Intelligent damage monitoring control method and system for steel-UHPC (Ultra High Performance Concrete) combined bridge deck slab

Abstract

The invention discloses an intelligent damage monitoring control method and system for a steel-UHPC (Ultra High Performance Concrete) combined bridge deck slab. The intelligent damage monitoring control method comprises the following steps: arranging a multi-source sensor in a key stress area of the steel-UHPC combined bridge deck slab; the data acquisition module synchronously acquires multiple physical quantity signals monitored by each sensor; the feature extraction module captures deep correlation characteristics among physical quantities to obtain enhanced feature vectors; the state recognition and prediction module dynamically recognizes the health state of the structure according to the enhanced feature vector and predicts the crack evolution trend; the active control module deduces local pressure stress needing to be applied to active regulation and control of the steel bridge deck slab and the UHPC according to the intelligent identification and evolution prediction result; after regulation and control are executed, the feedback optimization module collects operation data and feedback information of each module, optimizes and adjusts parameters and control strategies of the system, and feeds back the optimized parameters and strategies to each module to form a closed-loop self-adaptive control mechanism; according to the invention, full-life-cycle health management of the steel-UHPC combined bridge deck can be realized.

Description

Technical Field

[0001] This invention belongs to the field of bridge engineering technology, and more specifically, relates to an intelligent damage monitoring and control method and system for steel-UHPC composite bridge decks. Background Technology

[0002] With the continuous growth of traffic volume and the extension of bridge service life, bridge structures are subjected to multiple coupled effects, including heavy traffic loads, temperature loads, and corrosive media, posing a severe challenge to structural durability and service safety. Steel-UHPC composite bridge decks, as an emerging bridge superstructure system in recent years, have gradually gained widespread application in long-span bridges and heavy-load traffic bridges due to their superior mechanical properties. Compared with traditional steel-concrete composite bridge decks, steel-UHPC composite bridge decks possess high strength, high toughness, and high density, effectively improving the stiffness and fatigue performance of steel bridge decks, significantly extending structural service life, and reducing maintenance costs, demonstrating significant technological advantages and application prospects.

[0003] Despite the superior material properties and service performance of steel-UHPC composite bridge decks, numerous potential risks and challenges remain in practical engineering applications as service life increases. Firstly, under long-term traffic loads, initial defects inevitably exist within the welded joints connecting the UHPC layer and the steel bridge deck. During service, the coupling effect of loads and corrosive media easily induces the initiation and propagation of microcracks. Secondly, the bonding performance of the steel-UHPC interface is crucial to the synergistic stress distribution of the composite bridge deck. Discontinuities in material properties and differences in thermal expansion coefficients exist at the interface, making it prone to interface performance degradation during service, reducing the synergistic effect, and affecting the overall structural stiffness and durability.

[0004] Currently, traditional structural health monitoring relies primarily on a limited number of discrete sensors for passive data acquisition. This results in limited monitoring information, making it difficult to comprehensively perceive and identify key damage details early. Furthermore, existing monitoring methods generally lack proactive control capabilities, failing to take timely and effective intervention measures after anomalies are detected to prevent damage propagation and stress concentration. This delayed monitoring response easily leads to damage accumulation, accelerated structural fatigue deterioration, ultimately shortening the bridge's service life and even causing major safety accidents.

[0005] Therefore, there is an urgent need for a new type of intelligent monitoring and control system that integrates real-time perception of multi-source information, intelligent identification of potential damage, and active control. This system can achieve closed-loop management from early damage identification and trend prediction to stress optimization and control, thereby delaying the damage propagation process and improving the structural service performance and safety reliability. Summary of the Invention

