Self-adaptive high pier maintenance system and method

Abstract

The invention discloses a self-adaptive high pier maintenance system and method, and relates to the technical field of bridge engineering and building construction. The self-adaptive high pier maintenance system comprises three layers of annular spraying hose units capable of synchronously moving along with a hydraulic creeping formwork; the multi-mode sensing module is integrated with a weather station, a temperature sensor and a sonic sensor; a concrete curing digital twinning body is arranged in the intelligent execution decision-making module; the maintenance method comprises the steps of deploying a spraying pipeline and a sensor network, activating a digital twinborn body to divide maintenance partitions, collecting data in real time, recognizing crack risks, generating a four-dimensional decision matrix to regulate and control nozzle parameters, maintaining the tension of a hose to be 20 + / -5N during climbing, and rapidly transferring after maintenance is completed. According to the system, through multi-modal sensing, intelligent decision making and self-adaptive execution, all-dimensional precise curing of high pier concrete is achieved, the problems of uneven coverage, response lag and the like of a traditional method are effectively solved, and the curing quality and efficiency are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of bridge engineering and building construction technology, and in particular to an adaptive high pier curing system and method. Background Technology

[0002] As a key load-bearing structure in bridges, high-rise buildings, and other engineering projects, the quality of concrete construction of high piers directly affects the safety, durability, and service life of the overall structure. Traditional high pier construction, especially under hydraulic climbing formwork technology, faces severe challenges in concrete curing.

[0003] However, existing high-pier curing technologies have many defects and problems in application: existing spraying pipelines are mostly fixed or simply mobile designs, which cannot closely fit the pier contour, especially when the pier geometry changes, resulting in spraying blind spots or overlapping areas; traditional systems mainly rely on single sensors such as temperature or humidity, lacking multi-dimensional data fusion, and cannot capture environmental weather changes, internal hydration heat state or structural micro-crack risks in real time; at the same time, curing strategies are mostly based on preset rules or human experience, without combining dynamic simulation of concrete physical field, resulting in insufficient optimization of spraying parameters.

[0004] In response to the aforementioned technologies, a solution is proposed. Summary of the Invention

[0005] The purpose of this application is to provide an adaptive high-pitched maintenance system and method to solve the technical problems of rigid and uneven coverage, single monitoring methods, and experience-based decision-making logic in existing sprinkler systems.

[0006] This application provides an adaptive high-sill maintenance system and method, which adopts the following technical solution:

[0007] An adaptive high-sill maintenance system includes an adaptive annular sprinkler pipeline module, a multimodal sensing module, an intelligent execution decision module, and an adaptive decision module.

[0008] Adaptive annular spray pipe module: includes three layers of annular spray hose units that can move synchronously with the hydraulic climbing formwork. Each layer of the annular spray hose unit is fitted to the contour of the pier body through an elastic buckle unit. The annular spray hose unit adopts a rotating atomizing nozzle and the spray angle is dynamically adjustable within a range of 30°-60°.

[0009] Multimodal sensing module: includes a micro weather station unit and a temperature monitoring network unit, and also pre-embeds an acoustic signal acquisition unit along the vertical direction of the pier to acquire multimodal data of the pier;

[0010] Intelligent execution decision module: includes an edge computing control unit, which has a built-in concrete curing digital twin, which integrates a physical field simulation model and generates spraying strategy control signals based on an improved particle swarm optimization algorithm;

[0011] Climbing execution module: includes a magnetic base unit and a tension control unit. The magnetic base unit is disposed on the climbing formwork, and the tension control unit is connected to the spray hose to adjust the layout of the hose according to the movement trajectory of the climbing formwork.

[0012] By adopting the above technical solution, a three-layer ring-shaped spray hose unit that can move synchronously with the hydraulic climbing formwork is used in conjunction with an elastic buckle unit to achieve dynamic fitting between the pipeline and the pier outline, thus solving the problem of blind spots in coverage caused by geometric changes of the pier in traditional fixed spray systems.

[0013] The rotating atomizing nozzle dynamically adjusts the spray angle within the range of 30°-60° to adapt to different curvature changes in different pier sections, significantly improving the uniformity of the curing solution distribution;

[0014] The multimodal sensing module collects environmental temperature, humidity and wind speed data in real time through a micro weather station unit. Combined with the acoustic signal acquisition unit pre-embedded along the vertical direction of the pier, it constructs a concrete internal condition monitoring network to simultaneously acquire temperature field changes and structural acoustic emission signals, providing decision-makers with a multi-dimensional data source that integrates environmental and structural conditions.

[0015] The intelligent execution decision module relies on the concrete curing digital twin built into the edge computing control unit. This twin integrates a thermal-humidity-force coupled physical field simulation model and uses an improved particle swarm optimization algorithm to analyze multimodal sensing data, generating a spraying strategy control signal that takes into account both temperature difference control and humidity balance, thereby achieving precise dynamic optimization of curing parameters.

[0016] The climbing execution module fixes the pipeline system to the climbing formwork through the magnetic base unit, while the tension control unit adjusts the hose layout in real time according to the movement trajectory of the climbing formwork to avoid the pipeline from getting tangled or stress concentration during the climbing process. This design forms a closed-loop control from four dimensions: structural fit, state perception, intelligent decision-making and motion coordination, which can significantly improve the consistency and reliability of the curing quality of high pier concrete.

