Ice-pulling-resistant system and method for concrete dam flower-hanging pipe in high-cold region

Abstract

The invention discloses an ice-pulling-resistant system and method for a concrete dam hanging pipe in an alpine region, and relates to the technical field of hydraulic engineering, and the system comprises an intelligent monitoring module which employs a distributed optical fiber strain sensor group to capture the micro-strain of a pipe body in real time, and combines an ice layer thickness radar, a vertical temperature measurement chain and a high-dynamic water level gauge; an ice-water-structure multi-physical field sensing network is constructed; according to the dynamic regulation and control module, a phase change material heat blocking layer is arranged on the pipe wall, and solid-liquid phase change is induced through pulse heating to generate an interface lubricating film; the anchoring end is provided with a gradient stiffness unit, and a negative Poisson's ratio metamaterial cushion layer and a shape memory alloy damper are connected in series, so that the stiffness is increased in an exponential level along with the ice pulling force; and the edge calculation control module is used for fusing multi-source data based on an LSTM-physical hybrid model, generating an ice pulling force predicted value, dynamically triggering graded response, starting PCM heating to maintain the flexibility of a cushion layer in primary response, activating SMA rigidity jump in advanced response and cooperating with high-voltage electric pulse stripping.

Description

Technical Field

[0001] The present invention relates to the technical field of water conservancy projects, and in particular to an anti-ice pullout system and method for hanging flower pipes of concrete dams in high-altitude cold regions. Background Art

[0002] Cold regions such as Northeast and Northwest China and the Qinghai-Tibet Plateau experience low winter temperatures, often below -10°C. The freezing period can last up to four to six months, with large temperature swings between day and night and frequent freeze-thaw cycles. Concrete dams are subject to severe ice damage. Pipelines for water supply, drainage, and monitoring (such as pipes) within or around the dam are susceptible to the frost heave force of the ice. The ice expands during freezing, exerting upward "ice pull" forces on the pipes, potentially causing pipe deformation, cracking, loosening of joints, and even compromising the structural safety of the dam.

[0003] The adhesion of ice to the pipe surface, the volume expansion of ice as it freezes, and the compression of the pipe by the frost heave of the soil and dam material all contribute to upward pull forces. The impact of ice pull is even more pronounced on porous pipes due to their rough surfaces and pores, which are prone to water accumulation and freezing. As the core structure of a concrete dam, the reliability of its piping system directly impacts dam seepage control, water level monitoring, and operation and maintenance. Therefore, the development of specialized protection systems against ice pull hazards is crucial.

[0004] However, there are still problems in the existing technology that need to be solved urgently: the ice pullout calculation model is inaccurate, and the traditional theoretical formula ignores the coupling effects of ice layer heterogeneity, temperature gradient and dynamic water level, resulting in ice pullout prediction value deviation exceeding 30%, redundant or insufficient design safety factors, and difficult to guarantee structural reliability. Summary of the Invention

[0005] In order to solve the above technical problems, a system and method for anti-ice pullout of hanging flower pipes of concrete dams in high-altitude cold areas are provided. This technical solution solves the problem of inaccurate ice pullout force calculation model mentioned above.

[0006] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0007] An anti-ice pullout system for concrete dam hanging pipes in high-altitude cold regions, comprising:

[0008] Intelligent monitoring module: A distributed fiber optic strain sensor group embedded in the axial direction of the hanging pipe collects real-time micro-strain of the pipe under the action of ice pullout force; a temperature-water level coupled sensor array deployed in the ice layer in front of the dam, including an ice thickness radar, ice temperature gradient probe and water level fluctuation meter, dynamically monitors ice layer heterogeneity, temperature gradient and reservoir water level changes;

[0009] Dynamic Control Module: The phase change material thermal barrier layer covering the outer wall of the hanging flower tube triggers a solid-liquid transition through an active heating unit to remove ice adhesion. The gradient stiffness anchor unit at the anchor end of the tube body consists of a low-stiffness elastic pad and a high-stiffness damper in series. The damper stiffness increases exponentially with the increase of ice pullout force.

[0010] Edge computing control module: Receives data from the intelligent monitoring module, executes the ice pull force fusion algorithm, and dynamically adjusts the power of the active heating unit and the stiffness threshold of the damper based on the Fice prediction value.

[0011] Preferably, the intelligent monitoring module specifically includes:

[0012] Distributed fiber optic strain sensor group: The optical fiber is buried in the protective layer of the pipe wall along the axial direction of the hanging pipe, with measuring points arranged at predetermined intervals to fully cover the pipe. It uses low-temperature resistant single-mode optical fiber and is coated with a polyimide coating to resist frost heave and extrusion. Sensor types include FBG arrays and BOTDR systems.

[0013] By capturing the axial strain gradient of the pipe in real time, the ice pull-out force distribution is monitored; by using displacement resolution, the micro-deformation of the pipe is monitored; and by using the accumulated strain measurement error, the freeze-thaw cycle damage is monitored.

[0014] Preferably, the intelligent monitoring module specifically includes:

[0015] Temperature-water level coupled sensor array:

[0016] Ice thickness radar uses frequency modulated continuous wave radar. Multiple radar nodes are deployed in the waters in front of the dam to form a monitoring grid, which scans the entire area regularly. It also outputs ice thickness non-uniformity parameters.

[0017] Ice temperature gradient monitoring chain, the vertical temperature measurement chain is composed of multiple PT1000 platinum resistance sensors, which calculate the temperature difference between the upper and lower surfaces in real time and generate thermal stress components based on the ice thermal expansion coefficient; the probe shell is made of titanium alloy and the base is counterweighted;

[0018] High dynamic water level fluctuation meter, the main sensor is a radar water level gauge, the auxiliary pressure compensated water level sensor, the key output is the water level change rate, which is used for buoyancy component modeling.

[0019] Preferably, the dynamic control module specifically includes:

[0020] Phase change material heat blocking layer: The matrix is ​​microencapsulated paraffin, encapsulated in a silica reinforced skeleton to form a flexible film layer, and doped with carbon nanotubes;

[0021] Active heating unit: Pulsed carbon fiber heating wire embedded in the PCM layer; the trigger mechanism is that when the predicted ice pullout force value exceeds the established threshold, the power is turned on to raise the temperature to a set temperature, inducing the paraffin wax to transform from solid to liquid; the liquid paraffin wax forms a micron-level lubricating film at the interface between the pipe wall and the ice layer, and floats in conjunction with the water level.

[0022] Preferably, the dynamic control module specifically includes:

[0023] Gradient stiffness anchoring unit: low-stiffness elastic cushion, made of fluororubber foam, absorbs low-frequency vibration energy in the initial stage of ice pullout; high-stiffness damper, a shape memory alloy variable stiffness subunit; stiffness switching mechanism, including SMA in the austenitic state at room temperature, providing basic stiffness; when the Fice prediction value exceeds a predetermined threshold, power is applied to stimulate martensitic phase transformation, causing the elastic modulus to jump;

[0024] Anti-frost heave integrated, low-temperature phase change protection, SMA driver built-in PTC heater; anti-icing plug seal, moving parts are fully encapsulated with bellows and injected with low-temperature silicone grease.

