A concrete strength remote monitoring method and system suitable for complex environment

Abstract

The application discloses a kind of concrete strength remote monitoring method and system suitable for complex environment, belong to civil engineering structure health monitoring technical field.The remote monitoring method includes the following steps: step 1, the performance data of concrete structure and the environmental factor data related thereto are acquired, and the multi-source heterogeneous data is transmitted to data processing center in real time by the preset wireless communication protocol;Step 2, construct five-dimensional tensor data structure, and realize the accurate mathematical expression of the complex coupling relationship between environmental factors and material properties by tensor decomposition;Step 3, capture the nonlinear time-varying characteristics of concrete strength evolution;Step 4, quantify the age effect by strength development rate index;Step 5, based on the trend of strength development rate change, adaptively adjust the data sampling frequency and monitor the environmental condition fluctuation;Step 6, real-time evaluation of the safety state of concrete structure, ensure the safe and reliable operation of concrete structure under complex environment.

Description

Technical Field

[0001] This application relates to the field of structural health monitoring technology in civil engineering, and more specifically, to a method and system for remote monitoring of concrete strength suitable for complex environments. Background Technology

[0002] As the most commonly used structural material in construction engineering, the strength development process of concrete directly affects the safety and durability of the structure. In various large-scale infrastructure constructions, bridge constructions, high-rise buildings, and underground projects, real-time monitoring of concrete strength not only helps ensure construction quality but also optimizes construction progress and maintenance strategies, improving project management efficiency. Traditional concrete strength testing mainly relies on specimen preparation and destructive testing methods, such as the standard cube compressive strength test. This method has drawbacks such as long cycle time, excessive manual intervention, data lag, and inability to provide real-time feedback on the concrete strength development in the actual structure.

[0003] In recent years, with the development of intelligent sensing technology, wireless communication technology, and the Internet of Things, researchers have attempted to combine embedded sensors with wireless transmission methods to achieve real-time monitoring of physical parameters such as internal temperature and strain of concrete, and to predict its strength evolution process through empirical formulas or data models. However, existing systems generally suffer from problems such as poor sensor stability, unreliable wireless communication, low data fusion accuracy, high power consumption, and poor structural compatibility under complex environmental conditions (such as extreme climates, remote mountainous areas, underground tunnels, or areas with strong interference), making it difficult to operate continuously for long periods and limiting their widespread application in practical engineering.

[0004] Furthermore, most existing technologies focus on local monitoring or experimental environment verification, lacking systematic design and methodological support for complex field environments, and thus cannot meet the needs of high-intensity, long-term, large-scale, and multi-point distributed monitoring. At the same time, effectively fusing the collected multi-source data and constructing a reliable strength prediction model based on the properties of concrete materials remains a significant challenge.

[0005] In summary, how to construct a method and system for remote monitoring of concrete strength suitable for complex environments, and achieve real-time, stable, low-power, and high-precision monitoring of the concrete strength evolution process, has become an urgent technical problem to be solved. Summary of the Invention

[0006] In order to overcome a series of defects in the existing technology, the purpose of this application is to provide a method for remote monitoring of concrete strength in complex environments, which includes the following steps.

[0007] Step 1: Obtain performance data of the concrete structure and related environmental factors, and transmit the multi-source heterogeneous data to the data processing center in real time through a preset wireless communication protocol.

[0008] Step 2: Construct a five-dimensional tensor data structure and realize the precise mathematical expression of the complex coupling relationship between environmental factors and material properties through tensor decomposition.

[0009] Step 3: Introduce a time-dimensional recursive update mechanism to capture the nonlinear time-varying characteristics of concrete strength evolution.

[0010] Step 4: Quantify the age effect using the intensity development rate index.

[0011] Step 5: Based on the trend of intensity development rate change, adaptively adjust the data sampling frequency and monitor environmental condition fluctuations.

[0012] Step 6: Real-time assessment of the safety status of the concrete structure to ensure its safe and reliable operation in complex environments.

[0013] Furthermore, step 1 includes the following steps.

[0014] Based on finite element analysis, the distribution of key monitoring points in the structure is determined. Combined with load transfer paths and stress concentration areas, the monitoring layout scheme is optimized, and a multi-scale monitoring network topology is constructed.

[0015] A micro-level ion concentration probe array is pre-embedded inside the concrete. Ion selective electrode technology is used to detect changes in chloride ions, sulfate ions, and alkalinity in the pore fluid in real time. At the same time, pH sensors are deployed to monitor the evolution of the alkaline environment of the concrete.

[0016] A mesoscopic acoustic sensor network is deployed at key cross-sections of the structure to assess the internal density, porosity, and microcrack development of concrete by analyzing the acoustic propagation time and waveform characteristics.

[0017] By combining fiber optic grating technology with resistance strain measurement, a three-dimensional strain monitoring network is constructed to achieve real-time monitoring of concrete deformation, cracking, and displacement changes, and to establish a dynamic map of the strain field distribution.

[0018] Environmental parameter acquisition units, including high-precision temperature and humidity sensors, barometers, rain gauges, and ultraviolet intensity detectors, are deployed on the exterior of the structure to construct a microclimate monitoring network; at the same time, groundwater level and soil parameters are collected to establish a complete database of environmental impact factors.

[0019] A data transmission network is constructed using low-power wide-area IoT technology, and edge computing technology is combined to preprocess and compress the raw data. At the same time, the data is efficiently transmitted to the data processing center through a multi-hop self-organizing network protocol.

[0020] Furthermore, the micro-scale ion concentration probe array includes: a chloride ion sensor with a detection range of 0.01%-5% and an accuracy of ±0.005%; a sulfate ion sensor with a detection range of 0.02%-8% and an accuracy of ±0.01%; and a pH sensor with a detection range of 5-14 and an accuracy of ±0.2%.

[0021] The acoustic sensor network consists of 16-64 ultrasonic sensors with a working frequency of 20-100kHz and a measurement accuracy of ±1μs. It can detect microcracks with a width greater than 0.1mm. The sensor spacing is 1-3m, forming a multi-path acoustic wave propagation network.

[0022] The environmental parameter acquisition unit specifically includes: a temperature sensor with a measurement range of -30℃ to +80℃ and an accuracy of ±0.5℃; a humidity sensor with a measurement range of 5-95%RH and an accuracy of ±3%RH; a barometer with a measurement range of 300-1100hPa and an accuracy of ±1hPa; a rain gauge with a measurement range of 0-200mm / h and an accuracy of ±0.5mm; an ultraviolet intensity detector with a measurement range of 0-1200μW / cm² and an accuracy of ±10μW / cm²; a wind speed sensor with a measurement range of 0-50m / s and an accuracy of ±0.5m / s; and a sulfur dioxide concentration sensor with a measurement range of 0-10ppm and an accuracy of ±0.05ppm.

[0023] Furthermore, step 2 includes the following steps.

[0024] Principal component analysis was used to reduce the dimensionality of environmental data, extract the main variables, and wavelet transform was used for time-frequency analysis to construct the feature vector of environmental state.

[0025] By combining concrete mix proportion parameters, curing conditions, and measured strength grades, a multi-scale material model is used to quantitatively describe the microstructural characteristics of concrete, including pore distribution, hydration product ratio, and interfacial transition zone characteristics. By constructing a comprehensive material state descriptor, the precise digital expression of concrete material properties is achieved.

[0026] An exponentially weighted decay factor is introduced to process historical data and a recursive time series structure is constructed to capture the evolution law of concrete strength. At the same time, the strength development curves at different stages are simulated to achieve a precise mathematical expression of the time-varying characteristics of concrete strength.

[0027] By establishing a mapping relationship between microscopic physicochemical processes and macroscopic mechanical properties through a scale transformation function, the differences in monitoring data at different structural scales can be eliminated, enabling reliable inference from local monitoring points to overall structural performance.

[0028] By introducing performance degradation rate and limit state function, a comprehensive evaluation system based on performance indicators is constructed, and the weights of each indicator are set using reliability theory to form a multi-objective performance evaluation criterion.

[0029] Construct a five-dimensional tensor structure T(E,M,τ,S,P), where each dimension corresponds to environmental factors, material properties, time evolution, structural scale, and performance indicators, respectively.

[0030] T(E,M,τ,S,P) is decomposed into the product of the core tensor and five factor matrices. The main feature patterns of each dimension are extracted, and the coupling strength coefficient between each dimension is calculated to quantitatively characterize the influence of environmental factors and material properties on concrete performance at different times and scales.

[0031] Furthermore, step 3 includes the following steps.

[0032] A piecewise nonlinear model for concrete strength development is established, dividing the entire age period into a rapid growth period, a transition period, and a stable period. Each stage is connected by a piecewise continuous function to achieve a precise mathematical expression of the strength evolution throughout the entire age period.

[0033] A recursive prediction model based on a long short-term memory network is constructed. Historical monitoring data and five-dimensional tensor feature expression are used as inputs to learn the intrinsic law of concrete strength development and realize intelligent prediction and real-time correction of the strength development trajectory.

[0034] By integrating and accumulating environmental impact factors, the actual calendar age is converted into an equivalent age that accurately reflects the true degree of hydration, thereby achieving a unified expression of intensity development curves under different environmental conditions.

[0035] An exponentially weighted moving average method is used to enhance the importance of recent data, and a forgetting factor is set to gradually reduce the influence of long-term data. At the same time, the model parameters are forcibly calibrated at key age nodes to ensure that the prediction results are highly consistent with the actual intensity development.

[0036] By calculating the deviation between the measured intensity and the theoretical curve, multi-level early warning thresholds are set. When the deviation exceeds the allowable range, the anomaly analysis process is triggered, and Bayesian inference methods are introduced to identify influencing factors in order to distinguish between normal intensity fluctuations and potential quality problems.

[0037] Real-time monitoring data is aggregated and analyzed at different time scales, and wavelet transform is used to extract feature patterns at different time scales. At the same time, a cross-scale time correlation network is established to realize full-spectrum monitoring and prediction from short-term intensity fluctuations to long-term performance evolution.

[0038] Furthermore, step 4 includes the following steps.

[0039] By calculating the environment-material sensitivity matrix, the sensitivity of different types of concrete to environmental factors is quantified.

