At the final stage of data processing, the system performs multi-source data fusion and coordinate transformation. Considering that the self-stabilizing platform may still have microsecond-level residual attitude tilt angles under extreme conditions, the algorithm uses a quaternion rotation matrix to precisely map the original measurement vectors from the instrument coordinate system to the dam's geographic structural coordinate system. Simultaneously, a nonlinear deviation correction term caused by temperature drift and mechanical installation errors is introduced to eliminate systematic deviations that may be introduced by active stabilization control. Finally, the system outputs the rigorously calibrated true micro-deformation of the dam's monitoring points in the geographic coordinate system and generates a monitoring log containing deformation trend analysis for use by the dam safety management center in decision-making.

[0042] To verify the monitoring effectiveness of this embodiment under strong vibration conditions, we conducted a comparative test during the flood discharge of a large hydropower station. The table below shows a comparison of the data characteristics of the traditional hydrostatic level and the dynamic load isolation monitoring system described in this embodiment at the same measuring point and during the same time period.

[0043] Table 1: Comparison of Displacement Monitoring Data Quality under Strong Vibration Conditions

[0044] The data in the table above intuitively reflects the significant advantages of the technical solution in this embodiment under complex dynamic environments. Firstly, regarding vibration and noise suppression, traditional systems, lacking physical isolation mechanisms, have readings mixed with a large amount of high-frequency vibration induced by flood discharge, resulting in a noise peak-to-peak value as high as 5.24 mm, severely masking minute structural deformations. This system, however, utilizes a self-stabilizing platform to achieve physical-level noise reduction, lowering the noise level to 0.08 mm, greatly improving the signal-to-noise ratio and making micron-level deformations clearly visible. Secondly, in terms of response time, this system, leveraging its kilohertz-level sampling rate and the characteristics of light-speed measurement, shortens the response delay to less than 2 milliseconds, enabling it to completely capture the transient deformation process under impact loads. Traditional instruments, due to the inertia and viscosity of liquids, exhibit a lag of several seconds, failing to reflect dynamic processes. Furthermore, the significant reduction in reference drift demonstrates the reliability of the inertial stable reference, completely solving the measurement distortion problem caused by the reference point moving with the dam bedrock in traditional methods. With a true micro-deformation recognition rate of up to 98%, the system can effectively distinguish between invalid elastic vibrations and potentially hazardous plastic deformations, providing high-quality data support for dam safety assessments.

[0045] Example 2: This example focuses on describing an intelligent network collaborative monitoring system applied to the full life-cycle management of ultra-high concrete gravity dams. This example not only includes single-point dynamic load isolation measurement technology, but also integrates a multi-sensor redundancy mechanism, a distributed collaborative triggering network, and an intelligent data processing method based on environmental correlation elimination, aiming to solve the problems of monitoring reliability and data mining depth under long-term harsh natural environments.

[0046] In the hardware configuration of this embodiment, the measurement link between the self-stabilizing platform and the fixed end adopts a dual redundancy design. In addition to the laser interferometry measurement components in the main channel, a capacitive displacement sensor array based on the eddy current effect or the variable pole distance principle is symmetrically arranged in the tiny gap between the fixed end and the self-stabilizing platform. This design utilizes the impedance change of the induced current on the conductor surface or the change of the electric field between the plates to invert the displacement. In the system operation logic, the data processing module evaluates the quality of the optical signal in real time. When encountering the high humidity, dense fog, or dust environment unique to the dam corridor, which causes the laser beam energy to attenuate or even be blocked, the system automatically adjusts the weighting coefficients and smoothly switches to capacitive or eddy current sensing mode to ensure the continuity and integrity of the monitoring data.

[0047] At the control strategy level, this embodiment applies a fluid dynamics model for fine-tuning control of the self-stabilizing platform. Considering the perennial canyon wind field on the dam surface, the controller establishes a dynamic equilibrium mathematical model of the platform. By calculating the platform's rotational inertia tensor and air damping coefficient tensor in real time, it accurately outputs control commands to counteract wind load disturbances. This model-based control law enables the platform to quickly return to a horizontal zero position when subjected to nonlinear airflow disturbances, relying on the combination of inertia and active torque, further improving measurement stability.