[0025] Preferably, the dynamic control module specifically includes:

[0026] A hierarchical response strategy: in the primary response, the PCM heating unit is activated to remove ice adhesion; the elastic cushion layer maintains a low stiffness state to dissipate energy; in the advanced response, the PCM heating power is increased, the SMA damper high stiffness mode is activated to limit structural displacement, and encrypted data sampling is requested from the monitoring module;

[0027] Fail-safe mechanism, PCM layer failure backup, if the temperature sensor detects heating failure, switches to electric pulse de-icing mode; damper overload protection, when the displacement exceeds the established threshold, triggers the hydraulic buffer emergency braking.

[0028] Preferably, the edge computing control module specifically includes:

[0029] Ice pullout force fusion prediction model: real-time coupling of input parameters, receiving micro-strain data of the pipe body from the optical fiber strain sensor group, and dynamically correcting the ice-pipe contact area coefficient; integrating the non-uniform thickness of the ice radar, the gradient temperature difference of the ice temperature probe, and the rate of change of the water level gauge;

[0030] Generate ice pull force predictions:

[0031]

[0032] Where, F iceFice is the ice pull force prediction value, which represents the upward force generated by the ice on the object; K is the dynamic correction coefficient of the optical fiber strain data, which is determined by the optical fiber strain data and is used to correct the ice pull force prediction value; a is a temperature-related coefficient, which reflects the degree of influence of temperature change on the ice pull force; ΔT is the temperature change, which represents the change of the ice or the surrounding environment temperature within a certain time; δ is the thickness of the ice, which affects the strength and stability of the ice and further affects the size of the ice pull force; is the integral of the elastic modulus of the ice in the water depth direction, wherein E ice (z) is a function of the change of the elastic modulus of the ice with depth z, and the upper limit h of the integral represents the water depth; β is a water level change-related coefficient, which characterizes the influence mechanism of water level change on the ice pull force; is the water level change rate, which represents the change rate of the water level with time, and the unit is meters per second; the change of the water level changes the pressure balance between the ice and the water, affecting the ice pull force; ρ ice is the density of the ice, with the unit of kilograms per cubic meter, which is related to the material properties of the ice and affects the weight of the ice and the force acting on the object; g is the acceleration of gravity.

[0033] Preferably, the edge computing control module specifically comprises:

[0034] An adaptive learning mechanism, an LSTM time series prediction network, the input layer is the temperature change, the water level change rate and the strain peak value sequence within a certain time in the past; the output layer is the Fice prediction value and the confidence interval in a short time in the future; the parameters are updated online, and after completing an ice pull event, the back propagation algorithm adjusts the temperature-related coefficient, the water level change-related coefficient and the dynamic correction coefficient of the optical fiber strain data.

[0035] Preferably, the edge computing control module specifically comprises:

[0036] An edge hardware platform: the main control chip is NVIDIA Jetson AGX Orin, which realizes real-time operation based on the PREEMPT_RT patch of the Linux kernel; a storage architecture, the local RAM is 4TBNVMe SSD cache for short-term storage of raw data; cloud synchronization, key events are encrypted and uploaded to the center server daily;

[0037] Extreme environment protection, triple sealing for stainless steel shell, inert gas filling and silica gel waterproof interface; self-sustaining power supply, the main power supply is photovoltaic-diesel hybrid power supply, and the backup power supply is a super capacitor group.

[0038] Further, an ice pull resistance method for a hanging flower pipe of a concrete dam in an alpine region is used to realize an ice pull resistance system for a hanging flower pipe of a concrete dam in an alpine region as described above, comprising:

[0039] A distributed optical fiber sensor group embedded in the axial protection layer of the hanging flower pipe is used to collect micro-strains under the action of the ice pull force in real time.

[0040] Ice layer parameter collection: ice thickness radars are deployed in the grid in front of the dam to output non-uniformity parameters; chain temperature probes are inserted into the ice layer to measure vertical temperature differences; millimeter-wave radar water level gauges capture the rate of water level change;

[0041] The non-uniformity parameter is calculated by the standard deviation of ice thickness measured by radar grid:

[0042]

[0043] Where δ is the non-uniform thickness of the ice layer, which is calculated from ice thickness data measured by radar grids and reflects the degree of discreteness of the ice thickness; To take the reciprocal operation term, h is the water depth in meters, and the subsequent standard deviation calculation results are scaled so that the ice layer non-uniform thickness index is related to the water depth; is: find the average operation term, n is the total number of samples of ice thickness data measured by radar grid, which is used to calculate the average ice thickness; h i is the ice thickness at the i-th radar grid measurement point, in meters, which is the basic ice thickness measurement data; is the average ice thickness, which comprehensively reflects the average thickness of the ice layer within the radar grid measurement range; n is the total number of samples of ice thickness data measured by the radar grid, that is, the number of points where the radar measures ice thickness within a grid range. It is used to determine the average ice thickness and calculate the total deviation. It is dimensionless.

[0044] Build an edge computing fusion model, hierarchical anti-ice pullout regulation, primary response, trigger pulse heating of the phase change material layer, maintain the elastic cushion modulus of the anchoring system; advanced response, PCM power increase, forced ice melting and peeling, activate the SMA damper stiffness exponential growth.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] This invention proposes embedding distributed optical fibers along the entire axial length of the pipe to form a continuous strain field monitoring network. This overcomes the limitations of traditional point-based measurements in detecting the "blind spot" of ice pullout gradient distribution, accurately capturing pipe micro-deformations and freeze-thaw damage evolution. By integrating radar grid scanning, a vertical temperature measurement chain, and a water level fluctuation meter, a dynamic database of the ice-water-structure coupling field is established, addressing the pain point of traditional methods' invisibility of the internal state of the ice layer. Real-time spatiotemporal alignment and feature extraction of multi-source data are performed on the device side, replacing manual offline analysis and enabling a leap from "delayed judgment" to "millisecond-level prediction."

[0047] Through the solid-liquid phase transition of the PCM material, a micron-scale lubricating film is generated at the interface, enabling flexible ice peeling with extremely low energy consumption, overcoming the "strong adhesion-weak structure" contradiction. The synergistic mechanism of the low-stiffness energy-dissipating underlay and the SMA variable-stiffness damper dissipates energy through the metamaterial structure during the initial ice pullout. A critical state triggers a sudden increase in SMA phase transition stiffness, achieving a dynamic balance of "flexible support and rigid resistance." Combined with backup measures such as water-level buoyancy-assisted peeling and high-voltage electric pulse dielectric breakdown, this creates a fused active-passive defense matrix, significantly enhancing survivability under extreme conditions.