[0040] A piecewise temperature compensation function is established, dividing the temperature range into low-temperature, normal-temperature, and high-temperature zones. Different mathematical models are used in each zone to describe the influence of temperature on the hydration reaction rate, thereby achieving accurate prediction of concrete strength development across the entire temperature range.

[0041] By combining the size effect of concrete components, a humidity gradient-intensity distribution mapping relationship is constructed, thereby enabling reliable inference from surface humidity monitoring to internal intensity distribution.

[0042] Environmental conditions are quantitatively characterized by the comprehensive environmental impact index and used as a standardized input parameter.

[0043] The dynamic relationship between the strength development rate index and the environmental comprehensive effect index is described by differential equations, and the SDR-ECI response surface is constructed to quantify the acceleration or deceleration effect of strength development under different environmental conditions. At the same time, a material sensitivity factor is introduced to adjust the shape of the response surface to adapt to the characteristics of different concrete mix proportions, so as to achieve accurate prediction of the strength development law in changing environments.

[0044] Establish the correlation between equivalent age and actual strength development, calculate the strength contribution rate at different time periods, and establish a strength history evolution archive to record the long-term impact of key environmental events on concrete strength development.

[0045] Furthermore, step 5 includes the following steps.

[0046] The first and second derivatives of the intensity development rate are calculated in real time. The first derivative reflects the direction of the rate change, and the second derivative represents the magnitude of the acceleration. When an inflection point or increased fluctuation in the intensity development rate is detected, the optimal allocation of monitoring resources is achieved by intelligently adjusting the sampling density.

[0047] By integrating parameters such as temperature change rate, humidity gradient, carbon dioxide concentration fluctuation and chloride ion permeation rate, and using a multivariate weighting function to calculate a comprehensive sensitivity score, the system achieves accurate quantification and hierarchical monitoring of changes in environmental conditions.

[0048] Establish normal fluctuation ranges for environmental parameters, take into account seasonal variation characteristics and geographical location factors, dynamically update threshold settings, and identify environmental change patterns to distinguish between normal seasonal fluctuations and abnormal climate events.

[0049] A multi-level triggering mechanism is constructed, which divides environmental parameter fluctuations into three levels: attention level, warning level, and emergency level. At the same time, multiple parameter combination triggering conditions are considered to achieve intelligent response to complex environmental changes.

[0050] Lightweight computing modules are deployed at on-site monitoring nodes to perform real-time data preprocessing and preliminary analysis. When abnormal environmental conditions are detected, the local sampling strategy is adjusted immediately, and a data caching mechanism is activated to ensure that critical data can still be saved in the event of communication interruption. A clear task division and data synchronization mechanism is established with the central data processing center.

[0051] When a specific environmental event is detected, the system automatically enters a high-frequency sampling mode, activates additional environmental monitoring units, comprehensively records data throughout the entire process of the environmental event, performs a special assessment of concrete strength after the event ends, and constructs a causal relationship database between environmental event and strength change.

[0052] Furthermore, step 6 includes the following steps.

[0053] A multi-source heterogeneous data fusion framework was constructed to extract abnormal feature patterns from various types of sensor data and to establish a concrete strength evolution feature library.

[0054] It enables rolling forecasts of intensity values ​​for the next 7 days, 30 days, and 90 days.

[0055] Based on the structural importance level and service environment characteristics, a regional and graded strength safety threshold system is established. At the same time, the early warning threshold parameters are automatically calibrated in combination with the characteristics of seasonal environmental changes, and the safety margin is adjusted in a timely manner according to the structural aging rate.

[0056] Different levels of early warning response are triggered based on the degree of intensity decline: Level 1 warning is triggered when the intensity is lower than 90% of the design value, reminding people to pay attention to potential risks; Level 2 intervention warning is triggered when the intensity is lower than 85%, recommending preventive measures; Level 3 emergency warning is triggered when the intensity is lower than 80% or the rate of intensity decline exceeds 0.5 MPa / day, requiring immediate intervention measures to prevent further deterioration.

[0057] Based on the finite element mesh generation principle, a three-dimensional distribution cloud map of structural strength is generated, and a structural health assessment report containing strength numerical analysis, environmental impact assessment, safety risk classification, and maintenance recommendations is automatically generated.

[0058] By combining strength development trends and environmental sensitivity analysis, maintenance decision recommendations are automatically generated, including adjusting the inspection frequency, strengthening protective measures, and necessary reinforcement schemes, to ensure the long-term stability and safety of concrete structures in complex environments.

[0059] Furthermore, the method for generating the three-dimensional distribution cloud map is as follows.

[0060] Based on tetrahedral mesh generation technology, 5000-20000 cells are adaptively generated according to the structural complexity.

[0061] Each grid cell is associated with data from 2-3 nearby sensor nodes.

[0062] The intensity value within the mesh element is calculated using linear interpolation.

[0063] The Laplace smoothing algorithm is used to consider spatial correlation and smooth the transition between adjacent cells.

[0064] A three-color coloring scheme is adopted: red indicates dangerous areas with an intensity lower than 80% of the design value, yellow indicates warning areas with an intensity of 80%-90% of the design value, and green indicates safe areas with an intensity higher than 90% of the design value.

[0065] A simplified 3D rendering algorithm is used to achieve the desired visualization effect, ensuring smooth display on ordinary mobile devices.

[0066] The purpose of this application is also to provide a remote monitoring system for concrete strength suitable for complex environments, including the following modules.

[0067] The multi-level intelligent sensing and monitoring module enables comprehensive monitoring of the internal chemical environment, acoustic properties, and strain distribution of concrete through a sensor network at three levels: microscopic, mesoscopic, and macroscopic.

[0068] The environmental parameter acquisition and communication module is responsible for acquiring environmental parameters around the concrete structure and transmitting the data to the processing center via low-power IoT technology.

[0069] The data processing module is used to realize a unified mathematical expression of the coupling relationship between environmental factors, material properties, time evolution, structural scale and performance indicators.

[0070] The time-series strength evolution prediction module enables full-spectrum prediction of concrete strength from short-term fluctuations to long-term evolution under different environmental conditions, accurately capturing nonlinear time-varying characteristics.

[0071] The strength development rate assessment module is used to quantify the impact of different environmental conditions on the strength development rate of concrete, and to achieve accurate assessment of the age effect.

[0072] The adaptive monitoring and control module, based on intensity development rate derivative analysis and environmental sensitivity scoring, intelligently adjusts the sampling frequency and tracks environmental events, achieving optimal allocation of monitoring resources and accurate capture of key environmental changes.

[0073] The structural safety status assessment module integrates multi-source heterogeneous data to perform rolling intensity prediction, establishes a hierarchical safety threshold system and a multi-level early warning response mechanism, and generates a three-dimensional intensity distribution cloud map and a health assessment report.

[0074] Compared with the prior art, this application has the following beneficial effects.

[0075] This application enables high-precision, real-time, and full-lifecycle remote monitoring of concrete strength in complex environments. Through multi-source heterogeneous data fusion, tensor modeling, and intelligent prediction mechanisms, it improves the scientific rigor and reliability of structural safety assessments and effectively optimizes monitoring resource allocation and response strategies. Attached Figure Description

[0076] Figure 1 This is a schematic flowchart of a method for remote monitoring of concrete strength in complex environments, as disclosed in an embodiment of this application.

[0077] Figure 2 This is a structural schematic diagram of a remote concrete strength monitoring system suitable for complex environments, as disclosed in an embodiment of this application. Detailed Implementation

[0078] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the embodiments of this invention will be described in more detail below with reference to the accompanying drawings. In the drawings, the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The described embodiments are some embodiments of this invention, but not all embodiments.

[0079] Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0080] The embodiments and directional terms described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0081] like Figure 1 As shown, a method for remote monitoring of concrete strength suitable for complex environments includes the following steps.

[0082] Step 1: Obtain performance data of the concrete structure and related environmental factors, and transmit the multi-source heterogeneous data to the data processing center in real time through a preset wireless communication protocol.

[0083] Step 2: Construct a five-dimensional tensor data structure and realize the precise mathematical expression of the complex coupling relationship between environmental factors and material properties through tensor decomposition.

[0084] Step 3: Introduce a time-dimensional recursive update mechanism to capture the nonlinear time-varying characteristics of concrete strength evolution.

[0085] Step 4: Quantify the age effect using the intensity development rate index.

[0086] Step 5: Based on the trend of intensity development rate change, adaptively adjust the data sampling frequency and monitor environmental condition fluctuations.

[0087] Step 6: Real-time assessment of the safety status of the concrete structure to ensure its safe and reliable operation in complex environments.

[0088] As can be seen from the above, this remote monitoring method for concrete strength fully integrates multiple technologies, including multi-source data acquisition, tensor data structure modeling, time-dimensional recursive updating, strength development rate quantification, adaptive sampling adjustment, and real-time safety assessment, forming a highly efficient monitoring system. Each step complements the others, ensuring both the real-time and complete data transmission and achieving accurate description and prediction of concrete performance evolution under complex environments. The application of this method not only promptly captures and reflects the internal and external influences on concrete structures during service but also allows for rapid response to abnormal trends, ensuring the safe and stable operation of the structure. Through this integrated and intelligent technological approach, key indicators can be extracted from massive amounts of data, enabling the development of targeted maintenance and reinforcement measures, thereby significantly reducing safety risks and accident rates.

[0089] Furthermore, step 1 includes the following steps.

[0090] Based on finite element analysis, the distribution of key monitoring points in the structure is determined. Combined with load transfer paths and stress concentration areas, the monitoring layout scheme is optimized, and a multi-scale monitoring network topology is constructed.

[0091] A micro-level ion concentration probe array is pre-embedded inside the concrete. Ion selective electrode technology is used to detect changes in chloride ions, sulfate ions, and alkalinity in the pore fluid in real time. At the same time, pH sensors are deployed to monitor the evolution of the alkaline environment of the concrete.

[0092] A mesoscopic acoustic sensor network is deployed at key cross-sections of the structure to assess the internal density, porosity, and microcrack development of concrete by analyzing the acoustic propagation time and waveform characteristics.

[0093] By combining fiber optic grating technology with resistance strain measurement, a three-dimensional strain monitoring network is constructed to achieve real-time monitoring of concrete deformation, cracking, and displacement changes, and to establish a dynamic map of the strain field distribution.