[0017] The expression for the improved particle swarm optimization algorithm is:

[0018]

[0019] in, This represents the current velocity of the i-th particle in the j-th dimension parameter; This represents the inertia weight, with a value ranging from 0.4 to 0.9. and This represents the learning factor, with a value ranging from 1.5 to 2.0. Represents a random number within the interval [0,1]. This represents the historical best position of the i-th particle in the j-th dimension; This represents the globally optimal position of the group in the j-th dimension; This represents the gradient step size coefficient, with a value ranging from 0.01 to 0.1. The fitness function J represents the current particle position. gradient at; This represents the position vector of the i-th particle; This represents the temperature difference between the measured concrete temperature at the m-th monitoring point and the predicted value from the theoretical hydration heat model. Indicates the current spraying strategy The predicted temperature difference at point m below; This represents the difference between the measured humidity of the nth surface region and the predicted value from the evaporation model. Indicates the current spraying strategy Predicted humidity difference in the nth region; and This represents the weighting coefficient, α+β=1.

[0020] By adopting the above technical solution, the improved particle swarm optimization algorithm is used to dynamically optimize the spraying strategy. Its core principle lies in combining the position update mechanism of the traditional particle swarm optimization algorithm with the gradient descent direction of the fitness function. Specifically, a gradient term is introduced into the velocity update term. This allows the particle to move towards its historical optimal position. and the global optimal position of the group While moving, along the fitness function The negative gradient direction is used for local fine-grained search, thus effectively overcoming the defect of traditional algorithms that are prone to getting trapped in local optima; fitness function The design incorporates both the concrete temperature difference control objective and the humidity balance objective, with the temperature difference term... The requirement to minimize the difference between the measured temperature at each monitoring point and the predicted value of the hydration heat model, and the humidity term. The algorithm minimizes the difference between the measured humidity and the predicted value of the evaporation model for each surface area. The weighting coefficients α and β satisfy the constraint α+β=1, which enables the algorithm to dynamically adjust the priority of temperature and humidity control according to the needs of the maintenance stage.

[0021] The edge computing control unit calculates the position and velocity of particles in real time through iterative calculations and outputs a spraying strategy control signal that minimizes the fitness function value. This signal directly drives the angle, water volume, and timing of the rotating atomizing nozzle, which can simultaneously optimize the uniformity of the internal temperature field and the stability of the surface humidity field of the concrete, significantly improving the accuracy and adaptability of the curing strategy.

[0022] The acoustic signal acquisition unit is a fiber optic sensor array arranged along the axial direction of the pier. Its output signal is connected to the voiceprint recognition component, and the expression of the voiceprint recognition component is as follows:

[0023]

[0024] in, This represents the time-domain acoustic emission signal, derived from the original voltage waveform acquired by the acoustic signal acquisition unit; This represents a signal preprocessing operator that performs standardized operations. The feature space mapping function is used to extract the joint time-frequency domain feature vector of the acoustic signal; The pattern classification function outputs the probability distribution of the properties of acoustic emission events through a machine learning model. R represents the decision function, which determines the risk level of concrete structures based on probability thresholds; R represents the risk level output variable, which triggers the control signal source for targeted maintenance strategies, transforming the voiceprint recognition process into a quantifiable mathematical relationship.

[0025] By adopting the above technical solution, optical fiber sensor arrays deployed along the pier axis are used to collect acoustic emission signals inside the concrete. An acoustic signature recognition component converts the unstructured acoustic signals into quantifiable risk level outputs. First, a signal preprocessing operator performs standardization on the original voltage waveform to eliminate environmental noise interference. Second, a feature space mapping function extracts the joint time-frequency feature vector of the acoustic signal, capturing the unique energy distribution pattern and frequency abrupt change characteristics of crack formation. Then, a pattern classification function performs pattern matching on the feature vector based on a machine learning model, outputting the probability distribution of acoustic emission events of different natures. Finally, a decision function determines the risk level of the concrete structure based on a preset probability threshold, generating a risk level output variable. This risk level serves as a trigger signal for targeted maintenance strategies, enabling the system to dynamically adjust spraying parameters according to the risk of crack initiation.

[0026] The intelligent execution decision module includes a physics simulation model, and the coupled equations of the physics simulation model are expressed as follows:

[0027]

[0028] Where T represents the spatial distribution function of the internal temperature field of concrete; H represents the spatial distribution function of the pore humidity field of concrete; and t represents the time variable of concrete age. This represents the temperature conductivity tensor, which characterizes the ability of water to migrate. This represents the humidity diffusion coefficient tensor, characterizing the ability of moisture to migrate. Indicates the heat source term for hydration. It is a function of the degree of hydration reaction; This represents a function representing the rate of surface moisture evaporation. This indicates the spray water replenishment item. The input water volume for the spray; This represents the stress tensor field inside concrete. Represents the strain tensor field; Operators representing thermo-humid-mechanical coupling constitutive relations; Indicates the density of concrete; The vector representing gravitational acceleration is used to simulate the physical field using the coupled equations.

[0029] By adopting the above technical solution, a set of coupled thermal-humidity-mechanical multiphysics equations is constructed using a physical field simulation model. The core principle lies in accurately describing the multi-field interaction mechanism during the concrete curing process through mathematical equations: the temperature field equation combines the heat transfer capacity represented by the temperature conductivity tensor with the hydration heat source term to dynamically simulate the internal temperature evolution of concrete; the humidity field equation introduces the moisture migration capacity represented by the humidity diffusion tensor and simultaneously integrates the surface moisture evaporation function and the spray water replenishment term to quantify the impact of external intervention on humidity distribution; the stress field equation transforms the temperature gradient and humidity gradient into an internal stress field through the thermal-humidity-mechanical coupling constitutive relation operator, and then verifies the structural stability through the static equilibrium equation; the edge computing control unit inputs multimodal sensing data into the coupled equation set in real time to predict the spatial distribution and temporal evolution trend of the concrete humidity-stress field under different spraying strategies, realizing a full-dimensional dynamic simulation of the internal state of concrete during the curing process, providing a physical mechanism-level calculation basis for spraying parameter optimization, and significantly improving the synergy between temperature control and humidity regulation.