[0048] When PCM fails, it automatically switches to high-voltage pulse de-icing. Piezoelectric variable stiffness fills the gap when SMA delays occur. Radar strain inversion reconnects fiber breaks, achieving "uninterrupted defense." A hybrid photovoltaic-capacitor power supply combined with vibration energy recovery overcomes energy supply bottlenecks in polar infrastructure, ensuring continuous operation in extreme conditions such as snowstorms and road closures. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is an internal framework diagram of a concrete dam hanging pipe anti-ice pullout system in high-altitude cold areas;

[0050] Figure 2 The flowchart of a battery performance detection method that introduces implicit features. DETAILED DESCRIPTION

[0051] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are merely examples, and those skilled in the art may conceive of other obvious variations.

[0052] Reference Figure 1 As shown, a system for resisting ice pullout of concrete dam hanging pipes in high-altitude cold regions includes:

[0053] Intelligent monitoring module: A distributed fiber optic strain sensor group embedded in the axial direction of the hanging pipe collects real-time micro-strain of the pipe under the action of ice pullout force; a temperature-water level coupled sensor array deployed in the ice layer in front of the dam, including an ice thickness radar, ice temperature gradient probe and water level fluctuation meter, dynamically monitors ice layer heterogeneity, temperature gradient and reservoir water level changes;

[0054] Dynamic Control Module: The phase change material thermal barrier layer covering the outer wall of the hanging flower tube triggers a solid-liquid transition through an active heating unit to remove ice adhesion. The gradient stiffness anchor unit at the anchor end of the tube body consists of a low-stiffness elastic pad and a high-stiffness damper in series. The damper stiffness increases exponentially with the increase of ice pullout force.

[0055] Edge computing control module: Receives data from the intelligent monitoring module, executes the ice pull force fusion algorithm, and dynamically adjusts the power of the active heating unit and the stiffness threshold of the damper based on the Fice prediction value.

[0056] It should be noted that in the real-time perception-decision-execution chain, the distributed optical fiber captures the pipe strain at a 10Hz sampling rate → the edge computing unit completes the ice pull force fusion calculation within 200ms → the dynamic control module responds according to the dual-threshold strategy:

[0057] Primary response layer (0.7F c ≤Fice<0.9F c ): Start fractal topology heating of the PCM layer, linearly adjust the power, and simultaneously maintain the low stiffness state of the anchoring pad, allowing 20mm displacement buffer;

[0058] Critical response layer (Fice ≥ 0.9F c ): 100% of the PCM power melts the ice, triggering the exponential stiffness growth of the SMA damper and forcing the locking displacement to ≤5mm.

[0059] Cross-module data penetration: When the ice temperature gradient probe detects an abnormal jump in ΔT (>5°C / h), it automatically increases the fiber optic sampling rate to 50Hz and corrects the Fice prediction value in advance through edge computing.

[0060] Designed for extreme environment adaptability:

[0061] The -50℃ combined protection system uses a polyimide-nanoceramic composite coating on the optical fiber sensing group (3.2% elongation at break at -60℃), and silicon-based aerogel is embedded in the surface microgrooves to block frost heave stress. The PCM layer matrix is ​​upgraded to a fatty acid / expanded graphite-MXene composite (phase change enthalpy ≥ 210J / g, -45℃ viscosity <10mPa·s) to ensure low-temperature fluidity. The SMA damper moving pair is deposited with a diamond-like carbon film (friction coefficient 0.02) and filled with perfluoropolyether oil (pour point -70℃) to eliminate the cold welding effect.

[0062] Anti-electromagnetic interference capability. The edge computing main control chip (Kunpeng 920) is equipped with an electromagnetic shielding alloy cover (attenuation ≥ 60dB), and the sensor communication adopts differential orthogonal frequency division multiplexing (DQPSK modulation), with a bit error rate of <10 -9 .

[0063] The intelligent monitoring module specifically includes:

[0064] Distributed fiber optic strain sensor group: The optical fiber is buried in the protective layer of the pipe wall along the axial direction of the hanging pipe, with measuring points arranged at predetermined intervals to fully cover the pipe. It uses low-temperature resistant single-mode optical fiber and is coated with a polyimide coating to resist frost heave and extrusion. Sensor types include FBG arrays and BOTDR systems.

[0065] By capturing the axial strain gradient of the pipe in real time, the ice pullout force distribution is monitored; the micro-deformation of the pipe is monitored through displacement resolution; and the freeze-thaw cycle damage is monitored through the accumulated strain measurement error.

[0066] Temperature-water level coupled sensor array:

[0067] Ice thickness radar uses frequency modulated continuous wave radar. Multiple radar nodes are deployed in the waters in front of the dam to form a monitoring grid, which scans the entire area regularly. It also outputs ice thickness non-uniformity parameters.

[0068] Ice temperature gradient monitoring chain, the vertical temperature measurement chain is composed of multiple PT1000 platinum resistance sensors, which calculate the temperature difference between the upper and lower surfaces in real time and generate thermal stress components based on the ice thermal expansion coefficient; the probe shell is made of titanium alloy and the base is counterweighted;

[0069] High dynamic water level fluctuation meter, the main sensor is a radar water level gauge, the auxiliary pressure compensated water level sensor, the key output is the water level change rate, which is used for buoyancy component modeling.

[0070] It should be noted that the distributed optical fiber strain sensor group:

[0071] Fiber-enhanced protection system, the core adopts fluorine-doped radiation-resistant single-mode fiber (attenuation ≤ 0.18dB / km@1550nm, additional loss at -60℃ < 0.03dB / km);

[0072] The coating was upgraded to a polyimide-nano ceramic composite layer (thermal expansion coefficient 1.5×10 -6 / ℃), covered with a stainless steel micro-tube corrugated sheath (filled with silicon-based aerogel, thermal conductivity 0.018W / m·K) to block the transmission of frost heave stress;

[0073] Shape memory polymer buffer rings (Tg = -30°C) are set at key anchoring points to compensate for the thermal deformation difference between concrete and optical fiber.

[0074] Dual-mode sensing collaboration, FBG array (local high precision), femtosecond laser-written FBGs (reflectivity 90%, bandwidth 0.2nm) are arranged every 0.5m, and the wavelength demodulation accuracy is ±0.5pm. Dual-wavelength reference technology (1540nm + 1580nm) is used to eliminate temperature-strain cross-sensitivity.