[0094] Environmental parameter acquisition units, including high-precision temperature and humidity sensors, barometers, rain gauges, and ultraviolet intensity detectors, are deployed on the exterior of the structure to construct a microclimate monitoring network; at the same time, groundwater level and soil parameters are collected to establish a complete database of environmental impact factors.

[0095] A data transmission network is constructed using low-power wide-area IoT technology, and edge computing technology is combined to preprocess and compress the raw data. At the same time, the data is efficiently transmitted to the data processing center through a multi-hop self-organizing network protocol.

[0096] The distribution of key monitoring points in the structure is determined based on finite element analysis through the following steps: First, a three-dimensional solid model of the concrete structure is established using ANSYS Mechanical APDL or ABAQUS / Standard finite element analysis software. Then, C3D8R eight-node hexahedral elements are used for meshing, with element sizes controlled within 1 / 10 to 1 / 20 of the structural characteristic dimensions. Next, the stress distribution of the structure under design loads is calculated by solving the linear statics equations. Stress concentration areas are identified using the Von Mises equivalent stress criterion, and key monitoring sections are determined using the maximum principal stress theory. Finally, considering load transfer paths and fatigue-prone areas, a multi-objective optimization algorithm is used to determine the optimal layout of the monitoring points.

[0097] The construction of the graph theory-based monitoring network topology is achieved through the following steps: each monitoring node is regarded as a vertex of the graph, and the communication links between monitoring nodes are regarded as edges. The minimum spanning tree algorithm is used to ensure network connectivity while minimizing communication costs. A hierarchical star-mesh hybrid topology is adopted, in which micro-level sensors form star sub-networks, and the convergence nodes of each sub-network are interconnected through a mesh structure to achieve multi-hop data transmission and redundancy backup. The network topology meets the reliability requirement of connectivity ≥2, and the shortest path is calculated using Dijkstra's algorithm to optimize data transmission efficiency.

[0098] The project combines fiber Bragg grating technology with resistance strain measurement to construct a three-dimensional strain monitoring network. The specific steps are as follows: Distributed fiber Bragg grating sensors are used, with a working wavelength range of 1525-1565nm, a strain measurement range of ±5000με, a temperature measurement range of -40℃ to +80℃, and a spatial resolution of up to 1mm. The fiber optic sensors are arranged along three orthogonal axes and fixed to the steel reinforcement surface via prefabricated spiral channels or special clamps, forming a strain measurement matrix in a three-dimensional coordinate system. Simultaneously, 120Ω precision foil resistance strain gauges are deployed as a supplementary measurement method, using quarter-bridge, half-bridge, and full-bridge configurations. The measurement circuit converts strain into a voltage signal using the Wheatstone bridge principle; a calibration equation relating FBG wavelength drift to strain is established, and the coupling effect between strain and temperature is separated using a temperature compensation algorithm; 16-32 FBG sensors are integrated on a single optical fiber using wavelength division multiplexing (WDM) technology, and signal demodulation is performed using an optical time domain reflectometer (OTDR) and an optical frequency domain reflectometer (OFDR) to achieve distributed long-distance strain monitoring; a three-dimensional strain tensor matrix is ​​constructed, and the strain field distribution between monitoring points is calculated using a finite element interpolation algorithm, and the crack propagation direction is identified by combining principal strain theory to establish a dynamic map of the strain field distribution.

[0099] The data transmission network is configured as follows: It uses an STM32L4 series ultra-low-power microcontroller as the main control chip, integrating a 16-bit ADC converter. A multi-channel analog switch CD4051 is configured to achieve time-division multiplexing acquisition of multi-sensor signals, with an operating current of only 35μA, supporting parallel acquisition of up to 64 sensor signals. A 32GB industrial-grade eMMC flash memory chip (such as the Samsung KLM8G1GETF-B041) is configured as a local data cache, supporting power-off protection and data integrity verification. A real-time clock chip PCF8563 is integrated to provide accurate timestamps, and the storage capacity can support 30 days of continuous data recording offline. An ARM Cortex-A53 quad-core processor (such as the Raspberry Pi Compute Module 4) with a main frequency of 1.5GHz and 4GB of LPDDR4 memory is used, running an embedded Linux system and integrating a TensorFlow Lite inference engine. This enables local execution of data preprocessing, anomaly detection, and compression algorithms, with a processing capacity of up to 100,000 floating-point operations per second. The main communication method uses a Semtech SX1276. The LoRa chip operates in the 470-510MHz frequency band, with a transmit power of 20dBm, a transmission distance of 10-15km, and a data rate of 0.3-50kbps. It is equipped with a high-gain directional antenna (12dBi gain). Backup communication utilizes the Quectel BC95-GNB-IoT module, supporting Cat-NB1 / Cat-NB2 standards, with a standby power consumption of only 3μA, ensuring reliable data transmission even when the LoRaWAN signal is unstable. A high-efficiency DC-DC converter (such as the TI TPS63070) with a conversion efficiency of 95% is employed. A 20Ah 18650 lithium battery pack and a 30W solar panel are included, supporting MPPT (Maximum Power Point Tracking) to ensure continuous operation for over 6 months without external power. The system utilizes the multi-hop ad hoc on-demand network protocol AODV. DistanceVector enables automatic routing discovery and maintenance between devices, while integrating a TCP / IP protocol stack to support reliable connections with cloud servers. It uses the LZ77 lossless compression algorithm to reduce the amount of raw data by 60-80%, integrates digital filters (such as Kalman filters) to eliminate sensor noise, and uses edge AI algorithms to detect data anomalies and extract features. Only key information and abnormal data are uploaded to the data processing center, significantly reducing network bandwidth requirements.

[0100] The aforementioned series of steps, combining multi-dimensional sensor networks with advanced transmission and computing technologies, enables precise monitoring and analysis of concrete structures in complex environments. By optimizing the layout of monitoring points through finite element analysis and deploying a microscopic ion concentration probe array and a mesoscopic acoustic sensor network within the structure, key parameters such as chloride ions, sulfate ions, alkalinity changes, porosity, density, and microcrack propagation can be detected in real time. Combining fiber optic grating technology with strain measurement, a three-dimensional strain monitoring network achieves precise tracking of concrete deformation, cracking, and displacement. Real-time acquisition of external environmental parameters such as temperature, humidity, air pressure, rainfall, and ultraviolet intensity allows for correlation analysis of concrete structure performance changes with external environmental influencing factors. The combination of low-power wide-area IoT technology and edge computing not only improves data transmission efficiency and reduces energy consumption but also ensures efficient data transmission and timely feedback through multi-hop self-organizing network protocols.

[0101] Furthermore, the microscale ion concentration probe array includes...

[0102] The chloride ion sensor has a detection range of 0.01%-5% and an accuracy of ±0.005%.

[0103] The sulfate ion sensor has a detection range of 0.02%-8% and an accuracy of ±0.01%.

[0104] The pH sensor has a detection range of 5-14 and an accuracy of ±0.2.

[0105] Furthermore, the acoustic sensor network consists of 16-64 ultrasonic sensors with a working frequency of 20-100kHz, a measurement accuracy of ±1μs, and can detect microcracks with a width greater than 0.1mm. The sensor spacing is 1-3m, forming a multipath acoustic wave propagation network.

[0106] Furthermore, the environmental parameter acquisition unit specifically includes...

[0107] Temperature sensor with a measurement range of -30℃ to +80℃ and an accuracy of ±0.5℃.

[0108] Humidity sensor with a measurement range of 5-95%RH and an accuracy of ±3%RH.

[0109] The barometer has a measurement range of 300-1100 hPa and an accuracy of ±1 hPa.

[0110] Rain gauge with a measurement range of 0-200 mm / h and an accuracy of ±0.5 mm.

[0111] Ultraviolet intensity detector, with a measurement range of 0-1200μW / cm² and an accuracy of ±10μW / cm².

[0112] The wind speed sensor has a measurement range of 0-50 m / s and an accuracy of ±0.5 m / s.

[0113] The sulfur dioxide concentration sensor has a measurement range of 0-10ppm and an accuracy of ±0.05ppm.

[0114] Furthermore, step 2 includes the following steps.

[0115] Principal component analysis (PCA) was used to reduce the dimensionality of multi-source environmental data. The specific steps included: standardizing and preprocessing multi-dimensional environmental parameters, including temperature, humidity, chloride ion concentration, pH, strain, air pressure, and wind speed, to eliminate dimensional differences and numerical magnitude effects; constructing an environmental data covariance matrix and extracting the top 5-8 principal components through eigenvalue decomposition, achieving a cumulative contribution rate of over 85%, thus reducing data dimensionality while retaining key information; establishing a principal component loading matrix to identify dominant environmental factor combinations such as temperature-humidity coupling and ion concentration-pH correlation; using continuous wavelet transform to decompose environmental time-series data at multiple scales, extracting frequency domain features at different time scales such as 1-24 hours, 1-7 days, and 1-30 days; using discrete wavelet transform for signal denoising and feature extraction, and obtaining time-frequency localized features of environmental parameters through wavelet coefficient reconstruction; and fusing principal component weight coefficients, wavelet transform coefficients, and time-domain statistical features to construct a 128-dimensional environmental state feature vector containing mean, variance, skewness, kurtosis, energy density, and frequency distribution.

[0116] By combining concrete mix proportion parameters, curing conditions, and measured strength grades, a multi-scale material model is used to quantitatively describe the microstructural characteristics of concrete, including pore distribution, hydration product ratio, and interfacial transition zone characteristics. By constructing a comprehensive material state descriptor, the precise digital expression of concrete material properties is achieved.

[0117] An exponentially weighted decay factor is introduced to process historical data and a recursive time series structure is constructed to capture the evolution law of concrete strength. At the same time, the strength development curves at different stages are simulated to achieve a precise mathematical expression of the time-varying characteristics of concrete strength.

[0118] By establishing a mapping relationship between microscopic physicochemical processes and macroscopic mechanical properties through a scale transformation function, the differences in monitoring data at different structural scales can be eliminated, enabling reliable inference from local monitoring points to overall structural performance.

