Dynamic detection and analysis method and system for early performance of ultra-high performance concrete

Abstract

The invention relates to the technical field of concrete detection, in particular to a dynamic detection and analysis method and system for early performance of ultra-high performance concrete. According to the technical scheme, the method comprises the steps of establishing a raw material characteristic spectrum library based on multi-physics field coupling, constructing a dynamic detection array based on a time sequence, developing a multi-modal data fusion algorithm, establishing a nonlinear coupling effect decomposition model and implementing dynamic parameter inversion. According to the method, a technical chain of four-dimensional monitoring network-multi-physics field coupling analysis-cross-scale prediction-adaptive regulation and control is constructed, so that dynamic accurate detection and intelligent regulation and control of the early performance of the ultra-high performance concrete are realized, and the technical bottlenecks of a traditional method in the aspects of parameter coupling, cross-scale association and real-time regulation and control are solved; the early performance prediction precision and the maintenance efficiency are remarkably improved, the cracking risk is reduced, and reliable technical support is provided for application of the ultra-high performance concrete in major engineering.

Description

Technical Field

[0001] The present invention relates to the technical field of concrete detection, and particularly relates to a method and system for dynamically detecting and analyzing the early performance of ultra-high performance concrete. Background Art

[0002] Ultra-High Performance Concrete (UHPC), as a new type of building material, has been increasingly widely used in the field of modern construction engineering due to its excellent properties such as high strength, high durability, and high toughness. With the continuous improvement of the requirements for engineering quality and construction efficiency in the construction industry, accurately grasping the early performance of UHPC is of crucial significance for ensuring engineering quality and optimizing construction technology. In the detection of concrete performance, there are multi-parameter coupling problems:

[0003] Complex material composition: Ultra-high performance concrete is usually composed of multiple raw materials, such as cement, silica fume, fly ash, fibers, admixtures, etc. The interactions between these components are complex. Different proportions of these raw materials will have different effects on the early performance of concrete, and there is a coupling effect between various factors. For example, the incorporation of fibers can improve the toughness of concrete, but at the same time, it may affect the fluidity and early hydration reaction of concrete, making it difficult to separately isolate the influence of a certain factor on the early performance of concrete and bringing difficulties to detection and analysis.

[0004] Interrelated properties: The early strength, shrinkage, impermeability, and other properties of ultra-high performance concrete are interrelated and interact with each other. For example, the increase in early strength may be accompanied by an increase in shrinkage, and shrinkage may in turn affect the impermeability of concrete. When conducting detection and analysis, it is very difficult to study these properties completely independently, and it is necessary to comprehensively consider the relationships between multiple performance indicators, which increases the complexity and difficulty of detection and analysis.

[0005] In summary, the present application proposes a method and system for dynamically detecting and analyzing the early performance of ultra-high performance concrete. Summary of the Invention

[0006] The object of the present invention is to address the problem in the background art that the multi-parameter coupling in concrete performance detection increases the complexity and difficulty of detection and analysis, and to propose a method and system for dynamically detecting and analyzing the early performance of ultra-high performance concrete.

[0007] On the one hand, the present application provides a method for dynamically detecting and analyzing the early performance of ultra-high performance concrete, including the following steps:

[0008] Establish a raw material characteristic spectrum library based on multi-physical field coupling, including three-dimensional material fingerprint maps of cement matrix phase, mineral admixture phase, and fiber reinforcement phase;

[0009] Construct a dynamic detection array based on time series. Embed a distributed optical fiber sensor array inside the concrete specimen, and arrange a multi-spectral imaging device and an acoustic emission sensor array outside to form a four-dimensional monitoring network;

[0010] Develop a multi-modal data fusion algorithm to spatially and temporally align electrical impedance tomography data, hydration heat distribution data, and dielectric property data, and extract hidden layer coupling features through a variational autoencoder;

[0011] Establish a non-linear coupling effect decomposition model, and use tensor decomposition technology to separate the coupling effects of fiber orientation distribution, interfacial transition zone development, and hydration reaction process on early performance;

[0012] Implement dynamic parameter inversion, and synchronously solve the hydration kinetics parameters, shrinkage stress field distribution, and microscopic pore evolution equation by combining an improved particle swarm optimization algorithm;

[0013] Construct a cross-scale performance prediction model, and establish a mapping relationship from nano-scale C-S-H gel growth to macroscopic compressive strength development through a long short-term memory network;

[0014] Design an adaptive regulation strategy, dynamically adjust the curing environment parameters based on real-time performance deviation analysis, and establish a multi-objective optimization control model for temperature-humidity-constraint conditions;

[0015] Deploy an edge computing node network, construct a distributed computing architecture at the pouring site, and realize local feature extraction of monitoring data and incremental update of cloud models;

[0016] Establish a full-life cycle data chain, deeply associate the early performance detection data with the structural health monitoring data during the service period, and form a performance evolution digital twin.

[0017] Optionally, the three-dimensional material fingerprint includes:

[0018] The area ratio of X-ray diffraction characteristic peaks of the cement matrix phase, the gray scale distribution of backscattered electron images, and the nano-indentation modulus distribution;

[0019] The atomic force microscope surface roughness parameters of the mineral admixture phase, the Raman spectrum characteristic peak displacement, and the zeta potential-particle size distribution curve;

[0020] The micro-CT three-dimensional reconstruction parameters of the fiber reinforcement phase, the single fiber tensile stress relaxation curve, and the interfacial bond-slip constitutive relationship.