[0075] BOTDR system (full field coverage), pulse width 1ns (spatial resolution 1m), strain measurement range ±15000με; innovative frequency shift accumulation averaging algorithm (1000 scans), improves signal-to-noise ratio to >60dB;

[0076] Axial strain gradient monitoring, based on the one-dimensional stress wave equation, inverses the spatial distribution of ice pullout force:

[0077]

[0078] Where ε is the axial strain, which indicates the degree of deformation of the material in the axial direction, and its unit is usually infinite. x is the spatial position variable, which represents the position along the axis of the object, and the unit is meter; t is the time variable, and the unit is second; v p is the longitudinal wave velocity, which indicates the speed at which stress waves propagate in the material, in meters per second;

[0079] Quantification of freeze-thaw damage and definition of cumulative damage index:

[0080]

[0081] Where D f is the cumulative damage index, which is used to quantify the degree of cumulative damage caused by freeze-thaw cycles; N is the number of freeze-thaw cycles, which indicates the total number of freeze-thaw cycles experienced by the material; is the maximum strain amplitude of the i-th freeze-thaw cycle, indicating the maximum deformation degree of the material in a single freeze-thaw process; ε y is the yield strain of the pipe, which indicates the strain value at which the pipe begins to undergo plastic deformation; m is the material constant, which is 3.2 and is used to reflect the damage characteristics of the material; when D f When the value is greater than 0.6, a structural maintenance warning is triggered.

[0082] Temperature-water level coupled sensor array:

[0083] The ice thickness radar is anti-interference, and the transmitter uses K-band frequency-modulated continuous wave (24-26GHz, bandwidth 2GHz), with a range resolution of 7.5cm; the antenna array is configured with MIMO topology (4 transmitters and 8 receivers), and the beam scanning angle is ±60°.

[0084] Intelligent signal processing, embedded polarization synthetic aperture algorithm, horizontal polarization (HH) identification of ice-air interface; cross polarization (HV) suppression of wet snow clutter (suppression degree > 20dB); ice thickness inversion model introduces dynamic correction of dielectric constant: ε' r =3.17-0.009T ice +7.5×10 -5 S

[0085] Where ε′ r is the dielectric constant correction value, which is used to describe the changes in the dielectric properties of ice under specific conditions; T ice is the ice temperature, which indicates the temperature of the ice in degrees Celsius; S is the salt content in ice in ppm, which indicates the salt content in ice. The salt content has a significant effect on the dielectric constant of ice because it affects the conductivity and polarization properties of ice;

[0086] Strengthening the ice temperature gradient monitoring chain

[0087] Sensor optimization, PT1000 platinum resistance packaged in titanium alloy capillary (wall thickness 0.2mm), filled with fluorinated liquid heat-conducting glue (thermal conductivity 8W / m·K); base configuration tungsten alloy counterweight (density 18g / cm 3 ), ensure vertical attitude error <1°;

[0088] Thermal stress component modeling, real-time calculation of vertical temperature difference: ΔT=T top -T bottom

[0089] Output equivalent thermal stress: σ thermal =E ice ·α ice ·ΔT·h

[0090] In the formula, ΔT is the vertical temperature difference, indicating the difference between the top and bottom temperatures of the ice layer, in degrees Celsius; T top is the temperature at the top of the ice layer; T bottom is the temperature at the bottom of the ice layer; σ thermal is the equivalent thermal stress, indicating the stress inside the ice layer due to temperature changes; E ice is the dynamic elastic modulus of ice, used to describe the elastic properties of ice under stress; α ice is the thermal expansion coefficient of ice, indicating the expansion or contraction of ice when the temperature changes; h is the thickness of the ice layer, in meters, affecting the size of the thermal stress, which is usually determined by actual measurement;

[0091] High-dynamic water level monitoring redundancy design, main sensor (radar water level meter), working frequency band 78GHz (bandwidth 3GHz); wave interference algorithm, adaptive Kalman filter (state vector contains wave period / amplitude).

[0092] Auxiliary sensor (barometric compensation type), double pressure probe differential deployment (underwater 5m + above water surface), eliminating the influence of atmospheric pressure fluctuations; sampling rate 10Hz, range ±2m / s.

[0093] The dynamic regulation module specifically includes:

[0094] Phase change material thermal resistance layer: the base is microencapsulated paraffin, packaged in a silica reinforced skeleton to form a flexible film layer, and doped with carbon nanotubes;

[0095] Active heating unit: pulse type carbon fiber heating wire embedded in the PCM layer; the trigger mechanism is to heat to a certain temperature when the ice pulling force prediction value is greater than a certain threshold, inducing the solid-liquid conversion of paraffin; liquid paraffin forms a micron-level lubricating film at the pipe wall-ice layer interface, and cooperates with the water level float;

[0096] Gradient stiffness anchoring unit: low-stiffness elastic cushion, made of fluororubber foam, absorbs low-frequency vibration energy in the initial stage of ice pullout; high-stiffness damper, a shape memory alloy variable stiffness subunit; stiffness switching mechanism, including SMA in the austenitic state at room temperature, providing basic stiffness; when the Fice prediction value exceeds a predetermined threshold, power is applied to stimulate martensitic phase transformation, causing the elastic modulus to jump;

[0097] Anti-frost heave integrated, low-temperature phase change protection, SMA driver built-in PTC heater; anti-ice plug seal, moving parts are fully encapsulated with bellows and injected with low-temperature silicone grease;

[0098] A hierarchical response strategy: in the primary response, the PCM heating unit is activated to remove ice adhesion; the elastic cushion layer maintains a low stiffness state to dissipate energy; in the advanced response, the PCM heating power is increased, the SMA damper high stiffness mode is activated to limit structural displacement, and encrypted data sampling is requested from the monitoring module;

[0099] Fail-safe mechanism, PCM layer failure backup, if the temperature sensor detects heating failure, switches to electric pulse de-icing mode; damper overload protection, when the displacement exceeds the established threshold, triggers the hydraulic buffer emergency braking.

[0100] It should be noted that the phase change material (PCM) thermal blocking layer:

[0101] Microencapsulation: The paraffin matrix uses a docosanoic acid-palmitic acid eutectic system (phase transition temperature -3°C), encapsulated in a mesoporous silica shell (pore diameter 5nm, wall thickness 200nm) to improve leakage resistance. The skeleton reinforcement material is changed to a silicon carbide nanowire mesh (tensile strength 3GPa), replacing traditional silica particles, and the thermal conductivity path forms a three-dimensional interpenetrating network (thermal conductivity 6.8W / m·K).

[0102] Nano-doping strengthening: carbon nanotubes are upgraded to boron nitride nanotube-graphene hybrid fillers (mass ratio 2:1) to achieve anisotropic thermal conductivity (axial thermal conductivity > 400W / m·K); MXene quantum dots (Ti3C2Tx, size 2-5nm) are added to form an atomic-level lubrication layer at the ice-tube interface (the friction coefficient is reduced to 0.03).

[0103] Active heating unit:

[0104] Pulse heating topology, carbon fiber filaments are arranged using a fractal Hilbert curve (filling rate 92%), achieving a temperature uniformity of ±0.3°C on the tube surface; the drive circuit adds a zero voltage switching (ZVS) resonant topology, achieving an energy conversion efficiency of >95%;

[0105] Interface peeling dynamics guarantee, laser engraving micron-scale pit array on the tube wall (diameter 50μm, depth 20μm, density 10 4 pcs / m2 ), storing liquid paraffin to form a secondary lubricating film; in conjunction with the buoyancy of the water level, when the radar water level gauge detects dh / dt>5mm / min, the heating power is actively reduced by 20%, and the rising water flow is used to assist in ice peeling.