[0048] The core innovation of this embodiment lies in the construction of a distributed clock synchronization and triggering network. Dozens of dynamic load isolation monitoring nodes deployed on the dam are interconnected through an industrial fiber optic ring network. The system uses the IEEE 1588 precision time protocol to synchronize all nodes at the sub-microsecond level. In terms of triggering logic, this embodiment is no longer limited to single-point threshold determination, but introduces an adaptive triggering mechanism based on energy spectral density. The system performs sliding window spectral analysis on the raw displacement data stream and calculates the energy density in a specific frequency band in real time. When a key node detects an abnormal energy surge within the deformation characteristic frequency range and confirms that it is not background noise, the node not only immediately locks its local data, but also broadcasts a collaborative triggering signal to the entire network. After receiving the signal, neighboring nodes will force the activation of high-frequency acquisition mode even if they have not reached their local threshold. This collaborative mechanism enables the system to capture the propagation path and attenuation characteristics of deformation waves in the dam structure, thereby distinguishing between the expansion of local surface cracks and the overall structural displacement of the dam.

[0049] In the data analysis and processing phase, this embodiment emphasizes the elimination of multidimensional correlations among environmental impact factors. The data processing center simultaneously accesses multidimensional environmental data such as dam water level, air temperature, reservoir water temperature, and atmospheric pressure. The system utilizes machine learning algorithms to construct a baseline correlation model between environmental quantities and deformation quantities. When calculating the final actual deformation quantity, the algorithm first uses this model to subtract the thermal expansion and contraction components caused by seasonal temperature differences and the elastic deformation components caused by water level changes, thereby separating out those uncorrelated, anomalous micro-deformation abrupt changes. This processing method can sensitively identify the accumulation of minute plastic damage hidden beneath large-amplitude periodic deformation curves.

[0050] In addition, the system integrates trend prediction functionality based on historical data. Utilizing time-series prediction models such as Long Short-Term Memory (LSTM) networks, deep learning is applied to the cleaned deformation data to predict future deformation trends of the dam body. The system calculates a structural health score in real time based on the deviation between measured and predicted values, combined with the allowable deformation limits in the dam design specifications. If the score falls below the warning line or the predicted deformation rate deteriorates exponentially, the system automatically generates a diagnostic report including a deformation vector map and acceleration spectrum, and remotely sends it to the monitoring center.

[0051] To quantify the performance of this embodiment in intelligent data management and anomaly detection, we compared the data performance of the traditional continuous acquisition mode and the collaborative triggering mode of this embodiment during a one-year trial operation period, as shown in the table below.

[0052] Table 2: Comparison of the effectiveness of intelligent network collaborative monitoring and traditional methods

[0053] The data in the table above profoundly reveals the significant leap forward in the practicality of the intelligent collaborative monitoring model in engineering. Firstly, regarding data storage and processing efficiency, this embodiment, through an intelligent triggering mechanism, drastically reduces the annual data volume from 520TB to 4.8TB, a compression ratio exceeding 100 times. This not only significantly reduces hardware storage costs but, more importantly, eliminates massive amounts of invalid, stable data, shortening data retrieval and analysis time from days to minutes, achieving true real-time early warning. Despite the substantial reduction in data volume, the number of effective abnormal events captured did not decrease, and thanks to energy spectral density analysis, the false alarm rate plummeted from 45% to 2%, demonstrating the accuracy of the triggering logic. In terms of multi-point collaboration, sub-microsecond synchronization accuracy enables the system to record the propagation process of deformation waves, something completely impossible with traditional independent acquisition modes, providing a new perspective for dam structural dynamics analysis. Finally, thanks to multi-sensor fusion and non-contact measurement design, the equipment maintenance cycle is significantly extended, greatly reducing the manual operation and maintenance costs and risks in high-altitude and challenging areas, demonstrating extremely high economic benefits and application value.