[0021] Optionally, the four-dimensional monitoring network includes:

[0022] The Bragg grating array of the embedded optical fiber sensor to monitor the internal strain and temperature field evolution with a spatial resolution of 2 mm;

[0023] The multispectral imaging device acquires the dynamics of surface moisture evaporation and microcrack propagation in the 400 - 2500 nm band at a frequency of 10 Hz;

[0024] The acoustic emission sensor array uses waveform manifold learning technology to identify fiber pull - out, matrix cracking, and interface debonding events.

[0025] Optionally, the multimodal data fusion algorithm includes:

[0026] Establish a spatio - temporal registration coordinate system to perform voxel - level registration on the conductivity distribution data of electrical impedance tomography and the hydration heat infrared thermography data;

[0027] Develop a feature - level fusion module, and use a bidirectional gated recurrent unit weighted by an attention mechanism to process multi - source heterogeneous time - series data;

[0028] Construct a physical - constraint loss function, and embed the mass conservation equation and the energy balance equation into the neural network training process.

[0029] Optionally, the dynamic parameter inversion includes:

[0030] Establish a parameter sensitivity ranking model, and determine the priority of key inversion parameters based on the Morris global sensitivity analysis method;

[0031] Design a hybrid particle swarm optimization strategy, introduce the taboo list mechanism of taboo search into the particle update process to prevent falling into local optima;

[0032] Construct a multi - objective fitness function to synchronously optimize the hydration degree prediction error, the root - mean - square error of shrinkage strain, and the pore fractal dimension matching degree.

[0033] Optionally, the cross - scale performance prediction model includes:

[0034] Nano - scale module: A C - S - H gel growth kinetics model based on molecular dynamics simulation;

[0035] Micro - scale module: A meso - mechanical constitutive model considering the characteristics of the interfacial transition zone;

[0036] Macro - scale module: A three - dimensional damage evolution model integrating digital image correlation technology.

[0037] Optionally, the adaptive regulation strategy includes:

[0038] Establish an environmental response surface model, and determine the influence law of the temperature - humidity interaction on performance development through central composite design experiments;

[0039] Develop a fuzzy predictive control algorithm to dynamically adjust the steam curing parameters and external constraint conditions based on rolling - horizon optimization;

[0040] Build a security boundary warning system to calculate the deviation degree between the current maintenance path and the ideal performance development trajectory in real time.

[0041] Optionally, the edge computing node network includes:

[0042] Local feature extraction unit: Realize real-time wavelet packet decomposition of sensor data based on the parallel computing architecture of FPGA;

[0043] Fog computing layer: Deploy a lightweight convolutional neural network for key feature pattern recognition;

[0044] Cloud collaborative platform: Adopt a federated learning framework to realize privacy-preserving model updates for multi-project data.

[0045] Optionally, the full life cycle data chain construction method includes:

[0046] Establish a spatio-temporal coding rule to perform hash coding on the timestamp of early detection data and the structural space coordinates;

[0047] Develop a knowledge graph engine to construct a multi-dimensional relationship network of material composition - process parameters - early performance - long-term durability;

[0048] Design a transfer learning interface to transfer the laboratory-scale detection model to the digital twin of the actual engineering structure.

[0049] In a second aspect, the present application provides a dynamic detection and analysis system for the early performance of ultra-high performance concrete, including:

[0050] Multi-physical field coupling monitoring module, including a distributed fiber optic sensor array, a multi-spectral imager, an acoustic emission sensor, and an electrical impedance tomography device;

[0051] Edge intelligent processing terminal, equipped with a programmable logic device and an embedded neural network acceleration chip;

[0052] Cloud collaborative analysis platform, deployed with a coupling effect decomposition model library and a performance prediction digital twin;

[0053] Dynamic regulation execution mechanism, including an intelligent temperature control maintenance cover, an adaptive restraint device, and a microenvironment regulation system;

[0054] Data security interaction gateway, using blockchain technology to achieve tamper-proof storage and trusted sharing of detection data;

[0055] Among them, each module realizes data interaction through the industrial Internet of Things protocol, forming an "end-edge-cloud" collaborative architecture.

[0056] Compared with the prior art, the present application includes at least one of the following beneficial technical effects:

[0057] The present invention establishes a four-dimensional monitoring network system, breaks through the limitations of traditional single-point detection, and realizes multi-scale and full-process dynamic tracking from the nanoscale to the macroscale. At the same time, a coupling effect decomposition model is proposed, successfully solving the problem of the interactive influence of multiple parameters, and for the first time realizing the quantitative separation of the contribution of each constituent material to the early performance.

[0058] Furthermore, through the developed adaptive regulation system, the curing conditions are made to match the development of the material properties in real time, reducing the risk of early cracking, increasing the strength development rate, and building a full-life cycle data chain to break through the barriers between laboratory research and engineering applications, reducing the structural maintenance cost during the service period.

[0059] Even further, through the edge-cloud collaborative architecture, the data processing efficiency is greatly improved, meeting the real-time requirements of the construction site while ensuring data security.

[0060] By constructing a technology chain of "four-dimensional monitoring network - multi-physical field coupling analysis - cross-scale prediction - adaptive regulation", the present invention realizes the dynamic and accurate detection and intelligent regulation of the early performance of ultra-high performance concrete, solves the technical bottlenecks in parameter coupling, cross-scale correlation and real-time regulation of traditional methods, significantly improves the prediction accuracy of early performance and the curing efficiency, reduces the cracking risk, and provides reliable technical support for the application of ultra-high performance concrete in major projects. Description of the Drawings

[0061] Figure 1 It is a flowchart of a method for dynamically detecting and analyzing the early performance of ultra-high performance concrete;

[0062] Figure 2 It is a principle block diagram of a system for dynamically detecting and analyzing the early performance of ultra-high performance concrete. Detailed Embodiments

[0063] The technical solutions of the present invention will be further described below in conjunction with the drawings and specific embodiments.