[0119] By introducing performance degradation rate and limit state function, a comprehensive evaluation system based on performance indicators is constructed, and the weights of each indicator are set using reliability theory to form a multi-objective performance evaluation criterion.

[0120] Construct a five-dimensional tensor structure T(E,M,τ,S,P), where each dimension corresponds to environmental factors, material properties, time evolution, structural scale, and performance indicators, respectively.

[0121] T(E,M,τ,S,P) is decomposed into the product of the core tensor and five factor matrices. The main feature patterns of each dimension are extracted, and the coupling strength coefficient between each dimension is calculated to quantitatively characterize the influence of environmental factors and material properties on concrete performance at different times and scales.

[0122] The method involves constructing a comprehensive material state descriptor to achieve a precise digital expression of concrete material properties. Specifically, this includes the following steps: Based on X-ray diffraction and scanning electron microscopy analysis, the relative content and distribution characteristics of cement hydration products CSH gel, calcium hydroxide, and ettringite are quantified. Pore structure parameters, including porosity, average pore size, pore size distribution curve, and specific surface area, are determined using mercury intrusion porosimetry and nitrogen adsorption methods, establishing a digital mapping relationship between pore structure and permeability. The mechanical property gradient distribution of the interfacial transition zone is characterized using nanoindentation technology and microhardness testing, quantifying the aggregate-slurry interfacial bonding strength and micro-elastic modulus. Atom-level interactions from molecular dynamics simulations, representative volume units from micromechanics, and macroscopic continuum mechanics are combined. A material degradation index system is constructed, quantifying the degree of material performance degradation using key parameters such as carbonation depth, chloride ion diffusion coefficient, and sulfate attack rate. Finally, microstructure parameters, mechanical performance indicators, and durability evaluation results are integrated to form a material state description matrix containing 96 key parameters, achieving a comprehensive digital characterization of concrete material properties.

[0123] The process involves establishing a mapping relationship between microscopic physicochemical processes and macroscopic mechanical properties through a scale transformation function. This includes the following sub-steps: establishing a scale bridging function from nanoscale CSH gel structure to micrometer-scale cement paste properties, quantifying the contribution of intermolecular forces to the macroscopic elastic modulus of the material; establishing an effective medium theory for the aggregate-paste-interface three-phase composite system, realizing the mathematical transformation from microscopic component properties to macroscopic composite material properties; correlating the microscopic processes of microcrack initiation and propagation with macroscopic stiffness degradation and strength attenuation, establishing a cross-scale transfer function for damage evolution; training a scale transformation neural network, using microscopic structural parameters and physicochemical indices as input and macroscopic mechanical properties as output, to achieve intelligent modeling of nonlinear mapping relationships; considering the influence of time-varying processes such as hydration reactions, ion migration, and stress redistribution on cross-scale mapping relationships to achieve dynamic scale transformation; and verifying the accuracy of the scale transformation function through multi-scale finite element analysis, establishing a quantitative prediction model for the transformation of microscopic parameter changes to macroscopic performance responses, and achieving reliable inference from local monitoring data to overall structural performance.

[0124] This study introduces performance degradation rate and limit state function to construct a comprehensive evaluation system based on performance indicators. Reliability theory is used to set the weights of each indicator, forming a multi-objective performance evaluation criterion. Specifically, the steps include: quantifying the degradation rate constant and acceleration factor of each performance parameter by performing exponential fitting and power function regression on time-varying monitoring data of key indicators such as strength, stiffness, and durability; constructing a multi-level limit state function system, including the ultimate limit state of bearing capacity, the serviceability limit state, and the ultimate limit state of durability, and establishing structural failure criteria and early warning mechanisms by setting critical thresholds and safety factors; determining the initial weight allocation of strength, deformation, durability, and environmental adaptability indicators; calculating the reliability index and failure probability of each performance indicator, and dynamically adjusting the weight coefficients according to the reliability level to achieve risk-based weight optimization; using the TOPSIS method and fuzzy comprehensive evaluation method to combine quantitative indicators with qualitative evaluation, forming a comprehensive performance evaluation index that includes safety, applicability, and durability; and constructing a dynamic early warning threshold system, setting four warning levels—green safety, yellow warning, orange warning, and red danger—based on performance degradation trends and reliability changes to achieve real-time assessment and risk pre-control of the structural state.

[0125] The above series of steps, employing multi-level and multi-scale analytical methods, constructs a comprehensive framework for concrete performance evaluation. First, principal component analysis and wavelet transform are used to reduce the dimensionality of environmental data and perform time-frequency analysis, effectively extracting key environmental features. Combining concrete mix proportions, curing conditions, and measured strength grades, a multi-scale material model is used to accurately describe the microstructural characteristics of concrete, including pore distribution and interface transition zone characteristics, thus achieving a digital expression of concrete material properties. Simultaneously, by introducing an exponentially weighted decay factor and a recursive time-series structure, the time-varying characteristics of concrete strength are successfully captured, simulating strength development curves at different stages. Furthermore, a scale transformation function is used to eliminate the differences in monitoring data between different structural scales, ensuring reliable inferences from local to overall structural performance. Considering both performance degradation rate and limit state function, a multi-objective performance evaluation system is constructed, and appropriate weights are assigned to each index based on reliability theory. Finally, a five-dimensional tensor structure T(E,M,τ,S,P) is used to extract features of each dimension through tensor decomposition and quantify the coupling strength coefficient, thereby accurately characterizing the influence of environmental factors and material properties on concrete performance.

[0126] Furthermore, to achieve a precise mathematical expression of the time-varying properties of concrete strength, the following steps are included.

[0127] An exponentially weighted attenuation factor α(t) = exp(-λt) is introduced to weight the historical data, where α(t) is the exponentially weighted attenuation factor, representing the degree of attenuation of the influence of historical data on the current strength over time; λ is the attenuation constant, which controls the attenuation rate of the influence of historical data on the current concrete strength.

[0128] Using the recursive formula S recursive (t)=S(t-1)+∑ i=0 t-1 α(ti)·ΔS(i) processes time-series data, accumulating the strength increment ΔS(i) at each time step according to an exponential weighting factor to capture the gradual evolution of concrete strength, where S... recursive (t) represents the concrete strength calculated recursively, meaning the current strength is accumulated from the strength at the previous time step and the weighted increments over historical periods, reflecting the contribution of each time step to the overall strength change; S(t-1) represents the concrete strength value at time t-1; α(ti) represents the exponential weighted decay factor at time step ti, indicating the influence of historical data on the current strength, which gradually decreases over time.

[0129] Use S evolution (t)=S ∞ ·(1-exp(-kt)) describes the evolution of intensity from rapid initial growth to later stabilization, where S evolution (t) represents the ideal evolution curve of concrete strength, which describes the trend of strength change over time using an exponential growth model, emphasizing the overall development process from the initial state to the final steady state; S ∞ This indicates the stable strength value that concrete reaches after a relatively long period of time; k represents the strength growth rate constant, which controls the rate of strength growth of concrete at different stages.

[0130] By combining the recursive time sequence structure with the stage development curve, a comprehensive expression S is constructed. weighted (t)=S0+∑ i=1 t exp(-λ(ti))·ΔS(i)+S ∞ ·(1-exp(-kt)), which realizes a precise mathematical expression for the time-varying characteristics of concrete strength, where S weighted (t) represents the comprehensive strength expression combining the weighted effect of historical data and the overall evolution trend; S0 is the initial strength, representing the initial strength value of concrete at the very beginning.

[0131] In summary, by introducing an exponentially weighted attenuation factor and a recursive time-series structure, a precise mathematical expression of the time-varying characteristics of concrete strength is achieved. First, an exponentially weighted attenuation factor α(t) is used to weight historical data, ensuring that the influence of historical data gradually diminishes over time, thus reflecting the gradual evolution of concrete strength over time. A recursive formula is used to accumulate the strength increments of historical periods according to the weighting factor, capturing the contribution of each moment to the overall strength change. Next, an exponential growth model is used to describe the evolution of concrete strength from rapid initial growth to later stabilization, emphasizing the overall development trend of concrete strength. Finally, by combining the recursive time-series structure with the staged evolution curves, a comprehensive expression is constructed. This formula simultaneously considers the weighting effect of historical data and the overall evolution trend of concrete strength, providing a precise mathematical model of the time-varying characteristics of concrete strength, which helps to more accurately predict and analyze the strength changes of concrete at different stages.

[0132] Furthermore, the five-dimensional tensor decomposition employs the penalized term Tucker decomposition algorithm, wherein...

[0133] The decomposition orders are: environmental factors (2-4), material properties (3-6), time evolution (2-3), structural scale (2-4), and performance index (3-5).

[0134] An L1 regularization term is introduced to control the sparsity of the factor matrix, with the regularization coefficient ranging from 0.02 to 0.15.

[0135] The alternating least squares method is used to solve the decomposition problem, with a convergence threshold of 10. -4 The maximum number of iterations is 200.

[0136] By calculating the element distribution characteristics in the Tucker kernel tensor, the coupling degree between various dimensions is quantified, and the dominant interaction pattern is extracted.

[0137] Furthermore, step 3 includes the following steps.

[0138] A piecewise nonlinear model for concrete strength development is established, dividing the entire age period into a rapid growth period, a transition period, and a stable period. Each stage is connected by a piecewise continuous function to achieve a precise mathematical expression of the strength evolution throughout the entire age period.

[0139] A recursive prediction model based on a long short-term memory network is constructed. Historical monitoring data and five-dimensional tensor feature expression are used as inputs to learn the intrinsic law of concrete strength development and realize intelligent prediction and real-time correction of the strength development trajectory.

[0140] By integrating and accumulating environmental impact factors, the actual calendar age is converted into an equivalent age that accurately reflects the true degree of hydration, thereby achieving a unified expression of intensity development curves under different environmental conditions.

[0141] An exponentially weighted moving average method is used to enhance the importance of recent data, and a forgetting factor is set to gradually reduce the influence of long-term data;

[0142] An exponentially weighted moving average method is used to enhance the importance of recent data. A forgetting factor ω (0 < ω < 1) is set to gradually reduce the influence of long-term data. Historical monitoring data are weighted using the recursive formula S(t) = ω × X(t) + (1-ω) × S(t-1), where the weight of recent data increases exponentially and the weight of long-term data decreases exponentially, achieving adaptive smoothing of time-series data. At the same time, the model parameters are forcibly calibrated at key age nodes to ensure that the prediction results are highly consistent with the actual intensity development.