[0054] Reference Figure 2 This figure visually demonstrates the core advantages of dynamic load isolation technology. During flood discharge in large-scale water conservancy projects, strong high-frequency vibrations occur on the dam surface. Traditional monitoring equipment, directly fixed to the dam, collects signals filled with vibration noise, the amplitude of which far exceeds the actual quasi-static deformation, completely obscuring the true deformation signal. As the data shows, traditional methods exhibit numerical fluctuations ranging from several millimeters, displaying random high-frequency characteristics, making it impossible to identify valid information. In contrast, this invention, through a mechanically self-stabilizing platform and active damping control, filters out most high-frequency dynamic load interference at the physical level. The red curve shows minimal displacement changes with a clear linear trend, accurately reflecting the true micro-deformation of the dam under pressure. This "benchmark-stable" technical design allows the monitoring system to maintain arcsecond-level attitude stability and nanometer-level displacement resolution even under extreme dynamic interference environments, providing high signal-to-noise ratio foundational data for accurate assessment of dam structural health.

[0055] Reference Figure 3The diagram clearly illustrates the system's intelligent data management and anomaly identification logic. During long-term dam monitoring, constant background noise exists, and traditional continuous high-frequency acquisition generates massive amounts of invalid and redundant data, increasing the burden on transmission and storage. This invention introduces a trigger-based mechanism based on energy spectral density. Although there are energy fluctuations during time periods T1 and T2, the spectral characteristics do not reach the preset composite threshold, so the system only records low-power background data. At time T3, when the dam experiences structural displacement or significant impact, the signal energy density increases dramatically, far exceeding the red threshold baseline. The system immediately locks the high-precision measurement channel, achieving sub-microsecond data truncation and preservation. Through this mechanism, the system not only reduces more than 90% of invalid data redundancy but also eliminates false alarms caused by artifacts and environmental interference through frequency domain feature analysis, significantly improving the reliability of monitoring and early warning, and helping managers quickly locate structural anomalies.

[0056] Reference Figure 4 This figure demonstrates the system's deep data processing capabilities at the advanced level, specifically identifying minute structural damage from complex seasonal cyclic deformation. Dams are affected by environmental factors such as reservoir water level fluctuations, diurnal and seasonal temperature differences, resulting in significant periodic elastic expansion and contraction. The total monitored deformation exhibits obvious fluctuations, masking potential trend displacements. This invention uses multi-source sensor fusion to simultaneously collect environmental variables such as water level and temperature, and utilizes regression models or machine learning algorithms to calculate the theoretical deformation components corresponding to these environmental factors. Subtracting the environmentally related components from the total deformation curve yields the bottom residual net deformation curve. Although the value of this curve is only on the millimeter scale, it shows a continuously rising and irreversible trend, indicating the possible existence of slowly developing microcracks or plastic rheology within the dam body. This processing method elevates monitoring accuracy from the macroscopic geometric level to the level of diagnosing the health status of microstructures, helping engineers capture early signs of structural weakening months before visible damage appears in the dam, providing forward-looking data support for the safe operation and maintenance of the dam.