[0006] To address the aforementioned deficiencies or improvement needs of existing technologies, this invention provides an intelligent damage monitoring and control method and system for steel-UHPC composite bridge decks. Multiple types of high-precision sensors (such as fiber optic grating sensors, distributed strain monitoring systems, and acoustic emission sensors) are deployed in key areas of the bridge deck, including the steel-UHPC interface, the interior of the UHPC, and typical crack locations in the steel bridge deck. This enables real-time acquisition of multi-physical field information such as stress, strain, crack propagation, temperature, and corrosion potential. By fusing multi-source heterogeneous data, and based on intelligent sensing algorithms, micro-damage identification and evolution trend prediction are achieved, enhancing the ability to perceive and identify early-stage hidden damage and complex damage mechanisms. Simultaneously, combined with an active stress control module, intelligent actuators (such as piezoelectric actuators and shape memory alloy actuators) are used to regulate the local stress field, reducing stress concentration effects, suppressing crack propagation rates, and delaying the damage evolution process. Through deep fusion of multi-field information, accurate identification of damage evolution mechanisms, and real-time response of active stress regulation, the system can realize full life-cycle health management of steel-UHPC composite bridge decks, improve the safety, durability and reliability of bridges during service, provide a new technology path for intelligent operation and maintenance and life extension design of large bridge structures, and has important engineering application value and promotion prospects.

[0007] To achieve the above objectives, one aspect of the present invention provides an intelligent damage monitoring and control method for steel-UHPC composite bridge decks, comprising the following steps:

[0008] S1: Fiber optic grating sensors, piezoelectric ceramic sheet sensors and humidity-sensitive electrode sensors are deployed in the critical stress area of ​​the steel-UHPC composite bridge deck.

[0009] S2: The data acquisition module synchronously acquires the multi-physical quantity signals monitored by each sensor in step S1 through the shielded cable, and performs filtering and normalization processing on the acquired data to obtain the structural state feature vector at the current moment, forming a complete structural health monitoring database.

[0010] S3: The feature extraction module constructs a nonlinear interaction model. Based on the structural health monitoring data obtained in step S2, a feature interaction enhancement mechanism is introduced. Through nonlinear mapping and feature interaction, the deep-level correlation characteristics between physical quantities are captured, and the enhanced feature vector is obtained.

[0011] S4: The state recognition and prediction module establishes a dynamic health state recognition and evolution trend prediction model based on the gated recursion mechanism. According to the enhanced feature vector provided by the feature extraction module, it dynamically identifies the structural health state and predicts the crack evolution trend. The prediction results are sent to the active control module, and relevant information is fed back to the feedback optimization module.

[0012] S5: Based on the intelligent identification and evolution prediction results, the active control module derives the local compressive stress required to actively regulate the cracks in the steel bridge deck and UHPC cracks. The system generates the control command of the intelligent actuator and applies the local compressive stress field in real time to realize the active regulation of the expansion and stress state of the cracks in the steel bridge deck and UHPC cracks.

[0013] S6: After the control is executed, the feedback optimization module corrects the model parameters in real time based on the latest observation data; updates the feature extraction module, state recognition and prediction module and active control module; optimizes the control cost function through an adaptive learning mechanism, dynamically adjusts various control parameters, and realizes continuous improvement of system performance and self-learning evolution.

[0014] Furthermore, the key stress areas of the steel-UHPC composite bridge deck mentioned in step S1 include the steel-UHPC interface, the interior of the UHPC, and the typical crack locations of the steel bridge deck.

[0015] Furthermore, the structural state feature vector at the current moment in step S2 is calculated using equation (1):

[0016] X t =[x ε x AE x H x interface ,...] T (1)

[0017] Among them, X t x is the structural state feature vector at the current time t; ε Let x be the strain statistical eigenvector. AE x is the characteristic subvector of the acoustic radio frequency domain. H x is the feature vector of humidity change. interface It is the characteristic vector of interface slip and adhesion degradation.