[0143] The percentage deviation formula Δ=|Measured value - Theoretical value| / Theoretical value×100% is used to quantify the degree of strength deviation. Slight deviation thresholds (5-10%), moderate deviation thresholds (10-20%), and severe deviation thresholds (>20%) are set, and the threshold boundaries are dynamically adjusted according to the structural importance level and the severity of the service environment to achieve graded early warning management. When the deviation exceeds the allowable range, the anomaly analysis process is triggered, and Bayesian inference methods are introduced to identify influencing factors to distinguish between normal strength fluctuations and potential quality problems.

[0144] Real-time monitoring data is aggregated and analyzed at different time scales, and wavelet transform is used to extract feature patterns at different time scales. At the same time, a cross-scale time correlation network is established to realize full-spectrum monitoring and prediction from short-term intensity fluctuations to long-term performance evolution.

[0145] The process involves aggregating and analyzing real-time monitoring data across different time scales, extracting feature patterns at different time scales using wavelet transform, and establishing a cross-scale time correlation network. The steps include: dividing the real-time monitoring data into daily, weekly, and monthly time scales based on the time-dependent characteristics of concrete strength evolution, each used to capture performance change trends at different stages; statistically aggregating the raw data within each time scale, such as extracting the mean, variance, and peak value, to reduce high-frequency noise interference and improve feature stability; applying discrete wavelet transform to the aggregated multi-scale data to separate signal features in different frequency bands, revealing the implicit periodicity and abrupt changes in concrete performance evolution; extracting the main change patterns at each time scale from the wavelet decomposition results, constructing multi-dimensional feature vectors including abrupt changes, periodic changes, and trend changes; and establishing a dynamic correlation network between features at different scales through graph structures or time recursion mechanisms to reflect the impact path of short-term fluctuations on medium- and long-term performance evolution.

[0146] In summary, by constructing a piecewise nonlinear model and a recursive prediction model based on a Long Short-Term Memory (LSTM) network, accurate prediction and real-time correction of concrete strength development were achieved. First, the concrete age was divided into a rapid growth period, a transition period, and a stable period, and a piecewise continuous function was used to accurately represent the strength evolution process at each stage. Next, an LSTM network was used to combine historical monitoring data with five-dimensional tensor features to learn the inherent laws of strength development, intelligently predict the strength development trajectory, and convert the actual calendar age into an equivalent age through the integral accumulation of environmental impact factors, uniformly representing the strength curve under different environmental conditions. Simultaneously, an exponentially weighted moving average method was used to enhance the importance of recent data, and a forgetting factor was set to gradually reduce the influence of long-term data, ensuring prediction accuracy. At key age nodes, model parameters were forcibly calibrated to ensure a high degree of consistency between the predicted results and the actual strength. By calculating the deviation between the measured strength and the theoretical curve, an early warning threshold was set, and a Bayesian inference method was used to identify potential quality problems, distinguishing between normal fluctuations and abnormal situations. Furthermore, wavelet transform is used to extract feature patterns at different time scales, and a cross-scale time correlation network is established to provide full-spectrum monitoring and prediction from short-term strength fluctuations to long-term performance evolution, further enhancing the intelligent prediction capability of concrete strength time-varying characteristics.

[0147] Furthermore, the Long Short-Term Memory network has the following structure and parameters.

[0148] The input layer has dimensions of 20-40, corresponding to environmental parameters and material property feature vectors.

[0149] 2-3 LSTM hidden layers, each containing 32-128 neurons.

[0150] A fully connected output layer outputs intensity predictions for the next 7, 30, and 90 days.

[0151] The Adam optimizer was used for training, with an initial learning rate of 0.001-0.003 and a learning rate decay factor of 0.85-0.95 every 50 rounds.

[0152] The root mean square error was used as the loss function, the ratio of training data to validation data was 8:2, and the early stopping mechanism was triggered when the loss on the validation set did not improve for 10 consecutive rounds.

[0153] Furthermore, by accumulating environmental impact factors through integration, the actual calendar age is converted into an equivalent age that accurately reflects the true degree of hydration, thereby achieving a unified expression of intensity development curves under different environmental conditions, including the following steps.

[0154] The environmental impact factor λ(θ) at each time θ is calculated based on environmental parameters and expressed by the formula: Where T(θ) is the ambient temperature at time θ; R is the gas constant; E a RH is the activation energy, used to describe the sensitivity of the hydration reaction to temperature; RH(θ) is the ambient relative humidity at time θ; RH ref The reference humidity value is used to normalize the humidity effect; μ is the humidity sensitivity index, which describes the effect of humidity on the hydration rate; γ(θ) is other environmental correction factors.

[0155] By integrating and summing the environmental impact factors, the actual calendar age is converted into an equivalent age t. e The formula is: t e =∫0 t λ(θ)dθ, where dθ is a small time increment used for integral calculation of the equivalent age.

[0156] Based on the calculated equivalent age t e The original intensity development curve was corrected.

[0157] By adjusting the parameters of the environmental impact factor λ(θ) under different environmental conditions, the intensity development curve can be adapted to changes in temperature, humidity and other environmental factors, thereby achieving unified and accurate intensity prediction.

[0158] In summary, by introducing an integral summation method for environmental impact factors, the actual calendar age is converted into an equivalent age reflecting the true degree of hydration, thereby accurately unifying the intensity development curves under different environmental conditions. Specifically, firstly, the environmental impact factors at each moment are calculated based on environmental parameters (such as temperature and relative humidity), and quantified using formulas, taking into account the influence of temperature, humidity, and other correction factors on the hydration rate. Then, by integrating and summing the environmental impact factors, the equivalent age is calculated, thereby correcting the original intensity development curve and ensuring that the intensity curve accurately reflects the hydration process under different environments. By adjusting the parameters of the environmental impact factors, this method can flexibly adapt to changes in temperature, humidity, and other environmental factors, ensuring the uniformity and accuracy of the intensity development curve, and achieving accurate intensity prediction under variable environmental conditions.

[0159] Furthermore, step 4 includes the following steps.

[0160] By calculating the environment-material sensitivity matrix, the sensitivity of different types of concrete to environmental factors is quantified.

[0161] A piecewise temperature compensation function is established, dividing the temperature range into low-temperature, normal-temperature, and high-temperature zones. Different mathematical models are used in each zone to describe the influence of temperature on the hydration reaction rate, thereby achieving accurate prediction of concrete strength development across the entire temperature range.

[0162] By combining the size effect of concrete components, a humidity gradient-intensity distribution mapping relationship is constructed, thereby enabling reliable inference from surface humidity monitoring to internal intensity distribution.

[0163] Environmental conditions are quantitatively characterized by the comprehensive environmental impact index and used as a standardized input parameter.

[0164] The dynamic relationship between the strength development rate index and the environmental comprehensive effect index is described by differential equations, and the SDR-ECI response surface is constructed to quantify the acceleration or deceleration effect of strength development under different environmental conditions. At the same time, a material sensitivity factor is introduced to adjust the shape of the response surface to adapt to the characteristics of different concrete mix proportions, so as to achieve accurate prediction of the strength development law in changing environments.

[0165] Establish the correlation between equivalent age and actual strength development, calculate the strength contribution rate at different time periods, and establish a strength history evolution archive to record the long-term impact of key environmental events on concrete strength development.

[0166] The study quantifies the sensitivity of different types of concrete to environmental factors by calculating an environment-material sensitivity matrix. This includes the following steps: selecting representative environmental factors, such as temperature, humidity, chloride ion concentration, sulfate content, and carbon dioxide concentration, and normalizing them to ensure comparability across variables on a uniform scale; classifying samples based on concrete mix proportions, strength grades, admixture types, and curing methods to establish a basic database of multiple concrete material types; selecting response indicators for concrete under various environmental conditions, such as strength decay rate, crack development rate, and carbonation depth, and extracting quantitative response results based on long-term monitoring data; calculating the sensitivity coefficients between environmental factors and concrete responses to form a numerical sensitivity matrix; visualizing the sensitivity coefficient matrix using a heatmap to identify key coupling relationships between highly sensitive environmental factors and corresponding concrete types, and performing cluster analysis on different material types; and classifying concrete materials into high, medium, and low sensitivity levels based on their sensitivity intensity.

[0167] Specifically, the Arrhenius model is used for the normal and slightly higher temperature range of 10℃ to 60℃; the Hyperbolic Tangent model is used for the low temperature range of <10℃; and the Gaussian model is used for the high temperature range of >60℃.

[0168] The construction process of the comprehensive environmental impact index includes: based on the sensitivity analysis results of concrete strength evolution, screening out environmental parameters that have a significant impact on performance to form a unified set of input variables; performing dimensionless processing on the selected parameters to ensure their values ​​are in the range of [0,1]; designing a nonlinear weighting mechanism to enhance the sensitivity to changes in key parameters and suppress the fluctuations of secondary factors based on the response relationship between different parameters and concrete performance; and setting the weight coefficients corresponding to each parameter based on historical monitoring data to combine the standardized parameters and form the comprehensive environmental impact index.

[0169] In summary, by establishing an environment-material sensitivity matrix, a piecewise temperature compensation function, and a humidity gradient-strength distribution mapping relationship, the strength development of concrete under different environmental conditions can be accurately predicted. First, by quantifying the sensitivity of concrete to environmental factors, environmental adaptability analysis of different types of concrete is achieved. Then, a piecewise temperature compensation function is constructed, employing different mathematical models to describe the influence of temperature on the hydration reaction rate in different temperature ranges, ensuring accurate strength prediction across the entire temperature range. Simultaneously, by combining the component size effect, the relationship between humidity gradient and strength distribution is established, enabling inference from surface humidity monitoring to internal strength. By designing an environmental comprehensive effect index and integrating key environmental parameters using a nonlinear weighting function, a unified index is formed to quantify environmental impact. This scheme describes the dynamic relationship between strength development rate and environmental impact through differential equations and SDR-ECI response surfaces, and introduces a material sensitivity factor to adapt to the characteristics of different concrete mix proportions, achieving accurate prediction of strength development patterns. Furthermore, the correlation between equivalent age and actual strength development is established, recording the long-term impact of key environmental events on strength evolution and providing a comprehensive historical evolution archive.