[0030] The intelligent execution decision module performs the following operations:

[0031] A1. Input the real-time monitoring data into the physical field simulation model to predict the moisture-stress distribution of concrete;

[0032] A2. Optimize the spatial allocation and timing logic of spray parameters based on prediction results;

[0033] A3. Output a set of maintenance instructions including location identifiers, atomization characteristics, and water distribution.

[0034] By adopting the above technical solution, the intelligent execution decision module performs dynamic strategy transformation operations based on the physical field simulation results. Its core principle lies in establishing a seamless closed-loop control link from simulation prediction to spraying execution: In step A1, the real-time monitoring data collected by the multimodal perception module is input into the physical field simulation model in real time to generate spatial distribution predictions of the internal humidity field and stress field of the concrete; based on the prediction results, the spatial allocation of spraying parameters is optimized in step A2. The atomized particle size, spraying angle, and water volume ratio required for different pier areas are calculated by an improved particle swarm optimization algorithm, while optimizing the timing logic of each nozzle to avoid resource conflicts; finally, in step A3, a curing instruction set containing three-dimensional attributes such as location identification, atomization characteristics, and water volume allocation is output. This instruction set directly drives the rotating nozzle to perform targeted actions through the edge computing control unit, realizing the leap from global coarse control to local precise intervention in the curing strategy. Through iterative feedback of physical field simulation and real-time data, the spraying system can dynamically respond to the spatiotemporal changes in the state of the concrete structure, significantly improving the adaptability and control accuracy of the curing process.

[0035] The concrete curing digital twin includes a real-time mapping layer, a predictive simulation layer, and a strategy optimization layer. The real-time mapping layer receives monitoring data from the multimodal sensing module through a sensor interface. The predictive simulation layer calls the physical field simulation model to predict the evolution trend of the concrete humidity-stress field based on the output data of the real-time mapping layer. The strategy optimization layer generates spray control commands according to the evolution trend and distributes the spray control commands to the intelligent execution decision module.

[0036] By adopting the above technical solution, the three-layer collaborative architecture of the concrete curing digital twin is used to realize the dynamic generation and closed-loop control of curing strategies, and to construct a decision-making mechanism that integrates data-driven and model-driven approaches: The real-time mapping layer directly receives environmental data, temperature data and acoustic signals collected by the multimodal sensing module through the sensor interface to form a digital mirror of the physical state of the pier; the prediction simulation layer calls the physical field simulation model to perform coupled calculations on the multi-source data output by the real-time mapping layer, accurately simulates the spatiotemporal evolution trend of the concrete humidity field and stress field, and predicts potential risk areas in advance; the strategy optimization layer generates spatially differentiated spray control instructions based on the evolution trend data output by the prediction simulation layer, combined with an improved particle swarm optimization algorithm, including nozzle position identification, atomization characteristics and water distribution parameters; this instruction set is distributed to the intelligent execution decision module through the edge computing control unit to drive the spraying action, and at the same time feeds back the execution effect to the real-time mapping layer to start a new round of decision-making cycle.

[0037] The tension control unit dynamically adjusts the spray hose based on the climbing formwork's climbing height and inclination angle data, combined with the concrete curing digital twin, to calculate the ideal tension value. The expression for calculating the ideal tension value is:

[0038]

[0039] in, This represents the gravitational component per unit length of the spray hose; Represents the pipeline space configuration function. This is the coordinate set of the pipeline control points; Indicates the dynamic drag coefficient during climbing; This represents the motion state function of the climbing formwork. For instantaneous climbing height, The climbing formwork's inclination angle variation rate; Indicates environmental disturbance factors; This represents the time-varying wind load function, where t is the time variable; This represents the system safety margin constant.

[0040] By adopting the above technical solution, the ideal tension value is dynamically calculated using the pre-simulation data output by the tension control unit based on the concrete curing digital twin, establishing a mathematical mapping relationship between the pipeline mechanical state and the climbing formwork motion parameters: the pipeline spatial configuration function in the expression quantifies the three-dimensional layout of the spray hose on the pier surface through the control point coordinate set, and the gravity component term combines the unit length gravity of the hose with the spatial configuration to calculate the static tension benchmark in real time; the climbing formwork motion state function analyzes the impact of the instantaneous climbing height change and the climbing frame inclination rate on the dynamic load of the pipeline, and converts it into additional tension demand through the dynamic resistance coefficient; the time-varying wind load function quantifies the instantaneous tension fluctuation caused by environmental disturbance, and is weighted and compensated by the environmental disturbance factor; the system safety margin constant is preset to a fixed threshold to provide buffer redundancy for sudden external forces; the magnetic base unit rigidly fixes the pipeline system to the climbing formwork to ensure the stability of the force transmission path of tension control.

[0041] A health preservation method for an adaptive high-stilt health preservation system includes the following steps:

[0042] S1. Deploy a ring-shaped spray pipeline and sensor network through the climbing execution module, and use quantum dot fluorescent markers to assist in positioning verification;

[0043] S2. Activate the concrete curing digital twin, load the BIM model of the pier body and divide the dynamic curing zone grid, and assign a unique RFID coordinate code to each nozzle.