[0018] Furthermore, the feature enhancement in step S3 is achieved through a primary mapping and a secondary interaction, specifically as follows:

[0019]

[0020] Among them, z t Let W1 be the enhanced feature vector at time t, W2 be the first-order mapping weight matrix, and b be the second-order feature interaction weight matrix. e Let σ be the feature enhancement bias vector, σ(·) be the nonlinear activation function, and o represent the Hadamard product.

[0021] Further, step S4 includes:

[0022] S41: Establish a dynamic health status identification and evolution trend prediction model based on a gated recursive mechanism;

[0023] S42: Calculate the updated weights by combining the health status of the previous time step with the current characteristics;

[0024] S43: Based on the updated weights, the historical state and current characteristics are weighted to calculate the updated health status at the current moment, dynamically reflecting the damage accumulation process and providing support for continuous tracking of crack and fissure development.

[0025] S44: Based on the updated health status at the current moment, use the weight vector and bias term of the damage prediction output to predict the damage vector at future moments, including the steel plate crack length, UHPC crack length and interface slip.

[0026] Furthermore, the update weight g described in step S42 t Calculated using equation (3):

[0027] g t =σ(Wgh) t-1 +V g z t +b g (3)

[0028] Among them, g t The gating vectors are used to adjust the fusion ratio of historical information and new features; Wg and V g b is the weight matrix for the gating mechanism, which is applied to the historical state vector and the current feature sub-vector, respectively; g Here, σ is the bias term of the gated unit; σ(·) is the Sigmoid activation function; h t-1 This is the health state vector from the previous moment.

[0029] Furthermore, the updated health status at the current moment described in step S43 is represented by equation (4):

[0030]

[0031] Among them, h t h is the updated health state vector at the current moment. t-1 W is the health state vector from the previous time step, recursively obtained from the initial state h0; h and V h b is a recursive weight matrix for health status, corresponding to historical status and current features respectively; h Update the bias term for the health status; tanh(·) is the hyperbolic tangent activation function.

[0032] Furthermore, the damage vector at the future time step described in step S44 is represented by equation (5):

[0033]

[0034] Among them, S t+1 a represents the damage vector at a future time step; t+1 To predict the typical fatigue detail crack length of a steel bridge deck, l t+1 δ is the predicted UHPC crack length. t+1 W is the predicted interface slip. O，S W O，U W O，I These are the weight vectors for the steel structure, UHPC, and interface damage prediction outputs, respectively; b O，S b O，U b O，I These correspond to the bias terms in the prediction outputs for steel structures, UHPC, and interface damage, respectively.

[0035] Furthermore, the formula for calculating the control stress for crack propagation in the steel bridge deck in step S5 is expressed by equation (6):

[0036]

[0037] Where, σ pzt Applying local compressive stress to the piezoelectric actuator, σ max The maximum principal stress at the fatigue detail is (ΔK). th is the crack propagation threshold value, and Y is the crack geometry correction coefficient;

[0038] The calculation of crack control stress in UHPC is expressed by equation (7):

[0039]

[0040] Where, σ sma The local compressive stress applied to the shape memory alloy, σ t For UHPC tensile stress, G target The target value is set for the energy release rate control.

[0041] The second aspect of the present invention provides an intelligent damage monitoring and control system for steel-UHPC composite bridge decks, used to implement the intelligent damage monitoring and control method for steel-UHPC composite bridge decks, comprising: a data acquisition module, a feature extraction module, a state recognition and prediction module, an active control module, and a feedback optimization module;

[0042] Multiple monitoring areas requiring close monitoring are designated on the bridge using a steel-UHPC composite bridge deck. These monitoring areas are equipped with multi-source sensors for monitoring strain changes, crack acoustic emission characteristics, and interface microenvironment humidity. The multi-source sensors are connected to a data acquisition module. These sensors include a fiber optic grating sensor for monitoring strain changes, a piezoelectric ceramic sheet sensor for monitoring crack acoustic emission characteristics, and a humidity-sensitive electrode sensor for monitoring the microenvironment humidity at the steel-UHPC interface.