[0064] Embodiment 1

[0065] As Figure 1 shown, a method for dynamically detecting and analyzing the early performance of ultra-high performance concrete proposed by the present invention is described in detail for each step process below.

[0066] I. Establish a raw material characteristic spectrum library based on multi-physical field coupling, including three-dimensional material fingerprint maps of the cement matrix phase, mineral admixture phase, and fiber reinforcement phase. The three-dimensional material fingerprint maps include:

[0067] The area ratio of X-ray diffraction characteristic peaks, the gray-scale distribution of backscattered electron images, and the nano-indentation modulus distribution of the cement matrix phase;

[0068] Backscattered electron images were obtained using a field emission scanning electron microscope (FE-SEM, Hitachi SU8230). The gray distribution histogram was analyzed by ImageJ software to quantify the proportion of unhydrated particles. A nanoindentation tester (Hysitron TI950) was used to perform grid indentation tests (50×50 dot matrix, load 2 mN), and a probability density function of elastic modulus was established:

[0069]

[0070] where E i is the single-point modulus value, σ E is the distribution standard deviation, and N is the total number of measurement points;

[0071] Atomic force microscope surface roughness parameters, Raman spectroscopy characteristic peak displacement amounts, and zeta potential - particle size distribution curves of mineral admixture phases;

[0072] The surface of fly ash was scanned by an atomic force microscope (AFM, Bruker Dimension Icon), and three-dimensional topography parameters (Sa = 0.78 μm, Sq = 1.02 μm) were extracted. Sa: Arithmetic mean height of surface roughness; Sq: Root mean square height of surface roughness.

[0073] A laser particle size analyzer (Malvern Mastersizer 3000) was used to measure the zeta potential - particle size distribution curve, and an association equation between D50 and early hydration rate was established:

[0074]

[0075] where k 1 is a material constant, D 50 : Median particle size of mineral admixture; Zeta potential of mineral admixture, characterizing the surface charge characteristics of particles;

[0076] Micro-CT three-dimensional reconstruction parameters, single-filament tensile stress relaxation curves, and interfacial bond-slip constitutive relations of fiber-reinforced phases.

[0077] The fiber distribution was scanned by micro-CT (ZEISS Xradia 620), and the orientation tensor was calculated:

[0078]

[0079] where, is the direction cosine of the p-th fiber, and N is the total number of fibers;

[0080] The stress relaxation curve was obtained by a single-filament tensile test (Instron 5967), and the Prony series model was fitted:

[0081]

[0082] Among them, σ(t) is the stress value at time t, and σ 0 is the initial stress, and g i is the weight coefficient of the i-th relaxation term, and τ i is the time constant of the i-th relaxation term.

[0083] In this embodiment, through three-dimensional material fingerprint maps (X-ray diffraction, nanoindentation, micro-CT), the characteristics of the cement matrix, mineral admixtures, and fiber-reinforced phases are comprehensively quantified to provide high-precision input data. The performance fluctuations caused by raw material batch differences are reduced, and the reliability of concrete mix design is improved. Traditional methods rely on empirical mix ratios and cannot accurately quantify the influence of raw material characteristics on performance.

[0084] Second, construct a dynamic detection array based on time series. Buried distributed fiber optic sensor arrays inside the concrete specimens, and multi-spectral imaging devices and acoustic emission sensor arrays are arranged outside to form a four-dimensional monitoring network; the four-dimensional monitoring network includes:

[0085] The Bragg grating array of the embedded fiber optic sensor monitors the internal strain and temperature field evolution with a spatial resolution of 2 mm; a 4×4 fiber Bragg grating array (FBG, wavelength 1528 - 1562 nm) is pre-embedded with a spatial resolution of 2 mm and a strain measurement accuracy of ±1 με;

[0086] The multi-spectral imaging device collects the dynamics of surface moisture evaporation and microcrack propagation at a frequency of 10 Hz in the 400 - 2500 nm band;

[0087] The multi-spectral camera (Specim IQ) collects images at a rate of 10 fps in the 400 - 1000 nm band, and the evaporation rate is inverted through the moisture index (NDWI):

[0088]

[0089] Among them, ρ 860 is the reflectivity at the 860 nm band, and ρ 1240 is the reflectivity at the 1240 nm band.

[0090] The infrared thermal imager (FLIR A655sc) monitors the surface temperature field with a spatial resolution of 0.5 °C;

[0091] The acoustic emission sensor array uses waveform manifold learning technology to identify fiber pull-out, matrix cracking, and interface debonding events. Waveform manifold learning (t-SNE) is used to identify the damage types:

[0092] Fiber pull-out: frequency peak 80 - 120 kHz, duration > 200 μs;

[0093] Matrix cracking: frequency peak 150 - 250 kHz, rise time < 50 μs. Achieve full spatio - temporal coverage through a four - dimensional monitoring network (internal fiber optic sensing + external multi - spectral imaging + acoustic emission monitoring) to capture the dynamic details of early performance evolution. Real - time monitor the internal strain, temperature field and surface moisture evaporation of concrete to provide data support for early performance regulation.