[0057] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for monitoring micro-deformation of a dam body, characterized in that, Includes the following steps: S1. Construct and install a dynamic isolation monitoring system, dividing the system into a fixed-end module and a self-stabilizing platform module with universal joint suspension support; the former deploys a high-frequency inertial measurement unit, and the latter is equipped with an active damping actuator and a non-contact displacement detection array to establish an absolute spatial reference. S2. Perform system zero-position self-calibration and environmental parameter initialization. In the absence of dynamic load triggering, collect static relative position data between the fixed end module and the self-stabilizing platform module as the initial zero-bias matrix, and combine it with ambient temperature and humidity sensor data to perform reference compensation on the optical or electromagnetic characteristic parameters of the non-contact displacement detection array. S3. Real-time monitoring and calculation of the dynamic attitude change vector of the fixed end of the dam body. The three-dimensional angular velocity and acceleration components are collected by the high-frequency inertial measurement unit. The instantaneous attitude angle is output by the quaternion attitude calculation algorithm. The attitude data is used as a feedforward signal to input the self-stabilizing platform module controller. S4. Based on the dynamic load isolation principle, the controller of the self-stabilizing platform module drives the active damping actuator to generate a compensation torque with reverse matching amplitude according to the attitude change data. The vibration transmission is offset by the mechanical decoupling characteristics of the universal joint, so that the probe always maintains a constant attitude. S5. Using a non-contact displacement detection array of a self-stabilizing platform, a measurement beam is emitted and received towards a fixed target surface to continuously measure minute three-dimensional displacement vectors and obtain the original displacement data stream containing dam deformation information and residual noise. S6. Execute trigger-based data acquisition and judgment based on dynamic threshold, set composite trigger conditions, calculate the short-time energy characteristics of the original displacement data stream in real time, and lock the transient data of the current time window when the absolute value of the relative displacement or the signal energy characteristics exceed the threshold, and record the data synchronously. S7. Integrate multi-source data to calculate the actual deformation of the dam body. Combine the relative displacement vector recorded at the trigger time with the dynamic compensation residual of the self-stabilizing platform for vector superposition calculation. Eliminate systematic deviations caused by active stabilization control, output the true micro-deformation of the dam body monitoring points in the geographic coordinate system, and generate monitoring logs.

2. The method for monitoring micro-deformation of a dam body according to claim 1, characterized in that, It also includes optimizing the active damping control strategy of the self-stabilizing platform: In step S4, an active disturbance rejection control algorithm is used to drive the active damping actuator. In order to overcome the influence of universal joint friction and nonlinear aerodynamic interference on the stability of the platform, a compensation mechanism based on an extended state observer is introduced; by observing the angular position deviation and angular velocity fluctuation of the self-stabilizing platform in real time, the precise reverse compensation torque that needs to be applied is calculated.

3. The method for monitoring micro-deformation of a dam body according to claim 1, characterized in that, It also includes fine-tuning the self-stabilizing platform using a fluid dynamics model: In step S4, a mathematical model of the platform's dynamic equilibrium is established, and the real-time compensation driving torque required by the self-stabilizing platform is calculated using the following dynamic response equilibrium formula. : ; in, This represents the three-dimensional rotational inertia matrix of the self-stabilizing platform system. Represents the differential control gain coefficient; This represents the relative angular velocity deviation vector between the fixed end and the self-stabilizing platform; Represents the integral control gain coefficient; This represents the relative angular deviation vector between the fixed end and the self-stabilizing platform; Indicates the time variable of integration; This represents the viscous damping coefficient matrix caused by the universal joint bearing and air damping. This represents the absolute angular velocity vector of the self-stabilizing platform relative to inertial space; This represents the magnitude of the Coulomb frictional torque existing within the system; Represents a symbolic function.

4. The method for monitoring micro-deformation of a dam body according to claim 1, characterized in that, It also includes displacement measurement using multi-beam laser interferometry: In step S5, the non-contact displacement detection array uses a laser interferometry subsystem; this subsystem consists of three orthogonally arranged laser emitting probes and corresponding position-sensitive detectors or interference fringe receivers, which respectively measure the minute displacement components of the fixed end target surface in the X, Y, and Z axes; By using the stability of the laser wavelength as a benchmark for length measurement, the relative displacement value with nanometer-level resolution is obtained by analyzing the phase difference or the amount of movement of interference fringes between the reflected light wave and the reference light wave; at the same time, the system corrects the air refractive index on the optical path in real time.

5. The method for monitoring micro-deformation of a dam body according to claim 1, characterized in that, It also includes adaptive triggering logic based on energy spectral density: In step S6, in order to avoid false triggering caused by environmental background noise, the system does not simply rely on a single displacement amplitude threshold, but introduces a signal frequency domain energy analysis mechanism; by performing a sliding window fast Fourier transform on the original displacement data stream, the energy spectral density characteristics within a specific frequency band are extracted; when a sudden jump in energy is detected within the frequency range of the dam deformation characteristics, and the jump amplitude exceeds a preset multiple of the background noise base energy, it is determined to be a valid deformation event and data locking is triggered; after triggering, the system will automatically extend the recording time window forward and backward, and downsample and compress the data in the non-trigger period for storage.