[0043] The data acquisition module synchronously acquires multi-physical quantity signals sensed by multiple source sensors through shielded cables. After filtering and normalization, a complete structural health monitoring database is formed, and the processed data is transmitted to the feature extraction module.

[0044] The feature extraction module constructs a nonlinear interaction model to process the multi-source data collected by the data acquisition module, introduces a feature interaction enhancement mechanism, captures the deep-level correlation characteristics between physical quantities through nonlinear mapping and feature interaction, obtains enhanced feature vectors, and passes the extracted enhanced feature vectors to the state recognition and prediction module.

[0045] The state recognition and prediction module establishes a dynamic health state recognition and evolution trend prediction model based on a gated recursion mechanism. According to the enhanced feature vector provided by the feature extraction module, it dynamically identifies the structural health state and predicts the crack evolution trend. The prediction results are sent to the active control module, and relevant information is fed back to the feedback optimization module.

[0046] The active control module actively adjusts the local compressive stress required to be applied to the steel bridge deck cracks and UHPC cracks based on the intelligent identification and evolution prediction results, generates intelligent actuator control commands, and drives the intelligent actuator to actively adjust the expansion and stress state of the steel bridge deck cracks and UHPC cracks.

[0047] After the control is executed, the feedback optimization module corrects the model parameters in real time based on the latest observation data and updates the feature extraction, state recursion and control planning modules. It optimizes the control cost function through an adaptive learning mechanism, dynamically adjusts various control parameters, and feeds back the optimized parameters and strategies to each module to form a closed-loop adaptive control mechanism.

[0048] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects:

[0049] (1) The intelligent damage monitoring and control method and system for steel-UHPC composite bridge decks of the present invention addresses the problem that existing technologies cannot achieve comprehensive perception of the structural state during the service life of steel-UHPC composite bridge decks due to the complex internal stress and diverse damage evolution processes. A multi-source intelligent monitoring and active control system covering steel-UHPC composite bridge decks in cable-stayed bridge systems is constructed. The system deploys fiber optic grating sensors, piezoelectric ceramic sheet sensors, and humidity-sensitive electrode sensors in key stress areas of the bridge deck to monitor strain changes, crack acoustic emission characteristics, and interface microenvironment humidity, respectively, achieving synchronous monitoring of multiple physical quantities such as strain, acoustic emission, and humidity. The system is designed with a multi-source synchronous signal acquisition scheme, fusing strain, acoustic emission, and humidity signals to form multi-dimensional damage identification basic data. The data acquisition module synchronously acquires these multi-physical quantity signals of the multi-dimensional damage identification basic data through shielded cables, and performs filtering and normalization processing to form a structural state feature vector at the current moment, providing comprehensive data support for damage identification. This fusion of multi-source information overcomes the limitations of traditional monitoring methods that rely only on limited sensor data and have incomplete monitoring information.

[0050] (2) The intelligent damage monitoring and control method and system for steel-UHPC composite bridge decks of the present invention addresses the problem that the damage evolution process involves changes in multiple physical field signals, and that a single signal feature is insufficient to accurately describe the complex damage mechanism. It constructs a nonlinear interactive model through a feature extraction module to extract high-dimensional damage features and introduces a feature interaction enhancement mechanism. Through nonlinear mapping and feature interaction, it captures the deep-seated correlation characteristics between physical quantities, thereby improving the ability to identify minute damage changes. Compared to traditional methods, this module effectively enhances the expression of damage features through a primary mapping and secondary interaction process, making damage identification more accurate.