[0170] Furthermore, by combining the size effect of concrete components, a humidity gradient-intensity distribution mapping relationship is constructed to achieve a reliable inference from surface humidity monitoring to internal intensity distribution, including the following steps.

[0171] Using the humidity gradient formula G H (x)=(H(x)-H0) / x calculates the rate of change of humidity at different depths, thus characterizing the diffusion rate of moisture inside the component, where G H H(x) represents the rate of change of humidity at depth x, that is, the change of humidity with depth, which is used to reflect the diffusion rate of moisture in concrete components; H(x) represents the humidity value at depth x inside the concrete; H0 represents the humidity value on the surface of the concrete component.

[0172] Based on the component dimension D, we introduce the size effect function f(D) = (D / D0) -δThe effect of humidity on strength is corrected to reflect the strength change law caused by the decrease of humidity gradient in large-scale components. Here, f(D) is the size effect function, which is used to describe the correction effect of the size of the concrete component on the strength; D0 is the reference size, which is used to standardize the component size D; δ is the size effect correction factor, which represents the degree of influence of the component size on the strength.

[0173] Based on the strength formula σ(x)=σ0(1+ηG) H (x)) β f(D) couples the humidity gradient with the size effect to establish a mapping relationship between humidity gradient and strength distribution. Here, σ(x) represents the strength value of concrete at depth x; σ0 represents the strength value of concrete surface; η represents the sensitivity coefficient of the influence of humidity gradient on strength, reflecting the degree of influence of humidity change on concrete strength; β represents the nonlinear influence coefficient of humidity gradient on strength, which determines the amplitude of strength response when humidity gradient changes.

[0174] By using the measured surface humidity and the calculated humidity gradient, combined with the mapping relationship, the intensity distribution at different depths is calculated, enabling non-destructive inference of the internal strength of the component.

[0175] Furthermore, by incorporating the size effect of concrete components, a mapping relationship between humidity gradient and strength distribution was established, enabling reliable inference from surface humidity monitoring to internal strength distribution. First, the rate of change of humidity at different depths was calculated using the humidity gradient formula to characterize the diffusion rate of moisture within the component and reflect the impact of moisture on strength development. Then, considering the influence of component size on strength, a size effect function was used to correct for the effect of humidity on strength, especially in large-scale components where a decrease in the humidity gradient may lead to strength changes. Next, combining the humidity gradient and size effect, a mapping relationship between humidity gradient and strength distribution was established based on the strength formula. Finally, by applying the mapping relationship using measured surface humidity and calculated humidity gradient, the internal strength distribution of the component was inferred, thus achieving non-destructive strength assessment.

[0176] Furthermore, the dynamic relationship between the intensity development rate index and the comprehensive environmental impact index is described by differential equations, and the SDR-ECI response surface is constructed to quantify the acceleration or mitigation effects of intensity development under different environmental conditions, including the following steps.

[0177] By establishing the relationship between the strength development rate and time through differential equations, the dynamic law of concrete strength change over time can be accurately described.

[0178] Based on the dynamic changes of various environmental factors, a weighted function is used to fuse them to form a comprehensive environmental impact index.

[0179] The dynamic relationship between intensity development rate and the comprehensive environmental impact index is described by differential equations. A coupling equation between the two is established to quantify the direct impact of different environmental conditions on intensity development rate.

[0180] The relationship between intensity development rate and environmental comprehensive effect index is transformed into a response surface model. Numerical optimization methods are used to fit the intensity development changes under different environmental conditions and identify the sensitivity of each influencing factor.

[0181] Based on response surface analysis, the changing trend of intensity development rate under different environmental conditions is analyzed, and the acceleration or deceleration effect of intensity development under specific environmental conditions is quantified.

[0182] Furthermore, a material sensitivity factor is introduced to adjust the response surface morphology to adapt to the characteristics of different concrete mix proportions, thereby achieving accurate prediction of the strength development law in changing environments, including the following steps.

[0183] The influence of different concrete mix proportions on the strength development rate was analyzed, a material sensitivity factor was established, and the degree of influence of different mix proportions on strength evolution was quantified.

[0184] Based on the material properties of different proportions, a material sensitivity factor is introduced to adjust the shape of the response surface, so that it can dynamically reflect the differences in strength development of different material combinations under the same environmental conditions.

[0185] A mapping relationship between different concrete mix proportions and response surfaces is constructed to ensure that the surface can accurately reflect the influence of changes in concrete mix proportions on the strength development law.

[0186] Based on real-time monitored environmental and material data, the values ​​of the material sensitivity factor are adjusted so that the response surface can adapt to the strength development trend under different environmental conditions.

[0187] The strength development of concrete with different mix proportions is predicted by using dynamically adjusted response curves. By comparing with measured data, the selection of material sensitivity factors is optimized.

[0188] Through continuous monitoring and data feedback, the material sensitivity factors are constantly updated to ensure that the strength development law can be accurately predicted and adjusted in real time under different environmental changes and mixing ratios.

[0189] Furthermore, the correlation between equivalent age and actual intensity development is established, the intensity contribution rate at different time periods is calculated, and an intensity historical evolution archive is established, including the following steps.

[0190] Using the equivalent age, the evolution of concrete strength is described as f(t) = f0·(1-exp(-k1·t) e(t)), where f(t) represents the actual strength of concrete at time t; f0 represents the maximum strength of concrete under the theoretical limit condition; k1 represents the rate constant of concrete strength development; t e (t) represents the equivalent age at time t.

[0191] Within the time interval [t1, t2], the formula C(t1, t2) = [f(t2) - f(t1)] / [f(t2)] is used. max The contribution rate of each stage to the total strength growth is calculated, quantifying the relative impact of each stage on the evolution of concrete strength. Here, C(t1,t2) represents the contribution rate of concrete strength development within the time period [t1,t2]; f(t2) represents the actual strength at time t2; f(t1) represents the actual strength at time t1; and f(t... max ) indicates the maximum age t max At that time, the final strength of the concrete.

[0192] For critical environmental incidents, the environmental impact factor G will be used. r (t) is introduced, through An archive is constructed to systematically record the cumulative impact of long-term environmental events on changes in concrete strength, including f. history (t,G) represents the historical strength evolution of concrete at time t due to environmental influence G; α r This represents the influence coefficient of the r-th environmental factor on the development of concrete strength.

[0193] Furthermore, step 5 includes the following steps.

[0194] The first and second derivatives of the intensity development rate are calculated in real time. The first derivative reflects the direction of the rate change, and the second derivative represents the magnitude of the acceleration. When an inflection point or increased fluctuation in the intensity development rate is detected, the optimal allocation of monitoring resources is achieved by intelligently adjusting the sampling density.

[0195] By integrating parameters such as temperature change rate, humidity gradient, carbon dioxide concentration fluctuation and chloride ion permeation rate, and using a multivariate weighting function to calculate a comprehensive sensitivity score, the system achieves accurate quantification and hierarchical monitoring of changes in environmental conditions.

[0196] Establish normal fluctuation ranges for environmental parameters, take into account seasonal variation characteristics and geographical location factors, dynamically update threshold settings, and identify environmental change patterns to distinguish between normal seasonal fluctuations and abnormal climate events.

[0197] A multi-level triggering mechanism is constructed, which divides environmental parameter fluctuations into three levels: attention level, warning level, and emergency level. At the same time, multiple parameter combination triggering conditions are considered to achieve intelligent response to complex environmental changes.

[0198] Lightweight computing modules are deployed at on-site monitoring nodes to perform real-time data preprocessing and preliminary analysis. When abnormal environmental conditions are detected, the local sampling strategy is adjusted immediately, and a data caching mechanism is activated to ensure that critical data can still be saved in the event of communication interruption. A clear task division and data synchronization mechanism is established with the central data processing center.

[0199] When a specific environmental event is detected, the system automatically enters a high-frequency sampling mode, activates additional environmental monitoring units, comprehensively records data throughout the entire process of the environmental event, performs a special assessment of concrete strength after the event ends, and constructs a causal relationship database between environmental event and strength change.

[0200] In summary, by calculating the derivative of the intensity development rate in real time, the dynamic trend of intensity change is monitored, and the allocation of monitoring resources is optimized through intelligent adjustment of sampling density. A comprehensive sensitivity analysis is performed by integrating multiple environmental parameters (such as temperature change rate, humidity gradient, carbon dioxide concentration fluctuation, and chloride ion permeation rate) to accurately quantify environmental changes and conduct tiered monitoring. Simultaneously, a dynamically updated environmental parameter fluctuation threshold and a multi-level triggering mechanism are established, effectively identifying normal seasonal fluctuations and abnormal climate events, and providing intelligent responses. Furthermore, a lightweight computing module is deployed to process data in real time, respond to environmental anomalies, and ensure data storage, maintaining critical data even during communication interruptions. When specific environmental events are triggered, a high-frequency sampling mode is automatically activated, and a causal database between environmental events and intensity changes is established, providing accurate evidence for subsequent intensity assessments.

[0201] Furthermore, the specific parameters of the multi-level triggering mechanism are as follows:

[0202] Attention Level: A single environmental parameter exceeds the normal range by 1.5 times, or the overall sensitivity score reaches 0.6, and the sampling frequency is increased to 2 times / day.

[0203] Warning level: A single environmental parameter exceeds 2.0 times the normal range, or the overall sensitivity score reaches 0.75, or more than 2 environmental parameters simultaneously reach the attention level threshold. The sampling frequency is increased to 4 times / day, and a special intensity assessment procedure is initiated.

[0204] Emergency Level: A single environmental parameter exceeds 2.5 times the normal range, or the overall sensitivity score exceeds 0.85, or more than 3 environmental parameters simultaneously reach the warning level threshold. The sampling frequency is increased to 12 times / day, and the backup sensor is activated and an early warning notification is sent.

[0205] Furthermore, the lightweight computing module has the following characteristics.

[0206] Based on ARM Cortex-M series processors, with a main frequency of 40-80MHz and RAM capacity of 64-128KB.