[0106] Fail-safe mechanism is strengthened, electric pulse de-icing backup is provided, electrodes adopt interdigitated interdigital structure (spacing 0.5mm), 10kV / 100μs high voltage pulse is applied to induce dielectric breakdown of ice layer (energy consumption is only 0.05kWh / m 2 ); Failure criterion: if the temperature sensor detects ΔT < 1 / min and the strain continues to increase, it switches to the electric pulse mode.

[0107] Gradient stiffness anchor unit:

[0108] Low-rigidity elastic cushion, fluororubber foam is changed to a negative Poisson's ratio (-0.3) honeycomb structure, the radial expansion rate is greater than 15% when compressed, actively filling the concrete pores to block water seepage; built-in piezoelectric fiber array (PZT-5H, d 33 =650pC / N), converting vibration energy into electrical energy;

[0109] Low-frequency energy dissipation is optimized by adding a magnetorheological elastomer interlayer (carbonyl iron powder volume ratio 30%). The damping coefficient can be adjusted within the range of 5-50kN·s / m under magnetic field control.

[0110] High stiffness SMA damper, variable stiffness core mechanism, SMA wire surface coated with diamond nano-cone array (cone height 200nm, spacing 500nm), phase change cycle life> 10 7 Phase change trigger logic: When Fice ≥ 0.9, a 5A / 100ms current pulse is applied, and the austenite → martensite phase change response time is ≤ 0.5 seconds; the stiffness function is precisely controlled through closed-loop feedback from the laser displacement sensor;

[0111] Extreme environment protection and low-temperature phase change protection: PTC heaters are embedded in the micro-channel thermal management module (the working fluid is perfluorohexanone), with a startup time of less than 10 seconds at -40°C; SMA wire pre-loaded with two-way shape memory training (strain rate 4%) to avoid low-temperature phase change hysteresis;

[0112] Anti-ice plug sealing solution, bellows material upgraded to Nitinol-stainless steel composite layer (fatigue life> 5×10 8 perfluoropolyether-based low-temperature silicone grease (viscosity <500cP at -70°C) is injected, and a diamond-like carbon film (friction coefficient 0.02) is deposited on the surface of the moving pair.

[0113] Tiered Response Strategy:

[0114] Primary response (0.7F c ≤F ice<0.9F c ), PCM heating strategy:

[0115] P=50+20(F ice / F c -0.7)(kW / m 2 )

[0116] Where, P is the dynamic adjustment power; F ice is the predicted value of ice pullout force; F c is the critical ice pullout force value;

[0117] The target interface temperature is ≥2°C, verified online by an infrared thermal imager (sampling rate 30Hz);

[0118] Anchoring system status: elastic cushion modulus maintained ≤ 5MPa, allowing pipe displacement ≤ 20mm; magnetorheological interlayer activates low damping mode (c = 5kN·s / m), dissipating energy > 60%;

[0119] Advanced Response (F ice ≥0.9F c ), the PCM power is maximized and the heating power is increased to 100%, forcing ice to melt and peel off; if the strain does not decrease within 120 seconds, the electric pulse de-icing synergy (10kV pulse superposition) is triggered;

[0120] The SMA damper is activated and the stiffness increases exponentially to the design maximum value (k max =85GPa); send encrypted sampling instructions to the monitoring module: the fiber optic sampling rate is increased to 100Hz, and the radar scanning frequency is increased to 1 time / 10 seconds.

[0121] Fail-safe layered defense:

[0122] Damper overload protection, when the displacement is greater than 8mm, the magnetorheological fluid emergency brake is triggered (yield stress 500kPa, response <20ms); the hydraulic buffer adopts a multi-stage collapse design (peak energy absorption density 15J / cm 3 );

[0123] System-level redundancy: if the SMA phase change fails, it switches to the piezoelectric active variable stiffness mode (maximum output 200kN); the backup power supply uses a supercapacitor group (capacity 100F, -50℃ capacity retention rate >90%).

[0124] The edge computing control module specifically includes:

[0125] Ice pullout force fusion prediction model: real-time coupling of input parameters, receiving micro-strain data of the pipe body from the optical fiber strain sensor group, and dynamically correcting the ice-pipe contact area coefficient; integrating the non-uniform thickness of the ice radar, the gradient temperature difference of the ice temperature probe, and the rate of change of the water level gauge;

[0126] Generate ice pull force predictions:

[0127]

[0128] Where, F ice is the predicted value of ice pullout force, which represents the upward force exerted by ice on an object; K is the dynamic correction coefficient of optical fiber strain data, which is determined by optical fiber strain data and is used to correct the predicted value of ice pullout force; α is a temperature-related coefficient that reflects the degree of influence of temperature change on ice pullout force; ΔT is the temperature change, which represents the change in ice or ambient temperature over a certain period of time; δ is the thickness of ice, which affects the strength and stability of ice, and thus the magnitude of ice pullout force; is the integral of the elastic modulus of ice in the direction of water depth, where E ice (z) is the elastic modulus of ice as a function of depth z, and the upper limit of integration h represents the water depth; β is a coefficient related to water level change, which characterizes the mechanism of the influence of water level change on ice pullout force; is the water level change rate, which indicates the rate of change of water level over time, in meters per second. Water level changes will change the pressure balance between ice and water, affecting the ice pulling force; ρ ice is the density of ice, measured in kilograms per cubic meter, which is related to the material properties of ice and affects its weight and the force it exerts on objects; g is the acceleration due to gravity;

[0129] Adaptive learning mechanism, LSTM time series prediction network, the input layer is the temperature change, water level change rate, and strain peak sequence within a given time period in the past; the output layer is the Fice prediction value and confidence interval within a short period of time in the future; the parameters are updated online, and after each ice pull event, the back propagation algorithm adjusts the coefficients related to temperature, the coefficients related to water level change, and the dynamic correction coefficient value of the optical fiber strain data;

[0130] Edge hardware platform: The main control chip is NVIDIA Jetson AGX Orin, which operates in real time and is based on the PREEMPT_RT patch of the Linux kernel. The storage architecture uses 4TB of local RAM NVMe SSD to cache short-term raw data. Cloud synchronization is used to upload key events to the central server with daily encryption.

[0131] Extreme environment protection, triple seal for stainless steel shell, inert gas filling, silicone waterproof interface; self-sustaining power supply, the main power supply is the dam body photovoltaic-diesel hybrid power supply, and the backup power supply is a supercapacitor group.