[0044] S3. Real-time acquisition of environmental data, concrete temperature data and sound wave signals, and determination of crack risk level through the soundprint recognition component, triggering differentiated maintenance strategies based on the risk level;

[0045] S4. The strategy optimization layer generates a four-dimensional decision matrix and dynamically adjusts the working parameters of each nozzle.

[0046] S5. During the climbing process, based on the tension distribution data pre-simulated by the concrete curing digital twin, the reel mechanism is controlled to maintain the hose tension within the range of 20±5N.

[0047] S6. After maintenance is completed, remove the magnetic base unit and move it to the next construction section.

[0048] By adopting the above technical solution, firstly, in step S1, when deploying the annular spray pipeline through the magnetic base unit, quantum dot fluorescent markers are used simultaneously to perform optical-assisted positioning verification of the pipeline nodes to ensure the spatial coordinate accuracy of the sensor network; in step S2, after activating the concrete curing digital twin, the pier BIM model is loaded, and dynamic curing zoning grids are divided according to the structural curing characteristics, and a unique RFID coordinate code is assigned to each rotating atomizing nozzle to establish a precise mapping relationship between the spray unit and the three-dimensional space; in step S3, the acoustic emission signal inside the concrete collected by the acoustic sensor is processed by the acoustic fingerprint recognition component. When the pattern classification function identifies the time-frequency domain joint feature vector that matches the characteristics of crack generation, the decision function immediately marks the corresponding area as a high-risk level, triggering the automatic improvement of differentiated curing strategies. Step S4 involves adjusting the atomized particle size and water volume ratio in the area; Step S5 involves the strategy optimization layer generating a four-dimensional decision matrix based on physical field simulation prediction results, which includes spatial location, time series, atomization parameters, and water volume distribution, and dynamically controlling the working status of each RFID-coded nozzle; Step S6 involves the tension control unit adjusting the reel mechanism in real time based on the tension distribution data pre-simulated by the digital twin, using a proportional-derivative control algorithm to precisely maintain the hose tension within the target range; Finally, Step S6 involves disassembling the magnetic base unit after curing to achieve rapid system transfer; This design achieves full-process adaptive control of the curing process from deployment to transfer through five core links: optical verification, spatial coding, risk triggering, four-dimensional control, and tension closed loop, significantly improving the accuracy and construction continuity of high-pier concrete curing.

[0049] In step S3, feature extraction and pattern classification are performed on the acoustic signal. When an acoustic emission waveform that matches the characteristics of crack generation is identified, the corresponding concrete area is marked as high-risk. The atomization particle size of the high-risk area is increased to 140-150μm, and the water volume is increased by 18-20%. The spray atomization particle size is dynamically adjusted according to the curvature of the pier surface, ranging from 50-150μm, and the flow rate calibration error of the variable frequency water pump is less than ±3%.

[0050] By adopting the above technical solution, targeted maintenance intervention is performed after the voiceprint recognition component determines the risk level of concrete cracks, and a dynamic mapping mechanism between acoustic emission characteristics and spraying parameters is established: when the pattern classification function identifies a time-frequency domain joint feature vector that matches the characteristics of crack generation, the decision function immediately marks the corresponding concrete area as a high-risk level; the system automatically triggers a differentiated maintenance strategy based on this risk level, specifically by simultaneously increasing the atomization particle size of high-risk areas to the range of 140-150μm and increasing the water volume ratio by 18-20%. The technical principle is that increasing the atomization particle size can enhance the penetration ability of water molecules in the micro-crack area, while increasing the water volume compensates for the evaporation loss caused by changes in surface tension; non-high-risk areas maintain standard spraying parameters to ensure precise focus of maintenance resources.

[0051] In summary, this application includes at least one of the following beneficial technical effects:

[0052] 1. By using a three-layer annular spray hose unit that moves synchronously with the hydraulic climbing formwork, combined with an elastic buckle unit that dynamically fits the contour of the pier, and a rotating atomizing nozzle that dynamically adjusts the spray angle, the problem of uneven coverage in traditional spraying systems is solved.

[0053] 2. Integrate the multimodal perception module and the intelligent execution decision module, perform physical field simulation and improve the particle swarm optimization algorithm based on the concrete curing digital twin to generate spraying strategy, and dynamically optimize the spatial allocation and temporal logic of spraying parameters;

[0054] 3. By using the magnetic base unit and tension control unit of the climbing execution module, combined with the motion trajectory data pre-simulated by the digital twin, the ideal tension value is calculated and the pipeline layout is dynamically adjusted. The hose tension is automatically maintained during the climbing model movement to avoid pipeline chaos.

[0055] 4. The hierarchical structure of the digital twin enables seamless conversion of monitoring data into control commands, forming an adaptive feedback loop and improving the real-time response capability to environmental changes and structural risks. Attached Figure Description

[0056] Figure 1 This is the main control flowchart of the system of the present invention.

[0057] Figure 2 This is a flowchart of the adaptive surface adjustment process of the present invention. Detailed Implementation

[0058] The following is in conjunction with the appendix Figure 1 - Appendix Figure 2 This application will be described in further detail below.

[0059] This application discloses an adaptive high-stilt health maintenance system and method.

[0060] An adaptive high-sill maintenance system includes an adaptive annular sprinkler pipeline module, a multimodal sensing module, an intelligent execution decision module, and an adaptive decision module.

[0061] Adaptive annular spray pipe module: includes three layers of annular spray hose units that can move synchronously with the hydraulic climbing formwork. Each layer of the annular spray hose unit is fitted to the contour of the pier body through an elastic buckle unit. The annular spray hose unit adopts a rotating atomizing nozzle and the spray angle is dynamically adjustable within a range of 30°-60°.