[0094] III. Develop a multi - modal data fusion algorithm to perform spatio - temporal alignment on electrical impedance tomography data, hydration heat distribution data and dielectric property data, and extract hidden - layer coupling features through a variational auto - encoder; the multi - modal data fusion algorithm includes:

[0095] Establish a spatio - temporal registration coordinate system to perform voxel - level registration on the conductivity distribution data of electrical impedance tomography and the hydration heat infrared thermal image data; establish a voxel coordinate system (1mm 3 resolution), align the ERT and infrared data through the ICP algorithm, and the registration error < 0.3 mm.

[0096] Develop a feature - level fusion module, and use a bidirectional gated recurrent unit weighted by an attention mechanism to process multi - source heterogeneous time - series data;

[0097] class FusionGRU(nn.Module):

[0098] def __init__(self):

[0099] super().__init__()

[0100] self.attention = MultiHeadAttention(embed_dim = 128, num_heads = 4)

[0101] self.gru = BidirectionalGRU(input_size = 256, hidden_size = 128)

[0102] self.phys_constraint = PhysicsLoss(alpha = 0.5, beta = 0.3)

[0103] def forward(self, x1, x2):

[0104] attn_out = self.attention(x1, x2)

[0105] gru_out = self.gru(attn_out)

[0106] loss = self.phys_constraint(gru_out)

[0107] return gru_out, loss

[0108] Construct a physical constraint loss function, and embed the mass conservation equation and the energy balance equation into the neural network training process. Introduce Fick's second law in the loss function to constrain the moisture diffusion:

[0109]

[0110] where C is the moisture concentration, D is the diffusion coefficient, t is the time, and λ 1 is the weight coefficient of the physical constraint term, the Laplace operator, representing the second-order spatial derivative. Through a deep learning model with physical constraints (such as variational autoencoder), achieve precise alignment and fusion of electrical impedance tomography, hydration heat distribution, and dielectric property data, improve the utilization rate of multi-source heterogeneous data, and enhance the reliability of early performance prediction.

[0111] IV. Establish a non-linear coupling effect decomposition model, and use tensor decomposition technology to separate the coupling effects of fiber orientation distribution, interfacial transition zone development, and hydration reaction process on the early performance; Coupling effect decomposition (1) Tensor construction:

[0112] Construct a third-order tensor:

[0113]

[0114] where:

[0115] m = 6 (cement content, water-binder ratio, fiber content, curing temperature, humidity, degree of restraint);

[0116] n = 5 (compressive strength, shrinkage strain, permeability coefficient, porosity, elastic modulus);

[0117] p = 4 (1h, 6h, 24h, 72h);

[0118] (2) Tucker decomposition:

[0119] T = G × 1 U × 2 V × 3 W

[0120] where,

[0121] T: The third-order tensor represents the multi-parameter coupling effect;

[0122] G: The core tensor reveals the interaction pattern between parameters;

[0123] U, V, W: Factor matrices, representing the contribution degrees of material parameters, performance indicators, and time dimension respectively;

[0124] × i : Represents the product of the tensor in the i-th dimension.

[0125] Reveal the parameter interaction pattern through the core tensor G, and the factor matrix U characterizes the contribution degree of material parameters. Adopt tensor decomposition technology to separate the coupling effects of fiber distribution, interfacial transition zone development, and hydration reaction, and quantify the contribution degree of each factor to the performance. Provide a theoretical basis for material design and optimization, and guide the adjustment of mix ratio and process improvement.

[0126] V. Implement dynamic parameter inversion, and synchronously solve the hydration kinetic parameters, shrinkage stress field distribution, and microscopic pore evolution equation by combining with the improved particle swarm optimization algorithm; the dynamic parameter inversion includes:

[0127] Establish a parameter sensitivity ranking model, and determine the priority of key inversion parameters based on the Morris global sensitivity analysis method;

[0128] Design a hybrid particle swarm optimization strategy, introduce the taboo list mechanism of taboo search into the particle update process to prevent falling into local optimum; (1) Hybrid optimization algorithm:

[0129] Introduce the taboo search mechanism into the particle update formula:

[0130]

[0131] Among them, The velocity of the i-th particle at the (k + 1)-th iteration; w: Inertia weight; c 1 , c 2 : Learning factors; r 1 , r 2 : Random numbers, range [0, 1]; pbest i : The historical optimal solution of the i-th particle; gbest: Global optimal solution; Is the position of the i-th particle at the k-th iteration; η is the weight coefficient of the taboo search term, and TS() is the taboo term to avoid repeated search of historical inferior solutions.

[0132] Construct a multi-objective fitness function to synchronously optimize the prediction error of hydration degree, the root mean square error of shrinkage strain, and the matching degree of pore fractal dimension. Multi-objective fitness function:

[0133] Fitness = ω 1 ·RMSE hyd + ω 2 ·|Δε sh | + ω 3 ·(1 - FD match)

[0134] Among them, the weight coefficient ω i is dynamically adjusted by the entropy weight method. Fitness is the fitness value, which is used to evaluate the quality of the inversion result; ω 1 , ω 2 , ω 3 is the weight coefficient, which is dynamically adjusted by the entropy weight method; RMSE hyd : Root mean square error of hydration degree prediction; Δε sh : Prediction deviation of shrinkage strain; FD match is the matching degree of pore fractal dimension. Combining with the improved particle swarm optimization algorithm, the hydration kinetic parameters, shrinkage stress field and microscopic pore evolution equation are solved synchronously to achieve accurate inversion of complex properties. Real-time prediction of the early performance development of concrete provides a scientific basis for curing regulation.