[0132] It should be noted that the multi-physics coupling algorithm enhances:

[0133] Dynamic contact area correction mechanism: Fiber optic strain data is analyzed through strain energy density gradient analysis to reflect the contact surface changes caused by uneven ice adhesion in real time:

[0134]

[0135] Where A c is the corrected contact area, usually in square meters, indicating the actual contact area; A0 is the nominal contact area, in square meters, indicating the theoretical or initial contact area; γ is the correction factor, expressed as 0.15; L is the upper limit of integration, indicating the length range of the optical fiber strain measurement; ε is the strain, indicating the degree of deformation of the material; x is the spatial position variable, indicating the direction along the optical fiber;

[0136] Deep integration optimization of elastic modulus, introducing ice crystal orientation compensation factor:

[0137]

[0138] Where, E eff is the effective elastic modulus, in Pascals, which represents the elastic modulus corrected to take into account the anisotropy of the ice layer; E(z) is the elastic modulus that varies with depth, in Pascals, which represents the elastic modulus of the ice layer at depth z; h is the thickness of the ice layer; z is the depth variable; θ(z) is the lattice orientation angle, obtained by the ice temperature probe, which represents the orientation angle of the ice crystal at depth z and is used to compensate for the influence of ice anisotropy;

[0139] The buoyancy-thermal stress synergy term and the new water level-temperature cross term are used to quantify the thermal shock effect during sudden water level changes:

[0140]

[0141] Where, F cross is the water level-temperature cross-term force, in kilonewtons, which represents the quantitative value of the thermal shock effect when the water level changes suddenly; β is the cross-term coefficient, which is used to relate the water level change rate and the temperature change amplitude; is the water level change rate, in meters per second, indicating the rate of change of water level over time; ΔT is the temperature change amplitude, in degrees Celsius, indicating the temperature difference;

[0142] Real-time parameter calibration engine, fiber strain credibility verification, compares BOTDR and FBG data through Kolmogorov-Smirnov test, and activates radar strain inversion (compensation coefficient α + 0.15) if the deviation is greater than 8%;

[0143] Ice thickness radar data fusion uses three-dimensional Kriging interpolation method to convert discrete grid data into continuous field distribution, and the spatial resolution is improved to 0.2m×0.2m.

[0144] Triple Protection Sealed:

[0145]

[0146] Function Main power supply, photovoltaic input, dam body monocrystalline silicon panel (conversion efficiency 23%), maximum power point tracking (MPPT) efficiency > 99%; diesel backup, 12V DC direct injection diesel engine, -30℃ cold start time < 20 seconds;

[0147] Supercapacitor backup, hybrid capacitor bank (activated carbon / / Li4Ti5O 12 ), capacity 280F; low temperature self-heating: coated ceramic heating plate (-40℃ start, heating rate 15℃ / min).

[0148] Reference Figure 2 As shown, a battery performance detection method introducing implicit features includes:

[0149] The distributed optical fiber sensor group embedded in the axial protective layer of the hanging flower tube is used to collect the micro strain under the action of ice pull force in real time.

[0150] Ice layer parameter collection: ice thickness radars are deployed in the grid in front of the dam to output non-uniformity parameters; chain temperature probes are inserted into the ice layer to measure vertical temperature differences; millimeter-wave radar water level gauges capture the rate of water level change;

[0151] The non-uniformity parameter is calculated by the standard deviation of ice thickness measured by radar grid:

[0152]

[0153] Where δ is the non-uniform thickness of the ice layer, which is calculated from ice thickness data measured by radar grids and reflects the degree of discreteness of the ice thickness; To take the reciprocal operation term, h is the water depth in meters, and the subsequent standard deviation calculation results are scaled so that the ice layer non-uniform thickness index is related to the water depth; is: find the average operation term, n is the total number of samples of ice thickness data measured by radar grid, which is used to calculate the average ice thickness; h i is the ice thickness at the i-th radar grid measurement point, in meters, which is the basic ice thickness measurement data; is the average ice thickness, which comprehensively reflects the average thickness of the ice layer within the radar grid measurement range; n is the total number of samples of ice thickness data measured by the radar grid, that is, the number of points where the radar measures ice thickness within a grid range. It is used to determine the average ice thickness and calculate the total deviation. It is dimensionless.

[0154] Build an edge computing fusion model, hierarchical anti-ice pullout regulation, primary response, trigger pulse heating of the phase change material layer, maintain the elastic cushion modulus of the anchoring system; advanced response, PCM power increase, forced ice melting and peeling, activate the SMA damper stiffness exponential growth.

[0155] It should be noted that the ice layer heterogeneity parameters are:

[0156] The deep effect of the reciprocal term of water depth (1 / h). The traditional ice thickness standard deviation only reflects the absolute discreteness, while the water depth correlation correction reveals the coupling effect of ice layer mechanical behavior and water body constraints:

[0157] Shallow water (h<5m): The water's gravity constraint is weak, the ice layer is prone to bending and breaking, the δ value is magnified to 1.5 times, and the warning threshold is reduced to 8%;

[0158] Deep water area (h>10m): The rigid support of the water body is enhanced, the δ value is reduced to 0.7 times, and a higher discreteness is allowed (threshold 15%).

[0159] The radar grid is dynamically optimized, with a basic grid of 50m×50m, which is automatically encrypted to 10m×10m when δ>10%. Polarimetric interferometric synthetic aperture radar (PolInSAR) is used to invert ice crystal orientation through HH / HV polarization phase difference, thereby improving the accuracy of wet ice identification.

[0160] In summary, the advantages of the present invention are:

[0161] A multi-source fusion perception network consisting of "fiber-optic strain field + ice radar network + temperature chain + water level gauge" was constructed. Distributed optical fiber accurately captured pipe micro-deformation and freeze-thaw damage; ice thickness heterogeneity parameters were integrated with water depth constraints to dynamically assess ice layer structural strength; and temperature-water level cross-sensing enabled quantification of thermal buoyancy coupling effects. An edge computing platform performed real-time data fusion and feature extraction, replacing manual offline analysis and enabling the identification of ice shear risks to move from post-judgment to minute-level prediction.

[0162] The phase-change intelligent thermal blocking layer generates a micro-lubricating film at the interface through solid-liquid phase change, achieving lossless ice removal with 5% of the energy consumption of traditional electric heating; the stiffness programmable anchoring system combines a negative Poisson's ratio energy-absorbing pad with an SMA exponentially variable stiffness damper, which absorbs energy flexibly in the initial stage and locks rigidly in the critical state; the grading strategy triggers multi-modal collaboration such as water level buoyancy assistance and high-pressure pulse de-icing according to the predicted threshold to solve the problem of failure in response to extreme working conditions.

[0163] The sensing end uses fluorine-doped optical fiber anti-radiation coating and titanium alloy armored probe to withstand low temperatures of -55°C and ice impact; the execution end is designed with PCM leakage blocking microstructure and SMA phase change low-temperature protection module to ensure functional stability under freeze-thaw cycles.