[0207] It adopts a low-power operation mode with an average power consumption of less than 50mW.

[0208] Perform local data preprocessing tasks such as basic statistical analysis and simple feature extraction.

[0209] It has a 1-3MB circular cache, which can save 48-96 hours of critical data in the event of communication interruption.

[0210] An incremental data transmission protocol is adopted, which only sends changed data and abnormal features, reducing the amount of data transmission by 60-80%.

[0211] Furthermore, step 6 includes the following steps.

[0212] A multi-source heterogeneous data fusion framework was constructed to extract abnormal feature patterns from various types of sensor data and to establish a concrete strength evolution feature library.

[0213] It enables rolling forecasts of intensity values ​​for the next 7 days, 30 days, and 90 days.

[0214] Based on the structural importance level and service environment characteristics, a regional and graded strength safety threshold system is established. At the same time, the early warning threshold parameters are automatically calibrated in combination with the characteristics of seasonal environmental changes, and the safety margin is adjusted in a timely manner according to the structural aging rate.

[0215] Different levels of early warning response are triggered based on the degree of intensity decline: Level 1 warning is triggered when the intensity is lower than 90% of the design value, reminding people to pay attention to potential risks; Level 2 intervention warning is triggered when the intensity is lower than 85%, recommending preventive measures; Level 3 emergency warning is triggered when the intensity is lower than 80% or the rate of intensity decline exceeds 0.5 MPa / day, requiring immediate intervention measures to prevent further deterioration.

[0216] Based on the finite element mesh generation principle, a three-dimensional distribution cloud map of structural strength is generated, and a structural health assessment report containing strength numerical analysis, environmental impact assessment, safety risk classification, and maintenance recommendations is automatically generated.

[0217] By combining strength development trends and environmental sensitivity analysis, maintenance decision recommendations are automatically generated, including adjusting the inspection frequency, strengthening protective measures, and necessary reinforcement schemes, to ensure the long-term stability and safety of concrete structures in complex environments.

[0218] In summary, by constructing a multi-source heterogeneous data fusion framework, extracting abnormal features from sensor data, and establishing a concrete strength evolution feature library, rolling predictions of strength values ​​for the next 7, 30, and 90 days were achieved. Based on the structural importance and service environment characteristics, a regional and graded strength safety threshold system was established, capable of automatically calibrating and adjusting early warning threshold parameters. By setting a multi-level early warning response mechanism, different levels of warnings are triggered according to the degree of strength decline, ensuring timely intervention and preventing further deterioration. Furthermore, a three-dimensional strength distribution cloud map was generated using the finite element mesh principle, and a structural health assessment report was automatically generated. Combined with strength development trends and environmental sensitivity analysis, detailed maintenance decision recommendations were provided to ensure the long-term stability and safety of concrete structures in complex environments.

[0219] Furthermore, the multi-source heterogeneous data fusion framework includes...

[0220] Data-level fusion layer: Median filtering and simplified Kalman filtering are used to preprocess the raw data, reducing measurement noise by 10-20dB.

[0221] Feature-level fusion layer: Principal component analysis is used to extract key features from the data of each sensor, retaining more than 90% of the information content, while reducing the feature dimension by 50-70%.

[0222] Decision-level fusion layer: Combining fuzzy reasoning and support vector machines to build an integrated decision-making mechanism, comprehensively considering the reliability and confidence of each feature, the decision accuracy reaches 85-90%.

[0223] Furthermore, the rolling prediction employs the following strategy.

[0224] The 7-day forecast results are updated every 24 hours, with a forecast accuracy of ±3.5 MPa.

[0225] The 30-day forecast results are updated every 72 hours, with a forecast accuracy of ±5MPa.

[0226] The 90-day forecast results are updated every 7 days, with a forecast accuracy of ±7MPa.

[0227] An adaptive confidence interval is introduced, which automatically adjusts the interval width based on historical prediction accuracy, with a confidence level of 90%.

[0228] Furthermore, the method for generating the three-dimensional distribution cloud map is as follows.

[0229] Based on tetrahedral mesh generation technology, 5000-20000 cells are adaptively generated according to the structural complexity.

[0230] Each grid cell is associated with data from 2-3 nearby sensor nodes.

[0231] The intensity value within the mesh element is calculated using linear interpolation.

[0232] The Laplace smoothing algorithm is used to consider spatial correlation and smooth the transition between adjacent cells.

[0233] A three-color coloring scheme is adopted: red indicates dangerous areas with an intensity lower than 80% of the design value, yellow indicates warning areas with an intensity of 80%-90% of the design value, and green indicates safe areas with an intensity higher than 90% of the design value.

[0234] A simplified 3D rendering algorithm is used to achieve the desired visualization effect, ensuring smooth display on ordinary mobile devices.

[0235] like Figure 2 As shown, the purpose of this application is also to provide a remote monitoring system for concrete strength suitable for complex environments, including the following modules.

[0236] The multi-level intelligent sensing and monitoring module enables comprehensive monitoring of the internal chemical environment, acoustic properties, and strain distribution of concrete through a sensor network at three levels: microscopic, mesoscopic, and macroscopic.

[0237] The environmental parameter acquisition and communication module is responsible for acquiring environmental parameters around the concrete structure and transmitting the data to the processing center via low-power IoT technology.

[0238] The data processing module is used to realize a unified mathematical expression of the coupling relationship between environmental factors, material properties, time evolution, structural scale and performance indicators.

[0239] The time-series strength evolution prediction module enables full-spectrum prediction of concrete strength from short-term fluctuations to long-term evolution under different environmental conditions, accurately capturing nonlinear time-varying characteristics.

[0240] The strength development rate assessment module is used to quantify the impact of different environmental conditions on the strength development rate of concrete, and to achieve accurate assessment of the age effect.

[0241] The adaptive monitoring and control module, based on intensity development rate derivative analysis and environmental sensitivity scoring, intelligently adjusts the sampling frequency and tracks environmental events, achieving optimal allocation of monitoring resources and accurate capture of key environmental changes.

[0242] The structural safety status assessment module integrates multi-source heterogeneous data to perform rolling intensity prediction, establishes a hierarchical safety threshold system and a multi-level early warning response mechanism, and generates a three-dimensional intensity distribution cloud map and a health assessment report.

[0243] In summary, a multi-level intelligent sensing and monitoring module enables comprehensive monitoring of the internal chemical environment, acoustic properties, and strain distribution of concrete. Combined with an environmental parameter acquisition and communication module, data is transmitted via low-power IoT technology. The data processing module uniformly expresses the coupling relationship between environmental factors, material properties, temporal evolution, structural scale, and performance indicators. The time-series strength evolution prediction module accurately predicts short-term fluctuations and long-term evolution of concrete strength, while the strength development rate assessment module quantifies the impact of different environmental conditions on the strength development rate. The adaptive monitoring and control module intelligently adjusts the sampling frequency, optimizes resource allocation, and captures key environmental changes. The structural safety status assessment module combines multi-source data to perform rolling strength prediction, establishes graded safety thresholds and early warning mechanisms, generates a three-dimensional strength distribution cloud map and a health assessment report, and comprehensively evaluates the safety of the concrete structure.

[0244] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for remote monitoring of concrete strength suitable for complex environments, characterized in that, The method comprises the following steps: Step 1, obtaining performance data of the concrete structure and environmental factor data related thereto, and transmitting the multi-source heterogeneous data to a data processing center in real time through a preset wireless communication protocol; Step 2, constructing a five-dimensional tensor data structure, and realizing accurate mathematical expression of the complex coupling relationship between the environmental factors and the material properties through tensor decomposition; Step 3, introducing a time dimension recursive updating mechanism to capture the nonlinear time-varying characteristics of the concrete strength evolution; Step 4, quantifying the age effect through a strength development rate index; Step 5, adaptively adjusting the data sampling frequency and monitoring the environmental condition fluctuation based on the change trend of the strength development rate; Step 6, real-time evaluation of the safety state of the concrete structure to ensure the safe and reliable operation of the concrete structure under complex environment; Step 2 comprises the following steps: Dimensionality reduction of environmental data is performed by principal component analysis, main variables are extracted, and time-frequency analysis is performed by wavelet transform to construct a feature vector of the environmental state; Combining the concrete mix proportion parameters, curing conditions and measured strength grade, the microstructure characteristics of the concrete are quantitatively described based on the material multi-scale model, including pore distribution, hydration product ratio and interface transition zone characteristics, and the comprehensive material state is constructed to realize accurate digital expression of the concrete material properties; An exponential weighted decay factor is introduced to process historical data, and a recursive time sequence structure is constructed to capture the evolution law of the concrete strength; meanwhile, the strength development curves at different stages are simulated to realize accurate mathematical expression of the time-varying characteristics of the concrete strength; A scale conversion function is used to establish the mapping relationship between the micro physical and chemical processes and the macro mechanical properties to eliminate the differences in the monitoring data at different structural scales and realize reliable inference from local monitoring points to overall structural performance; A performance degradation rate and a limit state function are introduced to construct a comprehensive evaluation system based on performance indicators, and the reliability theory is used to set the weights of the indicators to form a multi-objective performance evaluation criterion; A five-dimensional tensor structure T(E, M, τ, S, P) is constructed, wherein each dimension corresponds to an environmental factor, a material property, a time evolution, a structural scale and a performance indicator; T(E, M, τ, S, P) is decomposed into the product form of a core tensor and five factor matrices, the main feature patterns of each dimension are extracted, and the coupling strength coefficients between the dimensions are calculated to quantitatively represent the influence of the environmental factors and the material properties on the concrete performance at different times and different scales; Step 4 comprises the following steps: The environmental-material sensitivity matrix is calculated to quantify the sensitivity of different types of concrete to environmental factors; A segmented temperature compensation function is established to divide the temperature range into low-temperature, normal-temperature and high-temperature zones, and different mathematical models are used in each zone to describe the influence of temperature on the hydration reaction rate, thereby realizing accurate prediction of the concrete strength development in the whole temperature range; A humidity gradient-strength distribution mapping relationship is constructed in combination with the size effect of the concrete member to realize reliable inference from surface humidity monitoring to internal strength distribution; An environmental comprehensive action index is used to quantitatively represent the environmental conditions and as a standardized input parameter; The dynamic relationship between the strength development rate index and the environmental comprehensive action index is described by a differential equation, and an SDR-ECI response surface is constructed to quantify the acceleration or deceleration effect of strength development under different environmental conditions; at the same time, a material sensitivity factor is introduced to adjust the response surface form to adapt to the characteristics of different concrete mix proportions, so as to realize the accurate prediction of the strength development law in the changing environment; The correlation between equivalent age and actual strength development is established, the strength contribution rate in different time periods is calculated, and a strength history evolution file is established to record the long-term influence of key environmental events on the strength development of concrete.