[0135] VI. Construct a cross-scale performance prediction model, and establish a mapping relationship from the growth of nano-scale C-S-H gel to the development of macroscopic compressive strength through a long short-term memory network; the cross-scale performance prediction model includes:

[0136] Nano-scale module: A kinetic model for the growth of C-S-H gel based on molecular dynamics simulation;

[0137] Micro-scale module: A meso-mechanical constitutive model considering the characteristics of the interfacial transition zone;

[0138] Macro-scale module: A three-dimensional damage evolution model integrating digital image correlation technology. Establish a mapping relationship from the growth of nano-scale C-S-H gel to the development of macroscopic compressive strength through a long short-term memory network (LSTM) to achieve seamless connection of multi-scale performance evolution. Accurately predict the early strength development of concrete and guide the construction progress arrangement.

[0139] VII. Design an adaptive regulation strategy, dynamically adjust the curing environment parameters based on real-time performance deviation analysis, and establish a multi-objective optimization control model for temperature-humidity-constraint conditions; the adaptive regulation strategy includes:

[0140] Establish an environmental response surface model, and determine the influence law of the temperature-humidity interaction on performance development through central composite design experiments;

[0141] Develop a fuzzy predictive control algorithm, and dynamically adjust the steam curing parameters and external constraint conditions based on rolling horizon optimization;

[0142] Construct a safety boundary warning system to calculate the deviation degree between the current curing path and the ideal performance development trajectory in real time.

[0143] Based on real-time performance deviation analysis, dynamically adjust the curing environment parameters (temperature, humidity, restraint conditions) to achieve real-time matching of curing conditions and material property development, significantly reduce the risk of early cracking, and improve the durability of concrete.

[0144] VIII. Deploy an edge computing node network to build a distributed computing architecture at the pouring site to achieve local feature extraction of monitoring data and incremental update of the cloud model; the edge computing node network includes:

[0145] Local feature extraction unit: Implement real-time wavelet packet decomposition of sensor data based on the parallel computing architecture of FPGA;

[0146] Fog computing layer: Deploy a lightweight convolutional neural network for key feature pattern recognition;

[0147] Cloud collaborative platform: Use the federated learning framework to achieve privacy-preserving model updates for multi-project data. Build an "edge-cloud" collaborative architecture at the construction site to achieve local processing of monitoring data and incremental update of the cloud model, improve data processing efficiency, meet real-time requirements, and ensure data security and privacy.

[0148] IX. Establish a full-life cycle data chain to deeply associate the early performance detection data with the structural health monitoring data during the service period to form a performance evolution digital twin. The method for constructing the full-life cycle data chain includes:

[0149] Establish a spatio-temporal coding rule to perform hash coding on the timestamp of the early detection data and the structural space coordinates;

[0150] Develop a knowledge graph engine to construct a multi-dimensional relationship network of material composition - process parameters - early performance - long-term durability;

[0151] Design a transfer learning interface to transfer the detection model at the laboratory scale to the digital twin of the actual engineering structure.

[0152] In this embodiment, deeply associate the early performance detection data with the structural health monitoring data during the service period to form a performance evolution digital twin. Provide a scientific basis for structural health monitoring and maintenance, and extend the service life of the project.

[0153] Embodiment 2

[0154] As Figure 2 shown, this embodiment provides a dynamic detection and analysis system for the early performance of ultra-high performance concrete, including:

[0155] Multi-physical field coupling monitoring module, including a distributed optical fiber sensor array, a multi-spectral imager, an acoustic emission sensor, and an electrical impedance tomography device;

[0156] Edge intelligent processing terminal, equipped with a programmable logic device and an embedded neural network acceleration chip;

[0157] Cloud collaborative analysis platform, deployed with a coupling effect decomposition model library and a performance prediction digital twin;

[0158] Dynamic regulation actuator, including an intelligent temperature control curing cover, an adaptive restraint device, and a microenvironment regulation system;

[0159] Data security interaction gateway, using blockchain technology to achieve tamper-proof storage and trusted sharing of detection data;

[0160] Among them, each module realizes data interaction through the industrial Internet of Things protocol, forming an "end-edge-cloud" collaborative architecture.

[0161] This embodiment provides a dynamic detection and analysis system for the early performance of ultra-high performance concrete (UHPC). Through the collaborative work of multiple modules, it realizes the full-process intelligent management from data collection, processing, analysis to regulation. The following is a detailed expansion description of the system:

[0162] 1. Multi-physical field coupling monitoring module

[0163] Distributed fiber optic sensor array:

[0164] Adopting Fiber Bragg Grating (FBG) technology, it monitors the evolution of internal strain and temperature field of concrete with a spatial resolution of 2mm; supports multi-channel synchronous acquisition, with a sampling frequency of 100Hz, a measurement accuracy of ±1με, and a temperature resolution of 0.1°C; built-in self-diagnosis function, which can detect the working state of the sensor in real time to ensure data reliability.

[0165] Multispectral imager: covering the 400 - 2500nm band, it acquires the dynamics of surface moisture evaporation and microcrack propagation at a frequency of 10Hz; equipped with a high-precision optical lens, with a spatial resolution of 0.1mm, and supports the calculation of multi-spectral indices (such as NDWI); built-in ambient light compensation algorithm, which can adapt to the complex lighting conditions of the construction site.

[0166] Acoustic emission sensor array: adopting broadband sensors (frequency range 20kHz - 1MHz), it captures microscopic damage signals at a sampling rate of 1MHz; equipped with a waveform manifold learning algorithm, which can automatically identify events such as fiber pull-out, matrix cracking, and interface debonding; supports three-dimensional sound source localization, with a localization accuracy of ±2mm.