[0062] Multimodal sensing module: includes a micro weather station unit and a temperature monitoring network unit, and also pre-embeds an acoustic signal acquisition unit along the vertical direction of the pier to acquire multimodal data of the pier;

[0063] Intelligent execution decision module: includes an edge computing control unit, which has a built-in concrete curing digital twin, which integrates a physical field simulation model and generates spraying strategy control signals based on an improved particle swarm optimization algorithm;

[0064] Climbing execution module: includes a magnetic base unit and a tension control unit. The magnetic base unit is disposed on the climbing formwork, and the tension control unit is connected to the spray hose to adjust the layout of the hose according to the movement trajectory of the climbing formwork.

[0065] Specifically, the adaptive annular sprinkler pipe module consists of three layers of annular sprinkler hoses that move synchronously with the climbing frame, installed on preset tracks at the top, middle, and bottom layers of the hydraulic climbing formwork system. These hoses maintain an adaptive fit to the pier contour via an elastic locking mechanism. The nozzles are intelligent rotating atomizing nozzles, with the spray angle dynamically adjustable from 30° to 60° via a servo motor, ensuring full coverage of the pier's curved surface at different heights. The bottom annular pipe connects to four vertically extendable water supply hoses with pressure compensation, extending to a constant-pressure water source on the ground; dedicated quick-connect hoses directly connect to the outlet of the intelligent variable frequency water pump on the ground.

[0066] Multimodal sensing module: integrates a micro weather station (temperature, wind speed, light intensity) and a distributed concrete temperature monitoring network (embedded NTC temperature sensor array) to collect multi-dimensional data on the environment and the heat of hydration of concrete in real time;

[0067] Intelligent execution decision module: The ground is equipped with a variable frequency water pump set with flow self-calibration function, which accurately adjusts water pressure and flow according to control signals; the explosion-proof control cabinet has a built-in edge computing controller, which supports 4G / WiFi dual-mode communication and manual / automatic seamless switching. Based on the improved particle swarm optimization algorithm, it integrates the concrete hydration thermodynamics model and the surface moisture evaporation model to establish a three-dimensional curing demand map and dynamically generate the optimal spraying strategy (including spraying area priority, atomized particle size distribution and water volume spatiotemporal allocation).

[0068] Climbing execution module: A servo motor-driven double reel mechanism is installed on the climbing formwork guide rail of the pier. The reel has a built-in high-precision tension sensor to collect the hose tension value in real time. The control box integrates a magnetic base positioning unit and a motion trajectory solver. It receives climbing formwork tilt angle and height increment data sent by the concrete curing digital twin through 4G / WiFi dual-mode communication. Based on the ideal tension value calculation formula, it dynamically calculates the pipeline space configuration compensation amount and outputs magnetic base displacement compensation signal and reel winding and unwinding control command to synchronously adjust the suspension curvature and stress distribution of the three-layer annular spray hose.

[0069] The expression for the improved particle swarm optimization algorithm is:

[0070]

[0071] in, This represents the current velocity of the i-th particle in the j-th dimension parameter; This represents the inertia weight, with a value ranging from 0.4 to 0.9. and This represents the learning factor, with a value ranging from 1.5 to 2.0. Represents a random number within the interval [0,1]. This represents the historical best position of the i-th particle in the j-th dimension; This represents the globally optimal position of the group in the j-th dimension; This represents the gradient step size coefficient, with a value ranging from 0.01 to 0.1. The fitness function J represents the current particle position. gradient at; This represents the position vector of the i-th particle; This represents the temperature difference between the measured concrete temperature at the m-th monitoring point and the predicted value from the theoretical hydration heat model. Indicates the current spraying strategy The predicted temperature difference at point m below; This represents the difference between the measured humidity of the nth surface region and the predicted value from the evaporation model. Indicates the current spraying strategy Predicted humidity difference in the nth region; and This represents the weighting coefficient, α+β=1.

[0072] Specifically, particle swarm location Randomly generated parameters represent different combinations of spraying parameters, such as nozzle angle, timing, and water volume. Baseline parameters from the hydration heat model and evaporation model are loaded, and then sensor data is collected in real time to calculate the fitness function. Through gradient terms The model bias is fed back to the particle motion direction to dynamically adjust the spraying strategy;

[0073] Global optimal solution The final spraying instructions include: the opening and closing sequence of each annular pipe, the nozzle oscillation mode, and the pump set output characteristics.

[0074] The acoustic signal acquisition unit is a fiber optic sensor array arranged along the axial direction of the pier. Its output signal is connected to the voiceprint recognition component, and the expression of the voiceprint recognition component is as follows:

[0075]

[0076] in, This represents the time-domain acoustic emission signal, derived from the original voltage waveform acquired by the acoustic signal acquisition unit; This represents a signal preprocessing operator that performs standardized operations. The feature space mapping function is used to extract the joint time-frequency domain feature vector of the acoustic signal; The pattern classification function outputs the probability distribution of the properties of acoustic emission events through a machine learning model. R represents the decision function, which determines the risk level of concrete structures based on probability thresholds; R represents the risk level output variable, which triggers the control signal source for targeted maintenance strategies, transforming the voiceprint recognition process into a quantifiable mathematical relationship.