2. The method for remote monitoring of concrete strength in complex environments according to claim 1, wherein, Step 1 includes the following steps: Based on finite element analysis, the distribution of key monitoring points of the structure is determined, the monitoring arrangement scheme is optimized combining with the load transmission path and stress concentration area, and a multi-scale covering monitoring network topology structure is constructed; Micro-level ion concentration probe arrays are embedded in the concrete, ion-selective electrode technology is used to detect the changes of chloride ions, sulfate ions and alkalinity in the pore solution in real time, and pH sensors are arranged to monitor the evolution of the alkaline environment of concrete; Mesoscopic-level acoustic sensor networks are deployed at key sections of the structure, and by analyzing the acoustic propagation time and waveform characteristics, the development of the density, porosity and micro-cracks in the concrete is evaluated; Combining fiber Bragg grating technology and resistance strain measurement, a three-dimensional strain monitoring network is constructed to realize real-time monitoring of concrete deformation, cracking and displacement changes, and a dynamic map of strain field distribution is established; Environmental parameter acquisition units are deployed outside the structure, including high-precision temperature and humidity sensors, barometers, rain gauges and ultraviolet intensity detectors to construct a microclimate monitoring network; at the same time, groundwater level and soil parameters are collected to establish a complete environmental influence factor database; Low-power wide-area Internet of Things technology is used to construct a data transmission network, and edge computing technology is used to preprocess and compress the original data, and through multi-hop self-organizing network protocol, the data is efficiently transmitted to the data processing center.

3. The method for remote monitoring of concrete strength in complex environments according to claim 2, wherein, The micro-level ion concentration probe array includes: a chloride ion sensor with a detection range of 0.01%-5% and an accuracy of ±0.005%; a sulfate ion sensor with a detection range of 0.02%-8% and an accuracy of ±0.01%; and a pH sensor with a detection range of 5-14 and an accuracy of ±0.2; The acoustic sensor network is composed of 16-64 ultrasonic sensors with a working frequency of 20-100 kHz and a measurement accuracy of ±1μs, which can detect micro-cracks with a width greater than 0.1mm, and the sensor spacing is 1-3m, forming a multi-path acoustic propagation network; The environmental parameter acquisition unit specifically includes: a temperature sensor with a measurement range of -30℃ to +80℃ and an accuracy of ±0.5℃; a humidity sensor with a measurement range of 5-95%RH and an accuracy of ±3%RH; a barometer with a measurement range of 300-1100hPa and an accuracy of ±1hPa; a rain gauge with a measurement range of 0-200mm / h and an accuracy of ±0.5mm; an ultraviolet intensity detector with a measurement range of 0-1200μW / cm² and an accuracy of ±10μW / cm²; a wind speed sensor with a measurement range of 0-50m / s and an accuracy of ±0.5m / s; and a sulfur dioxide concentration sensor with a measurement range of 0-10ppm and an accuracy of ±0.05ppm.

4. The method for remote monitoring of concrete strength in complex environments of claim 1, wherein, Step 3 includes the following steps: A segmented nonlinear model of concrete strength development is established, the entire age period is divided into a rapid growth period, a transition period, and a stable period, and each stage is connected by a segmented continuous function to achieve accurate mathematical expression of the strength evolution throughout the age period; A recursive prediction model based on a long short-term memory network is constructed, historical monitoring data and five-dimensional tensor feature expression are taken as inputs, the internal law of concrete strength development is learned, and intelligent prediction and real-time correction of the strength development trajectory are realized; By integrating environmental impact factors, the actual calendar age is converted into an equivalent age that accurately reflects the true hydration degree, thereby realizing unified expression of the strength development curve under different environmental conditions; The exponential weighted moving average method is used to enhance the importance of recent data, and a forgetting factor is set to gradually reduce the influence of long-term data; at the same time, model parameters are calibrated at key age nodes to ensure that the prediction results are highly consistent with the actual strength development; By calculating the deviation of the measured strength from the theoretical curve, multi-level warning thresholds are set, when the deviation exceeds the allowed range, the abnormal analysis process is triggered, and the Bayesian inference method is introduced to identify the influencing factors, to distinguish between normal strength fluctuations and potential quality problems; Real-time monitoring data is aggregated and analyzed according to different time scales, and wavelet transform is used to extract feature patterns at different time scales, and a cross-scale time correlation network is established to realize full-spectrum monitoring and prediction from short-term strength fluctuations to long-term performance evolution.

5. The method for remote monitoring of concrete strength in complex environments of claim 1, wherein, Step 5 includes the following steps: The first and second derivatives of the strength development rate are calculated in real time, the first derivative reflects the direction of rate change, and the second derivative represents the acceleration size, when the strength development rate is detected to have a turning point or intensified fluctuations, the optimal allocation of monitoring resources is realized through intelligent control of sampling density; Temperature change rate, humidity gradient, carbon dioxide concentration fluctuation, and chloride ion permeation rate parameters are integrated, a multivariate weight function is used to calculate the comprehensive sensitivity score, and accurate quantification and hierarchical monitoring of environmental condition changes are realized; The normal fluctuation range of environmental parameters is established, and seasonal variation characteristics and geographical location factors are considered, the threshold settings are dynamically updated, and environmental change patterns are identified, thereby distinguishing between normal seasonal fluctuations and abnormal weather events; A multi-level triggering mechanism is constructed to divide environmental parameter fluctuations into three levels: attention, warning, and emergency, and a multi-parameter combination triggering condition is considered to achieve intelligent response to complex environmental changes; A lightweight computing module is deployed at the field monitoring node to perform real-time data preprocessing and preliminary analysis, and to adjust the local sampling strategy immediately when an abnormal environmental condition is detected, while starting a data caching mechanism to ensure that critical data can still be saved in the case of communication interruption, and establishing a clear task division and data synchronization mechanism with the central data processing center; When a specific environmental event is detected, it automatically enters a high-frequency sampling mode, activates additional environmental monitoring units, records comprehensive data during the entire environmental event, and performs a special evaluation of concrete strength after the event, and builds an environmental event-strength change causality database.

6. The method for remote monitoring of concrete strength in complex environments of claim 1, wherein, Step 6 includes the following steps: A multi-source heterogeneous data fusion framework is constructed to extract abnormal feature patterns from various types of sensor data and build a concrete strength evolution feature library; Rolling prediction of strength values for the next 7 days, 30 days, and 90 days is achieved; According to the importance level of the structure and the characteristics of the service environment, a regional and graded strength safety threshold system is established, and the warning threshold parameters are automatically calibrated in combination with seasonal environmental change characteristics, and the safety margin is adjusted in time according to the structure aging rate; Different levels of warning responses are triggered according to the degree of strength decline: when the strength is less than 90% of the design value, a first-level prompt warning is triggered to remind potential risks; when the strength is less than 85%, a second-level intervention warning is triggered to suggest preventive measures; when the strength is less than 80% or the strength decline rate exceeds 0.5 MPa / day, a third-level emergency warning is triggered, requiring immediate intervention measures to prevent further deterioration; Based on the finite element mesh division principle, a three-dimensional distribution cloud map of structural strength is generated, and a structure health assessment report containing strength numerical analysis, environmental impact assessment, safety risk classification, and maintenance recommendations is automatically generated; Combining the strength development trend and environmental sensitivity analysis, maintenance decision recommendations are automatically generated, including detection frequency adjustment, protection measure strengthening, and necessary reinforcement schemes to ensure the long-term stability and safety of concrete structures in complex environments.

7. The method for remote monitoring of concrete strength in complex environments according to claim 6, wherein, The generation method of the three-dimensional distribution cloud map is: Based on tetrahedral mesh partitioning technology, 5000-20000 grid cells are adaptively generated according to the complexity of the structure; Each grid cell is associated with data from 2-3 nearby sensor nodes; Linear interpolation method is used to calculate the strength value within the grid cell; Laplace smoothing algorithm is used to consider spatial correlation and smooth the transition between adjacent cells; A three-color coloring scheme is used, with red indicating dangerous areas where the strength is less than 80% of the design value, yellow indicating warning areas where the strength is between 80% and 90% of the design value, and green indicating safe areas where the strength is greater than 90% of the design value; A simplified three-dimensional rendering algorithm is used to achieve visual effects, ensuring smooth display on ordinary mobile terminals.

8. A system for remote monitoring of concrete strength in complex environments for implementing a method for remote monitoring of concrete strength in complex environments according to any one of claims 1 to 7, characterized in that, The following modules are included: A multi-level intelligent sensing monitoring module that monitors the chemical environment, acoustic properties, and strain distribution within the concrete through a three-level sensor network: microscopic, mesoscopic, and macroscopic. An environmental parameter acquisition and communication module is responsible for obtaining environmental parameters around the concrete structure and transmitting data to the processing center through low-power Internet of Things technology. A data processing module is used to realize the unified mathematical expression of the coupling relationship between environmental factors, material properties, time evolution, structure size and performance indicators. A time series strength evolution prediction module realizes the full spectrum prediction of concrete strength from short-term fluctuations to long-term evolution under different environmental conditions, accurately capturing nonlinear time-varying characteristics. A strength development rate evaluation module is used to quantify the influence of different environmental conditions on the development rate of concrete strength and accurately evaluate the age effect. An adaptive monitoring control module intelligently adjusts the sampling frequency and tracks environmental events based on strength development rate derivative analysis and environmental sensitivity scoring, achieving optimal allocation of monitoring resources and accurate capture of key environmental changes. A structure safety state evaluation module integrates multi-source heterogeneous data for strength rolling prediction, establishes a hierarchical safety threshold system and a multi-level early warning response mechanism, and generates a three-dimensional strength distribution cloud map and a health assessment report.

Images