[0167] Electrical impedance tomography device: adopting a 16-electrode array, it acquires the conductivity distribution inside the concrete at a frequency of 10Hz; equipped with an adaptive regularization algorithm to improve the imaging resolution and anti-noise ability; supports multi-parameter fusion to realize the synchronous monitoring of moisture distribution and pore structure.

[0168] 2. Edge intelligent processing terminal

[0169] Hardware configuration: Equipped with FPGA (Field Programmable Gate Array) and embedded neural network acceleration chips (such as NVIDIA Jetson Nano); supports parallel processing of multi-channel sensor data, with a computing power of 4 TOPS (trillion operations per second); built-in large-capacity storage module (maximum support of 1TB) to meet the long-term data caching requirements.

[0170] Software functions: Real-time wavelet packet decomposition: Denoises and extracts features from sensor data;

[0171] Lightweight convolutional neural network: Achieves key feature pattern recognition (such as crack type classification);

[0172] Local data compression: Adopts lossless compression algorithm to reduce data transmission bandwidth requirements.

[0173] 3. Cloud collaborative analysis platform

[0174] Coupling effect decomposition model library: Contains multi-parameter interaction analysis models based on tensor decomposition; supports online model training and updating to adapt to different engineering scenario requirements.

[0175] Performance prediction digital twin: Builds a cross-scale performance prediction model based on Long Short-Term Memory Network (LSTM); supports visualization of real-time performance evolution and provides prediction results of key indicators such as early strength and shrinkage strain.

[0176] Data management and analysis: Uses a distributed database (such as MongoDB) to store massive monitoring data; provides multi-dimensional data analysis tools (such as clustering analysis, trend prediction), and supports user-defined report generation.

[0177] 4. Dynamic regulation actuator

[0178] Intelligent temperature control curing cover: Adopts PID control algorithm, with a temperature control accuracy of ±0.5°C; supports multi-zone independent temperature control to adapt to the curing requirements of mass concrete.

[0179] Adaptive constraint device: Equipped with a hydraulic servo system to achieve precise adjustment of the constraint force (adjustment range 0 - 10 MPa); built-in stress feedback mechanism to adjust the constraint conditions in real time to prevent early cracking.

[0180] Microenvironment regulation system: Integrates humidification, ventilation, and heating functions, with a humidity control range of 30% - 95% and an accuracy of ±3%; supports remote control to adjust the curing parameters in real time through a mobile terminal.

[0181] 5. Data security interaction gateway

[0182] Application of blockchain technology: An alliance chain architecture is adopted to achieve tamper-proof storage of detection data; data traceability is supported to ensure the credibility and transparency of monitoring data.

[0183] Secure communication protocol: Industrial Internet of Things protocols (such as MQTT, OPC UA) are adopted to ensure the security of data transmission; data encryption and identity authentication are supported to prevent unauthorized access.

[0184] 6. "Edge-Cloud" collaborative architecture

[0185] Edge side (device layer): Responsible for data collection and local processing, reducing the computing load on the cloud;

[0186] Supports the function of resuming data transmission after network disconnection to ensure data integrity.

[0187] Edge side (edge computing layer): Realizes data feature extraction and preliminary analysis, improving the system response speed; supports multi-terminal collaborative work to meet the needs of large-scale construction sites.

[0188] Cloud side (cloud platform): Provides high-performance computing resources, supports complex model training and optimization; supports multi-project data sharing and collaborative analysis to improve the overall efficiency of the system.

[0189] 7. System working process

[0190] Data collection: The multi-physical field coupling monitoring module collects the internal strain, temperature, moisture distribution and acoustic emission signals of concrete in real time;

[0191] Edge processing: The edge intelligent processing terminal performs noise reduction, feature extraction and preliminary analysis on the data;

[0192] Cloud analysis: The cloud collaborative analysis platform decomposes the coupling effect and predicts the performance, generating control instructions;

[0193] Dynamic regulation: The dynamic regulation actuator adjusts the curing conditions according to the instructions to achieve precise control of early performance;

[0194] Data storage and sharing: The data security interaction gateway stores the monitoring data in the blockchain to ensure secure and trusted sharing of data.

[0195] This system realizes the dynamic and precise detection and intelligent regulation of the early performance of UHPC through the collaborative work of multiple modules, solves the technical bottlenecks in parameter coupling, cross-scale correlation and real-time regulation of traditional methods, significantly improves the prediction accuracy of early performance and the curing efficiency, reduces the cracking risk, and provides reliable technical support for the application of ultra-high performance concrete in major projects.

[0196] Example 3

[0197] Based on Embodiment 2, this embodiment of the early performance dynamic detection and analysis system further includes an intelligent diagnosis and early warning module, a system optimization and expansion module, and a system workflow optimization module.

[0198] The intelligent diagnosis and early warning module includes:

[0199] Real-time analysis of monitoring data to identify abnormal performance development trends (such as early cracking, lagging strength development); providing graded early warnings (mild, moderate, severe) and targeted control suggestions. Using deep reinforcement learning algorithms (such as DQN) to build an anomaly detection model; integrating an expert knowledge base, combining historical data with real-time monitoring data to improve the accuracy of diagnosis; supporting multi-dimensional early warning indicators (such as strain rate, temperature gradient, acoustic emission energy). When abnormal strain rate in a local area is detected, automatically trigger an early warning and adjust the curing conditions, identify areas with uneven fiber distribution, and guide the improvement of subsequent construction processes.