[0164] Example:

[0165] Step 1: A distributed optical fiber strain sensor group was buried along the axial direction of the hanging flower tube (tube length 28m). Low-temperature resistant fluorine-doped single-mode optical fiber was used, and FBG measurement points were arranged every 0.5m (wavelength demodulation accuracy ±0.4με). BOTDR units (spatial resolution 1m) were deployed at the root, middle, and end of the tube body.

[0166] Step 2: Set up a temperature-water-level coupled array in the waters in front of the dam: ice thickness radars are arranged in a 50m x 50m grid (K-band frequency-modulated wave, 7.5cm resolution); vertical temperature measurement chains (eight PT1000 platinum resistance thermometers, temperature difference measurement accuracy ±0.1°C) are inserted into ice holes; and radar water level gauges are installed on the maintenance tower (sampling rate 10Hz, accuracy ±1mm).

[0167] Step 3: Install the control module and coat the pipe wall with a PCM thermal barrier layer: microencapsulated docosanoic acid-palmitic acid eutectic (phase transition point -3°C), 3mm thick; embedded carbon fiber heating wires are arranged according to the Hilbert curve (resistance 15Ω / m);

[0168] The anchoring end is equipped with a gradient stiffness unit: a bottom layer of fluororubber negative Poisson's ratio cushion (Poisson's ratio -0.3, thickness 50 mm); an upper layer of SMA damper (NiTi alloy wire diameter 2 mm, prestrain 4%);

[0169] Step 4: Activate the edge platform, load the LSTM-physics hybrid model on NVIDIA Jetson AGX Orin, and initialize the parameters: temperature coefficient λ = 0.35, water level coefficient β = 7.2, strain correction γ = 0.15; set the ice pullout threshold: primary response F c 1=300kN, advanced response F c 2=380kN;

[0170] Step 5: Ice condition data collection: The fiber optic sensor captures strain anomalies: the strain peak in the middle of the pipe reaches +152με (baseline value ±20με);

[0171] Ice radar scans show an average ice thickness of 0.82 m and a standard deviation of σ = 0.18 m. Calculated heterogeneity parameter δ = (1 / h)(σ / μ) = 0.22(0.18 / 0.82) = 4.8%.

[0172] Temperature measurement chain output: ice surface temperature -28°C vs ice bottom temperature -1.2°C → vertical temperature difference ΔT = 26.8°C;

[0173] Water level meter records: rate of change dh / dt = 4.3 mm / min (continuously rising);

[0174] Step 6: Edge intelligent decision-making, multi-source data input LSTM model, output prediction: F ice =328kN (confidence interval 95% ±18kN), triggering the primary response command;

[0175] Step 7: Primary response executed, PCM thermal block layer activated: carbon fiber loading 50kW / m 2Pulse power (duty cycle 60%), the interface temperature rises to 2°C within 10 seconds; liquid paraffin seeps out of the micro-pits, forming a lubricating film with a friction coefficient of 0.03;

[0176] Anchor system configuration: Fluororubber cushion layer maintains low stiffness (5MPa), magnetorheological elastomer damping is set to 5kN·s / m; real-time monitoring strain is reduced to +85με (a 44% reduction);

[0177] Step 8: Advanced response upgrade, PCM power increase: heating power increased to 100kW / m 2 , synchronously triggering a 10kV / 100μs high-voltage pulse (inducing dielectric breakdown of the ice layer);

[0178] SMA damper activation: Applying a 5A current pulse → austenite → martensite phase transformation → exponential increase in stiffness to 58GPa (response time 0.5 seconds);

[0179] Edge command linkage: The fiber optic sampling rate is increased to 100Hz, and the radar scanning frequency is increased to 1 time / minute;

[0180] Step 9: Event termination processing: the PCM layer stops heating, and the SMA damper is demagnetized to restore the foundation stiffness. The edge platform records the event: peak Fice = 412 kN, duration 185 minutes. Backpropagation updates the model parameters: λ → 0.37, β → 7.5 (learning rate 0.001).

[0181] Step 10: Fail-safe response (simulating PCM partial failure), temperature sensor detects abnormality: heating zone ΔT < 1 / min → automatically switches to high-voltage electric pulse de-icing mode; supercapacitor group releases 10kV pulse (energy consumption 0.05kWh / m 2 ) to ensure uninterrupted system operation.

[0182] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions merely illustrate the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A system for preventing ice pullout of concrete dam hanging pipes in high-altitude cold regions, characterized in that: include: Intelligent monitoring module: A distributed fiber optic strain sensor group embedded in the axial direction of the hanging pipe collects real-time micro-strain of the pipe under the action of ice pullout force; a temperature-water level coupled sensor array deployed in the ice layer in front of the dam, including an ice thickness radar, ice temperature gradient probe and water level fluctuation meter, dynamically monitors ice layer heterogeneity, temperature gradient and reservoir water level changes; Dynamic Control Module: The phase change material thermal barrier layer covering the outer wall of the hanging flower tube triggers a solid-liquid transition through an active heating unit to remove ice adhesion. The gradient stiffness anchor unit at the anchor end of the tube body consists of a low-stiffness elastic pad and a high-stiffness damper in series. The damper stiffness increases exponentially with the increase of ice pullout force. Edge computing control module: Receives data from the intelligent monitoring module, executes the ice pull force fusion algorithm, and dynamically adjusts the power of the active heating unit and the stiffness threshold of the damper based on the Fice prediction value.

2. The anti-ice pullout system for concrete dam hanging pipes in high-altitude cold regions according to claim 1 is characterized in that: The edge computing control module specifically includes: Ice pullout force fusion prediction model: real-time coupling of input parameters, receiving micro-strain data of the pipe body from the optical fiber strain sensor group, and dynamically correcting the ice-pipe contact area coefficient; integrating the non-uniform thickness of the ice radar, the gradient temperature difference of the ice temperature probe, and the rate of change of the water level gauge; Generate ice pull force predictions: Where, F ice is the predicted value of ice pullout force, which represents the upward force exerted by ice on an object; K is the dynamic correction coefficient of optical fiber strain data, which is determined by optical fiber strain data and is used to correct the predicted value of ice pullout force; α is a temperature-related coefficient that reflects the degree of influence of temperature change on ice pullout force; ΔT is the temperature change, which represents the change in ice or ambient temperature over a certain period of time; δ is the thickness of ice, which affects the strength and stability of ice, and thus the magnitude of ice pullout force; is the integral of the elastic modulus of ice in the direction of water depth, where E ice (z) is the elastic modulus of ice as a function of depth z, and the upper limit of integration h represents the water depth; β is a coefficient related to water level change, which characterizes the mechanism of the influence of water level change on ice pullout force; is the water level change rate, which indicates the rate of change of water level over time, in meters per second. Water level changes will change the pressure balance between ice and water, affecting the ice pulling force; ρ ice is the density of ice, measured in kilograms per cubic meter. It is related to the material properties of ice and affects its weight and the force it exerts on objects. g is the acceleration due to gravity.