6. The method for monitoring micro-deformation of a dam body according to claim 1, characterized in that, The process also includes spatial coordinate transformation and error correction of the measurement data: In step S7, the measured relative displacement vector needs to be mapped from the instrument coordinate system to the dam's structural coordinate system; a quaternion rotation matrix is ​​used to perform a spatial rotation transformation on the original displacement vector, while introducing a nonlinear deviation correction term caused by temperature drift and mechanical installation errors; the true deformation vector of the dam body in the geographic coordinate system is calculated using the following spatial coordinate solution formula. : ; in, This represents the true deformation and displacement vector of the dam body monitoring points in the geographic coordinate system after correction; This represents the rotation transformation matrix from the instrument measurement coordinate system to the geographic coordinate system, which is determined by the system's mounting Euler angles. Decide; This represents the inverse operation of a matrix; This represents the original relative displacement vector measured by the displacement detection array; This represents the high-frequency random noise vector of the system estimated by the filtering algorithm; This represents the lever arm vector from the probe center to the center of rotation of the fixed end; This represents the vector cross product operation; This represents the small angular velocity vector remaining after compensation on the self-stabilizing platform; Indicates the sampling time interval.

7. The method for monitoring micro-deformation of a dam body according to claim 1, characterized in that, It also includes a redundant measurement mechanism based on the capacitive sensing principle: In step S5, a differential capacitive displacement sensor or eddy current sensor is arranged in the slit between the fixed end and the self-stabilizing platform; the sensor uses the change in capacitance caused by the change in distance between the two plates, or the change in impedance caused by eddy current induced on the surface of the metal conductor, to invert the small displacement; in the data processing unit, the analog voltage signal obtained by the capacitive / inductive sensor is converted from analog to digital and then weighted and fused with the optical measurement data; when the optical path is blocked by dust or water mist, the system automatically increases the weight of the capacitive sensing data.

8. The method for monitoring micro-deformation of a dam body according to claim 1, characterized in that, It also includes distributed clock synchronization and trigger network coordination: In step S6, when a single monitoring point meets the triggering conditions, the monitoring node sends a coordinated trigger broadcast signal to adjacent monitoring nodes deployed in other locations on the dam through an industrial wireless network or fiber optic ring network; all nodes that receive the broadcast signal will synchronously start the high-frequency data acquisition mode even if they have not reached the local trigger threshold, and use the IEEE 1588 precise time protocol to align the data timestamps of all nodes at the sub-microsecond level.

9. The method for monitoring micro-deformation of a dam body according to claim 1, characterized in that, It also includes multidimensional correlation analysis and elimination of environmental impact factors: In step S7, the data processing module not only calculates the geometric deformation, but also simultaneously reads the temperature field distribution, reservoir water level and atmospheric pressure data on the dam surface; it uses multiple regression analysis or machine learning algorithms to establish a benchmark correlation model between environmental quantities and deformation quantities; when calculating the final deformation quantity, it first uses the model to deduct the conventional periodic components caused by thermal expansion and contraction and hydraulic elastic deformation, and separates the non-correlated abnormal micro-deformation mutations.

10. The method for monitoring micro-deformation of a dam body according to claim 1, characterized in that, It also includes trend prediction and health score based on historical data: After step S7, the system stores the calculated real micro-deformation into the historical database and constructs a time series prediction model; the model uses autoregressive integral moving average or long short-term memory network to extrapolate the future deformation trend of the dam body. The system calculates the structural health score of the dam in real time based on the degree of deviation between the measured deformation and the predicted value; When the score falls below the safety warning line or the predicted deformation rate shows an exponential growth trend, the system automatically generates a graded alarm signal and sends a diagnostic report to the dam safety monitoring center.

Images