[0077] Specifically, a complete technology chain of signal acquisition → feature extraction → intelligent decision-making was constructed to realize the physical conversion of signals, that is, the elastic waves generated by micro-cracks inside concrete are modulated by optical signals from fiber optic sensors to generate electrical signals. Through hybrid features Simultaneously capturing the frequency domain resonance characteristics and time domain burst characteristics of acoustic emission events, using a CNN model. By studying the nonlinear separable boundary between hydration noise and crack sound signature, we can overcome the limitations of misjudgment in the traditional threshold method. Then, the output quantity R is directly related to the adjustment of the nozzle's water volume and atomization parameters, forming a risk-driven targeted maintenance mechanism.

[0078] The intelligent execution decision module includes a physics simulation model, and the coupled equations of the physics simulation model are expressed as follows:

[0079]

[0080] Where T represents the spatial distribution function of the internal temperature field of concrete; H represents the spatial distribution function of the pore humidity field of concrete; and t represents the time variable of concrete age. This represents the temperature conductivity tensor, which characterizes the ability of water to migrate. This represents the humidity diffusion coefficient tensor, characterizing the ability of moisture to migrate. Indicates the heat source term for hydration. It is a function of the degree of hydration reaction; This represents a function representing the rate of surface moisture evaporation. This indicates the spray water replenishment item. The input water volume for the spray; This represents the stress tensor field inside concrete. Represents the strain tensor field; Operators representing thermo-humid-mechanical coupling constitutive relations; Indicates the density of concrete; The vector representing gravitational acceleration is used to simulate the physical field using the coupled equations.

[0081] Specifically, the model achieves accurate predictions through a four-fold coupling mechanism. The dynamic reflection of the driving effect of hydration heat release on the temperature field T, in which By linking age (t) and environmental temperature using maturity theory, Quantifying the impact of ambient temperature and humidity on moisture loss from concrete surfaces. The active control of the humidity field H by the sprinkler system is characterized, and then the constitutive equation is... By introducing the coefficients of thermal expansion and humidity expansion, the thermal and humidity changes are transformed into strain. The stress balance equation, which correlates internal stress σ with external load, outputs the critical stress zone that may lead to cracking, directly driving the spatial allocation of spray parameters.

[0082] high Region → Increase water allocation priority;

[0083] Low H region → Increase atomization frequency;

[0084] High T area → Trigger cooling spray mode.

[0085] The tension control unit dynamically adjusts the spray hose based on the climbing formwork's climbing height and inclination angle data, combined with the concrete curing digital twin, to calculate the ideal tension value. The expression for calculating the ideal tension value is:

[0086]

[0087] in, This represents the gravitational component per unit length of the spray hose; Represents the pipeline space configuration function. This is the coordinate set of the pipeline control points; Indicates the dynamic drag coefficient during climbing; This represents the motion state function of the climbing formwork. For instantaneous climbing height, The climbing formwork's inclination angle variation rate; Indicates environmental disturbance factors; This represents the time-varying wind load function, where t is the time variable; This represents the system safety margin constant.

[0088] Specifically, Overcoming hose sagging caused by gravity, Suppressing dynamic shocks caused by climbing acceleration Compensate for random vibrations caused by wind loads. To prevent overload fracture, when the climbing formwork begins to ascend, the digital twin dynamically calculates the ideal tension value under the current height increment and the rate of change of the inclination angle based on the pre-stored 3D model of the pier and real-time inclination angle data. The edge computing control unit uses this ideal tension value as the target input and compares it with the measured value of the tension sensor in real time. It then uses a proportional-derivative control algorithm to generate reel winding and unwinding instructions. The proportional term quickly responds to static tension deviations, while the derivative term predictively compensates for inertial effects based on the rate of change of height.

[0089] A health preservation method for an adaptive high-stilt health preservation system includes the following steps:

[0090] S1. Deploy a ring-shaped spray pipeline and sensor network through the climbing execution module, and use quantum dot fluorescent markers to assist in positioning verification;

[0091] S2. Activate the concrete curing digital twin, load the BIM model of the pier body and divide the dynamic curing zone grid, and assign a unique RFID coordinate code to each nozzle.

[0092] S3. Real-time acquisition of environmental data, concrete temperature data and sound wave signals, and determination of crack risk level through the soundprint recognition component, triggering differentiated maintenance strategies based on the risk level;

[0093] S4. The strategy optimization layer generates a four-dimensional decision matrix and dynamically adjusts the working parameters of each nozzle.

[0094] S5. During the climbing process, based on the tension distribution data pre-simulated by the concrete curing digital twin, the reel mechanism is controlled to maintain the hose tension within the range of 20±5N.

[0095] S6. After maintenance is completed, remove the magnetic base unit and move it to the next construction section.

[0096] Specifically, in actual implementation, ring pipe supports with displacement compensation function can be installed in the 3rd, 6th and 9th sections of the hydraulic climbing formwork, and the optical coordinates of the pipeline nodes can be verified by quantum dot fluorescent markers. Simultaneously, micro weather station units and vertically embedded acoustic signal acquisition units are deployed to form a monitoring network.

[0097] Activate the concrete curing digital twin and load the pier BIM model. Divide the dynamic curing zone grid according to the structural curing curing. Assign a unique RFID coordinate code to each 316L stainless steel smart rotating nozzle and complete the simulation verification of spray coverage.

[0098] During the real-time acquisition phase, the voiceprint recognition component is called to process the acoustic emission signal. When the machine learning model identifies crack features, a graded early warning mechanism is triggered, automatically marking the corresponding area as a red risk level and increasing the atomization particle size of the RFID-coded nozzle to 140-150μm.