[0200] Digital twin enhancement function. Based on Embodiment 2, enhance the real-time interaction and simulation capabilities of the digital twin; support virtual reality (VR) and augmented reality (AR) visualization to provide an immersive monitoring experience. Use the Unity 3D engine to build a three-dimensional visualization model, and map the internal state of concrete in real time; integrate a physics engine (such as PhysX) to simulate the performance evolution process under different curing conditions; support multi-user collaborative operations to achieve remote consultation and decision-making support. View the internal strain field distribution of concrete through VR devices to identify potential cracking risk areas; use AR technology to overlay virtual data on-site to guide curing operations.

[0201] Multi-objective optimization control strategy. Based on Embodiment 2, introduce a multi-objective optimization algorithm to achieve intelligent control of curing conditions; comprehensively consider multiple objectives such as strength development, shrinkage control, and energy consumption optimization, and use NSGA-II (Non-dominated Sorting Genetic Algorithm) to solve the multi-objective optimization problem;

[0202] Construct the objective function:

[0203] Fitness = ω 1 ·f 1 (strength)+ω 2 ·f 2 (shrinkage)+ω 3 ·f 3 (energy)

[0204] Where f 1 、f 2 、f 3 are the objective functions of strength, shrinkage, and energy consumption respectively, and ω iis the weight coefficient. It supports dynamic weight adjustment to adapt to different engineering requirements. On the premise of ensuring strength development, it optimizes curing energy consumption, reduces construction costs, balances shrinkage control and strength development, and reduces the risk of early cracking.

[0205] In this embodiment, the system optimization and expansion module specifically includes:

[0206] Expansion of the multi-physical field coupling monitoring module, adding new sensor types: ultrasonic sensor: monitoring the evolution of the internal pore structure of concrete, frequency range 50 kHz - 1 MHz; microwave humidity sensor: real-time measurement of the internal moisture distribution of concrete, accuracy ±0.5%.

[0207] Data fusion optimization: Introduce the Kalman filter algorithm to improve the accuracy of multi-source data fusion; develop an adaptive sampling strategy to dynamically adjust the sampling frequency according to performance evolution.

[0208] The upgrade of the edge intelligent processing terminal specifically includes:

[0209] Hardware upgrade: Adopt a new generation of AI chips (such as Huawei Ascend 910), and the computing power is increased to 16 TOPS;

[0210] Add a 5G communication module to support high-speed data transmission and low-latency control.

[0211] Software function expansion: Add an anomaly detection algorithm to support local real-time early warning; integrate a data encryption module to ensure the security of edge-side data.

[0212] Enhancement of the cloud collaborative analysis platform, supporting multi-project data comparison and analysis to identify common laws and individual differences; providing an intelligent report generation tool to automatically generate inspection reports and control suggestions.

[0213] Performance optimization: Adopt a distributed computing framework (such as Spark) to improve the efficiency of large-scale data processing; introduce model compression technology to reduce the consumption of cloud computing resources.

[0214] It should be noted that the system workflow optimization module in this embodiment specifically includes the following aspects:

[0215] Data collection and preprocessing: Add ultrasonic and microwave humidity sensor data to enrich the monitoring dimension; adopt the Kalman filter algorithm to improve data quality.

[0216] Edge intelligent processing: Add an anomaly detection function to achieve local real-time early warning; support 5G high-speed transmission to improve the data upload efficiency.

[0217] Cloud analysis and decision-making: Enhance the digital twin function to support VR / AR visualization; adopt a multi-objective optimization algorithm to generate the optimal control strategy.

[0218] Dynamic regulation and feedback: Dynamically adjust the maintenance conditions according to the multi-objective optimization results; provide real-time feedback on the regulation effects to optimize subsequent decisions.

[0219] In this embodiment, by adding an intelligent diagnosis and early warning module, enhancing the digital twin function, and adopting a multi-objective optimization and regulation strategy, the intelligence level and practicality of the system are further improved. Through the deep reinforcement learning algorithm, the abnormal recognition accuracy rate is increased to over 95%; VR / AR visualization is supported to provide immersive monitoring and decision-making support. Through the multi-objective optimization algorithm, on the premise of ensuring performance, the maintenance energy consumption is reduced by 20%-30%.

[0220] The above specific embodiments are only several alternative embodiments of the present invention. Based on the technical solution of the present invention and the relevant inspirations of the above embodiments, those skilled in the art can make various alternative improvements and combinations to the above specific embodiments.

Claims

1. A method for dynamic detection and analysis of early performance of ultra-high performance concrete, characterized in that: The following steps are involved: Establish a raw material characteristic spectrum library based on multi-physical field coupling, including three-dimensional material fingerprints of cement matrix phase, mineral admixture phase, and fiber reinforcement phase; A dynamic detection array based on time series was constructed. A distributed optical fiber sensor array was buried inside the concrete specimen, and a multi-spectral imaging device and an acoustic emission sensor array were arranged outside to form a four-dimensional monitoring network. Develop a multimodal data fusion algorithm to align electrical impedance tomography data, hydration heat distribution data, and dielectric property data in time and space, and extract hidden layer coupling features through variational autoencoders; A nonlinear coupling effect decomposition model was established, and the tensor decomposition technology was used to separate the coupling effects of fiber orientation distribution, interface transition zone development, and hydration reaction process on early performance. Implement dynamic parameter inversion and combine the improved particle swarm optimization algorithm to simultaneously solve the hydration dynamics parameters, shrinkage stress field distribution and microscopic pore evolution equations; Construct a cross-scale performance prediction model and establish a mapping relationship from nanoscale CSH gel growth to macroscopic compressive strength development through long short-term memory networks; Design an adaptive control strategy, dynamically adjust the maintenance environment parameters based on real-time performance deviation analysis, and establish a multi-objective optimization control model of temperature-humidity-constraint conditions; Deploy an edge computing node network and build a distributed computing architecture at the pouring site to achieve local feature extraction of monitoring data and incremental updates of cloud models; Establish a full life cycle data chain, deeply associate early performance test data with structural health monitoring data during service, and form a digital twin of performance evolution.