3. The anti-ice pullout system for concrete dam hanging pipes in high-altitude cold regions according to claim 2 is characterized in that: The edge computing control module specifically includes: Adaptive learning mechanism, LSTM time series prediction network, the input layer is the temperature change, water level change rate, and strain peak sequence within a given time in the past; the output layer is the Fice prediction value and confidence interval in the short time in the future; the parameters are updated online, and after each ice pull event, the back propagation algorithm adjusts the coefficients related to temperature, the coefficients related to water level change, and the dynamic correction coefficient value of the optical fiber strain data.

4. The anti-ice pullout system for concrete dam hanging pipes in high-altitude cold regions according to claim 3 is characterized in that: The edge computing control module specifically includes: Edge hardware platform: The main control chip is NVIDIA Jetson AGX Orin, which operates in real time and is based on the PREEMPT_RT patch of the Linux kernel. The storage architecture uses 4TB of local RAM NVMe SSD to cache short-term raw data. Cloud synchronization is used to upload key events to the central server with daily encryption. Extreme environment protection, triple seal for stainless steel shell, inert gas filling, silicone waterproof interface; self-sustaining power supply, the main power supply is the dam body photovoltaic-diesel hybrid power supply, and the backup power supply is a supercapacitor group.

5. The anti-ice pullout system for concrete dam hanging pipes in high-altitude cold regions according to claim 4 is characterized in that: The dynamic control module specifically includes: Phase change material heat blocking layer: The matrix is ​​microencapsulated paraffin, encapsulated in a silica reinforced skeleton to form a flexible film layer, and doped with carbon nanotubes; Active heating unit: Pulsed carbon fiber heating wire embedded in the PCM layer; the trigger mechanism is that when the predicted ice pullout force value exceeds the established threshold, the power is turned on to raise the temperature to a set temperature, inducing the paraffin wax to transform from solid to liquid; the liquid paraffin wax forms a micron-level lubricating film at the interface between the pipe wall and the ice layer, and floats in conjunction with the water level.

6. The anti-ice pullout system for hanging flower pipes of concrete dams in high-altitude cold regions according to claim 5 is characterized in that: The dynamic control module specifically includes: Gradient stiffness anchoring unit: low-stiffness elastic cushion, made of fluororubber foam, absorbs low-frequency vibration energy in the initial stage of ice pullout; high-stiffness damper, a shape memory alloy variable stiffness subunit; stiffness switching mechanism, including SMA in the austenitic state at room temperature, providing basic stiffness; when the Fice prediction value exceeds a predetermined threshold, power is applied to stimulate martensitic phase transformation, causing the elastic modulus to jump; Anti-frost heave integrated, low-temperature phase change protection, SMA driver built-in PTC heater; anti-icing plug seal, moving parts are fully encapsulated with bellows and injected with low-temperature silicone grease.

7. The anti-ice pullout system for concrete dam hanging pipes in high-altitude cold regions according to claim 6 is characterized in that: The dynamic control module specifically includes: A hierarchical response strategy: in the primary response, the PCM heating unit is activated to remove ice adhesion; the elastic cushion layer maintains a low stiffness state to dissipate energy; in the advanced response, the PCM heating power is increased, the SMA damper high stiffness mode is activated to limit structural displacement, and encrypted data sampling is requested from the monitoring module; Fail-safe mechanism, PCM layer failure backup, if the temperature sensor detects heating failure, switches to electric pulse de-icing mode; damper overload protection, when the displacement exceeds the established threshold, triggers the hydraulic buffer emergency braking.

8. The anti-ice pullout system for concrete dam hanging pipes in high-altitude cold regions according to claim 7 is characterized in that: The intelligent monitoring module specifically includes: Distributed fiber optic strain sensor group: The optical fiber is buried in the protective layer of the pipe wall along the axial direction of the hanging pipe, with measuring points arranged at predetermined intervals to fully cover the pipe. It uses low-temperature resistant single-mode optical fiber and is coated with a polyimide coating to resist frost heave and extrusion. Sensor types include FBG arrays and BOTDR systems. By capturing the axial strain gradient of the pipe in real time, the ice pull-out force distribution is monitored; by using displacement resolution, the micro-deformation of the pipe is monitored; and by using the accumulated strain measurement error, the freeze-thaw cycle damage is monitored.

9. The anti-ice pullout system for concrete dam hanging pipes in high-altitude cold regions according to claim 8 is characterized in that: The intelligent monitoring module specifically includes: Temperature-water level coupled sensor array: Ice thickness radar uses frequency modulated continuous wave radar. Multiple radar nodes are deployed in the waters in front of the dam to form a monitoring grid, which scans the entire area regularly. It also outputs ice thickness non-uniformity parameters. Ice temperature gradient monitoring chain, the vertical temperature measurement chain is composed of multiple PT1000 platinum resistance sensors, which calculate the temperature difference between the upper and lower surfaces in real time and generate thermal stress components based on the ice thermal expansion coefficient; the probe shell is made of titanium alloy and the base is counterweighted; High dynamic water level fluctuation meter, the main sensor is a radar water level gauge, the auxiliary pressure compensated water level sensor, the key output is the water level change rate, which is used for buoyancy component modeling.

10. A method for resisting ice pullout of hanging flower pipes of concrete dams in high-altitude cold regions, used to implement the anti-ice pullout system for hanging flower pipes of concrete dams in high-altitude cold regions as described in claims 1-9, characterized in that: include: The distributed optical fiber sensor group embedded in the axial protective layer of the hanging flower tube is used to collect the micro strain under the action of ice pull force in real time. Ice layer parameter collection: ice thickness radars are deployed in the grid in front of the dam to output non-uniformity parameters; chain temperature probes are inserted into the ice layer to measure vertical temperature differences; millimeter-wave radar water level gauges capture the rate of water level change; The non-uniformity parameter is calculated by the standard deviation of ice thickness measured by radar grid: Where δ is the non-uniform thickness of the ice layer, which is calculated from ice thickness data measured by radar grids and reflects the degree of discreteness of the ice thickness; To take the reciprocal operation term, h is the water depth in meters, and the subsequent standard deviation calculation results are scaled so that the ice layer non-uniform thickness index is related to the water depth; is: find the average operation term, n is the total number of samples of ice thickness data measured by radar grid, which is used to calculate the average ice thickness; h i is the ice thickness at the i-th radar grid measurement point, in meters, which is the basic ice thickness measurement data; is the average ice thickness, which comprehensively reflects the average thickness of the ice layer within the radar grid measurement range; n is the total number of samples of ice thickness data measured by the radar grid, that is, the number of points where the radar measures ice thickness within a grid range. It is used to determine the average ice thickness and calculate the total deviation. It is dimensionless. Build an edge computing fusion model, hierarchical anti-ice pullout regulation, primary response, trigger pulse heating of the phase change material layer, maintain the elastic cushion modulus of the anchoring system; advanced response, PCM power increase, forced ice melting and peeling, activate the SMA damper stiffness exponential growth.

Images