[0099] The strategy optimization layer generates a decision matrix based on the physical field simulation prediction results, which includes four-dimensional attributes such as spatial coordinates, time window, atomization angle, and water ratio. This matrix is ​​dynamically distributed to each nozzle through a 5G edge gateway. During the climbing process, the tension control unit uses a proportional-derivative control algorithm to drive the dual reel mechanism to retract and extend the double-layer steel wire reinforced polyurethane hose in real time, based on the tension distribution pre-simulated by the digital twin, so that the pipeline tension is stably maintained within the range of 20±5N. After maintenance, a special quick-release device is used to disassemble the magnetic base unit, and the entire system is transferred and installed within 4 hours.

[0100] In step S3, feature extraction and pattern classification are performed on the acoustic signal. When an acoustic emission waveform that matches the characteristics of crack generation is identified, the corresponding concrete area is marked as high-risk, and the atomized particle size of the high-risk area is increased to 140-150μm and the water volume is increased by 18-20%.

[0101] The spray atomization particle size is dynamically adjusted according to the curvature of the pier surface, ranging from 50 to 150 μm, and the flow rate calibration error of the variable frequency water pump is less than ±3%.

[0102] Specifically, the stepper motor drives the nozzle rotary cup to reduce the rotation speed by 18-20%, increasing the atomized particle size to 140-150μm. At the same time, the variable frequency water pump increases the pressure of the water supply pipeline in this area by 18-20%. Increasing the droplet size can enhance the water penetration ability in the crack area, and increasing the water volume can compensate for the water loss caused by potential micro-cracks.

[0103] For surface adaptability adjustment, the system dynamically controls the rotational angular velocity of each nozzle stepper motor based on the pre-stored pier curvature data in the BIM model. In specific implementation, when the climbing formwork ascends to the pier area with a curvature radius of less than 5m, the nozzle automatically increases the rotational speed of the rotating cup to reduce the atomized particle size to 50-70μm. The principle is that surfaces with smaller curvature radii require finer droplets to avoid water accumulation, while areas with gentle curvature (radius > 15m) use 100-150μm particle size to improve coverage efficiency. This process is achieved through a built-in curvature-particle size mapping algorithm, and the flow rate calibration is corrected in real time by the edge controller to correct the pump frequency, ensuring that the error is less than ±3%.

[0104] The implementation principle of this application embodiment is as follows: First, the system collects sound wave signals and temperature and humidity data in real time through the multimodal perception module. When the voiceprint recognition component extracts the sound emission waveform that matches the characteristics of concrete cracks, it determines the risk level based on the machine learning classification model and maps the coordinates of the high-risk area to the real-time mapping layer of the digital twin.

[0105] Then, the edge computing control unit triggers targeted regulation based on the risk level: for high-risk areas, the rotation speed of the associated nozzle cup is reduced by 18-20% to increase the atomized particle size to 140-150μm, while the output pressure of the variable frequency water pump is increased by 18-20%. The fluid dynamics principle of increasing the particle size to enhance water permeability is used to compensate for the water loss caused by micro-cracks.

[0106] Secondly, based on the BIM-pre-stored pier curvature data, the spray characteristics are dynamically adjusted through a curvature-particle size mapping algorithm: in areas with a curvature radius of less than 5m, the rotation speed of the rotary cup is automatically increased to reduce the atomized particle size to 50-70μm to avoid water accumulation; in flat areas, a particle size of 100-150μm is used to improve coverage efficiency, and the edge controller calibrates the water pump flow error to ±3% in real time, ultimately achieving adaptive and uniform wetting of the curved surface.

[0107] The embodiments described in this specific implementation are preferred embodiments of this application and are not intended to limit the scope of protection of this application. Identical components are represented by the same reference numerals. Therefore, all equivalent changes made to the structure, shape, and principle of this application should be covered within the scope of protection of this application.

Claims

1. An adaptive high-sill maintenance system, comprising an adaptive annular sprinkler pipeline module, a multimodal sensing module, an intelligent execution decision-making module, and an adaptive decision-making module, characterized in that: Adaptive annular spray pipe module: includes three layers of annular spray hose units that can move synchronously with the hydraulic climbing formwork. Each layer of the annular spray hose unit is fitted to the contour of the pier body through an elastic buckle unit. The annular spray hose unit adopts a rotating atomizing nozzle and the spray angle is dynamically adjustable within a range of 30°-60°. Multimodal sensing module: includes a micro weather station unit and a temperature monitoring network unit, and also pre-embeds an acoustic signal acquisition unit along the vertical direction of the pier to acquire multimodal data of the pier; Intelligent execution decision module: includes an edge computing control unit, which has a built-in concrete curing digital twin, which integrates a physical field simulation model and generates spraying strategy control signals based on an improved particle swarm optimization algorithm; Climbing execution module: includes a magnetic base unit and a tension control unit. The magnetic base unit is disposed on the climbing formwork, and the tension control unit is connected to the spray hose to adjust the layout of the hose according to the movement trajectory of the climbing formwork.

2. The adaptive high-stilt health maintenance system according to claim 1, characterized in that: The expression for the improved particle swarm optimization algorithm is: in, This represents the current velocity of the i-th particle in the j-th dimension parameter; This represents the inertia weight, with a value ranging from 0.4 to 0.

9. and This represents the learning factor, with a value ranging from 1.5 to 2.

0. Represents a random number within the interval [0,1]. This represents the historical best position of the i-th particle in the j-th dimension; This represents the globally optimal position of the group in the j-th dimension; This represents the gradient step size coefficient, with a value ranging from 0.01 to 0.