2. The method for dynamic detection and analysis of early performance of ultra-high performance concrete according to claim 1, characterized in that: The three-dimensional material fingerprint comprises: X-ray diffraction characteristic peak area ratio, backscattered electron image grayscale distribution, and nanoindentation modulus distribution of cement matrix phase; Atomic force microscope surface roughness parameters, Raman spectrum characteristic peak shift, zeta potential-particle size distribution curve of mineral admixture phase; Micro-CT three-dimensional reconstruction parameters of fiber-reinforced phase, single-filament tensile stress relaxation curve, and interface bonding-slip constitutive relationship.

3. The method for dynamic detection and analysis of early performance of ultra-high performance concrete according to claim 1, characterized in that: The four-dimensional monitoring network comprises: Bragg grating array of embedded fiber optic sensors to monitor the evolution of internal strain and temperature fields with a spatial resolution of 2 mm; The multispectral imaging device collects the dynamics of surface water evaporation and microcrack growth at a frequency of 10 Hz in the 400-2500 nm band; The acoustic emission sensor array uses waveform manifold learning technology to identify fiber pullout, matrix cracking and interface debonding events.

4. The method for dynamic detection and analysis of early performance of ultra-high performance concrete according to claim 1, characterized in that: The multimodal data fusion algorithm includes: A spatiotemporal registration coordinate system was established to perform voxel-level registration of the conductivity distribution data of electrical impedance tomography with the hydration thermal infrared imaging data; Develop a feature-level fusion module that uses a bidirectional gated recurrent unit weighted by an attention mechanism to process multi-source heterogeneous time series data; Construct a physical constraint loss function and embed the mass conservation equation and energy balance equation into the neural network training process.

5. The method for dynamic detection and analysis of early performance of ultra-high performance concrete according to claim 1, characterized in that: The dynamic parameter inversion includes: A parameter sensitivity ranking model was established to determine the priority of key inversion parameters based on the Morris global sensitivity analysis method; Design a hybrid particle swarm optimization strategy and introduce the taboo table mechanism of taboo search into the particle update process to prevent falling into the local optimum; A multi-objective fitness function is constructed to simultaneously optimize the hydration degree prediction error, shrinkage strain root mean square error and pore fractal dimension matching.

6. The method for dynamic detection and analysis of early performance of ultra-high performance concrete according to claim 1, characterized in that: The cross-scale performance prediction model comprises: Nanoscale module: CSH gel growth kinetics model based on molecular dynamics simulation; Microscale module: micromechanical constitutive model considering the characteristics of the interface transition zone; Macroscale module: 3D damage evolution model integrating digital image correlation technology.

7. The method for dynamic detection and analysis of early performance of ultra-high performance concrete according to claim 1, characterized in that: The adaptive control strategy includes: An environmental response surface model was established, and the influence of temperature and humidity interaction on performance development was determined through central composite design experiments; Develop fuzzy predictive control algorithms to dynamically adjust steam maintenance parameters and external constraints based on rolling horizon optimization; Build a safety boundary early warning system to calculate in real time the deviation between the current maintenance path and the ideal performance development trajectory.

8. The method for dynamic detection and analysis of early performance of ultra-high performance concrete according to claim 1, characterized in that: The edge computing node network comprises: Local feature extraction unit: Real-time wavelet packet decomposition of sensor data based on FPGA parallel computing architecture; Fog computing layer: deploy lightweight convolutional neural networks to perform key feature pattern recognition; Cloud-based collaborative platform: Uses a federated learning framework to implement privacy-preserving model updates for multi-project data.

9. The method for dynamic detection and analysis of early performance of ultra-high performance concrete according to claim 1, characterized in that: The full life cycle data link construction method includes: Establishing spatiotemporal coding rules to hash the timestamps of early detection data with the spatial coordinates of the structure; Develop a knowledge graph engine to build a multi-dimensional relationship network of material composition-process parameters-early performance-long-term durability; Design a transfer learning interface to migrate the laboratory-scale detection model to the digital twin of the actual engineering structure.

10. A dynamic detection and analysis system for early performance of ultra-high performance concrete, characterized in that: include: A multi-physics field coupled monitoring module, which includes a distributed fiber optic sensor array, a multi-spectral imager, an acoustic emission sensor, and an electrical impedance tomography device; Edge intelligent processing terminal, equipped with programmable logic devices and embedded neural network acceleration chips; A cloud-based collaborative analysis platform, which deploys a coupling effect decomposition model library and a performance prediction digital twin; Dynamic control actuator, including intelligent temperature control and maintenance cover, adaptive restraint device and micro-environment adjustment system; Data security interaction gateway, using blockchain technology to achieve tamper-proof storage and trusted sharing of detection data; Among them, each module realizes data interaction through the industrial Internet of Things protocol, forming an "end-edge-cloud" collaborative architecture.

Images