1. The fitness function J represents the current particle position. gradient at; This represents the position vector of the i-th particle; This represents the temperature difference between the measured concrete temperature at the m-th monitoring point and the predicted value from the theoretical hydration heat model. Indicates the current spraying strategy The predicted temperature difference at point m below; This represents the difference between the measured humidity of the nth surface region and the predicted value from the evaporation model. Indicates the current spraying strategy Predicted humidity difference in the nth region; and This represents the weighting coefficient, α+β=1.

3. The adaptive high-stilt health maintenance system according to claim 1, characterized in that: The acoustic signal acquisition unit is a fiber optic sensor array arranged along the axial direction of the pier. Its output signal is connected to the voiceprint recognition component, and the expression of the voiceprint recognition component is as follows: in, This represents the time-domain acoustic emission signal, derived from the original voltage waveform acquired by the acoustic signal acquisition unit; This represents a signal preprocessing operator that performs standardized operations. The feature space mapping function is used to extract the joint time-frequency domain feature vector of the acoustic signal; The pattern classification function outputs the probability distribution of the properties of acoustic emission events through a machine learning model. R represents the decision function, which determines the risk level of concrete structures based on probability thresholds; R represents the risk level output variable, which triggers the control signal source for targeted maintenance strategies, transforming the voiceprint recognition process into a quantifiable mathematical relationship.

4. The adaptive high-stilt health maintenance system according to claim 1, characterized in that: The intelligent execution decision module includes a physics simulation model, and the coupled equations of the physics simulation model are expressed as follows: Where T represents the spatial distribution function of the internal temperature field of concrete; H represents the spatial distribution function of the pore humidity field of concrete; and t represents the time variable of concrete age. This represents the temperature conductivity tensor, which characterizes the ability of water to migrate. This represents the humidity diffusion coefficient tensor, characterizing the ability of moisture to migrate. Indicates the heat source term for hydration. It is a function of the degree of hydration reaction; This represents a function representing the rate of surface moisture evaporation. This indicates the spray water replenishment item. The input water volume for the spray; This represents the stress tensor field inside concrete. Represents the strain tensor field; Operators representing thermo-humid-mechanical coupling constitutive relations; Indicates the density of concrete; The vector representing gravitational acceleration is used to simulate the physical field using the coupled equations.

5. The adaptive high-stilt health maintenance system according to claim 4, characterized in that: The intelligent execution decision module performs the following operations: A1. Input the real-time monitoring data into the physical field simulation model to predict the moisture-stress distribution of concrete; A2. Optimize the spatial allocation and timing logic of spray parameters based on prediction results; A3. Output a set of maintenance instructions including location identifiers, atomization characteristics, and water distribution.

6. The adaptive high-stilt health maintenance system according to claim 1, characterized in that: The concrete curing digital twin includes a real-time mapping layer, a predictive simulation layer, and a strategy optimization layer. The real-time mapping layer receives monitoring data from the multimodal sensing module through a sensor interface. The predictive simulation layer calls the physical field simulation model to predict the evolution trend of the concrete humidity-stress field based on the output data of the real-time mapping layer. The strategy optimization layer generates spray control commands according to the evolution trend and distributes the spray control commands to the intelligent execution decision module.

7. The adaptive high-stilt health maintenance system according to claim 1, characterized in that: The tension control unit dynamically adjusts the spray hose based on the climbing formwork's climbing height and inclination angle data, combined with the concrete curing digital twin, to calculate the ideal tension value. The expression for calculating the ideal tension value is: in, This represents the gravitational component per unit length of the spray hose; Represents the pipeline space configuration function. This is the coordinate set of the pipeline control points; Indicates the dynamic drag coefficient during climbing; This represents the motion state function of the climbing formwork. For instantaneous climbing height, The climbing formwork's inclination angle variation rate; Indicates environmental disturbance factors; This represents the time-varying wind load function, where t is the time variable; This represents the system safety margin constant.

8. A health preservation method for an adaptive high-pitched health preservation system, applicable to the adaptive high-pitched health preservation system according to any one of claims 1-7, characterized in that, Includes the following steps: S1. Deploy a ring-shaped spray pipeline and sensor network through the climbing execution module, and use quantum dot fluorescent markers to assist in positioning verification; S2. Activate the concrete curing digital twin, load the BIM model of the pier body and divide the dynamic curing zone grid, and assign a unique RFID coordinate code to each nozzle. S3. Real-time acquisition of environmental data, concrete temperature data and sound wave signals, and determination of crack risk level through the soundprint recognition component, triggering differentiated maintenance strategies based on the risk level; S4. The strategy optimization layer generates a four-dimensional decision matrix and dynamically adjusts the working parameters of each nozzle. S5. During the climbing process, based on the tension distribution data pre-simulated by the concrete curing digital twin, the reel mechanism is controlled to maintain the hose tension within the range of 20±5N. S6. After maintenance is completed, remove the magnetic base unit and move it to the next construction section.

9. The health preservation method of the adaptive high-pitched health preservation system according to claim 8, characterized in that: In step S3, feature extraction and pattern classification are performed on the acoustic signal. When an acoustic emission waveform that matches the characteristics of crack generation is identified, the corresponding concrete area is marked as high-risk, and the atomized particle size of the high-risk area is increased to 140-150μm and the water volume is increased by 18-20%.

10. The health preservation method of the adaptive high-pitched health preservation system according to claim 8, characterized in that: The spray atomization particle size is dynamically adjusted according to the curvature of the pier surface, ranging from 50 to 150 μm, and the flow rate calibration error of the variable frequency water pump is less than ±3%.

Images