[0051] (3) The intelligent damage monitoring and control method and system for steel-UHPC composite bridge decks of the present invention addresses the obvious temporal and irreversible characteristics of the damage evolution process of steel-UHPC composite bridge decks. The evolution process of cracks from initiation to propagation and from interface bonding to slippage exhibits strong cumulative and nonlinear variation characteristics. The system establishes a dynamic health status identification and evolution trend prediction model based on a gated recursive mechanism to capture the state change process of the structure at different stages of service. Specifically, the state identification and prediction module adopts a gated recursive mechanism to dynamically identify the structural health status and predict the crack evolution trend. It adjusts the fusion ratio of historical information and new features through the gated vector to avoid misjudgment caused by small fluctuations. At the same time, it uses the health status recursive model to accurately reflect the damage accumulation process, providing support for continuous tracking of crack and fissure development. Compared with traditional static or simple prediction models, this module can more accurately capture changes in the structural health status, providing a more reliable basis for taking maintenance measures in advance.

[0052] (4) The intelligent damage monitoring and control method and system for steel-UHPC composite bridge deck of the present invention, the active control module derives the local compressive stress to be applied for active regulation based on the intelligent identification and evolution prediction results, generates actuator control commands based on the prediction results, drives the piezoelectric actuator and shape memory alloy wire and other actuators to apply local compressive stress field in real time, actively regulates crack propagation and stress state, and delays the further development of initial damage or existing damage in typical fatigue details of steel bridge deck and UHPC; this method breaks through the bottleneck of traditional monitoring methods lacking active regulation capability, and can take effective intervention measures in a timely manner after detecting abnormalities, delay damage propagation, and improve the safety and durability of the structure.

[0053] (5) The intelligent damage monitoring and control method and system for steel-UHPC composite bridge deck of the present invention, after the control is executed, the feedback optimization module corrects the model parameters in real time based on the latest observation data and updates the feature extraction, state recursion and control planning modules; the control cost function is optimized through an adaptive learning mechanism, and the control parameters are dynamically adjusted to achieve continuous improvement of system performance and self-learning evolution; this enables the system to continuously adapt to changes in structural state and actual damage evolution, and always maintain the best operating state.

[0054] (6) The intelligent damage monitoring and control method and system for steel-UHPC composite bridge decks of the present invention integrates multiple physical quantity information such as strain changes in the monitoring area, acoustic emission characteristics of cracks, and humidity of the interface microenvironment to achieve real-time identification and trend prediction of key damage elements; it can form a response strategy in the early stage of damage, actively regulate the local stress state of the structure, and dynamically intervene in crack propagation path and interface slip behavior, thereby significantly improving the safety and maintainability of steel-UHPC composite bridge decks during service. The present invention provides more accurate decision support for bridge maintenance by dynamically predicting key indicators such as crack length and interface slip. The feedback optimization module supports real-time strategy updates after regulation response, with a short response cycle, enabling continuous iterative regulation and effectively improving the long-term stability and adaptability of the structure. The piezoelectric actuator applies high-frequency stress disturbance, which significantly reduces the stress amplitude of fatigue details, and the shape memory alloy wire controls the residual opening width after crack closure to be extremely small, all of which help to extend the service life of the bridge. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the structure of an intelligent damage monitoring and control system for a steel-UHPC composite bridge deck according to an embodiment of the present invention;

[0056] Figure 2 This is a flowchart illustrating an intelligent damage monitoring and control method for steel-UHPC composite bridge deck according to an embodiment of the present invention.

[0057] Figure 3This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention.

[0058] In all the accompanying drawings, the same reference numerals indicate the same technical features, specifically: 1-bridge with steel-UHPC composite bridge deck, 2-monitoring area, 3-cable, 4-data acquisition module, 5-feature extraction module, 6-state identification and prediction module, 7-active control module, 8-feedback optimization module, 9-equipment room, 10-data processing center. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0060] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, when an element is referred to as "fixed to," "set on," or "provided on" another element, it can be directly on or indirectly on the other element. When an element is referred to as "connected to" another element, it can be directly connected to or indirectly connected to the other element. The terms "mounted," "connected," "linked," and "provided with" should be interpreted broadly. For example, it can refer to a fixed connection, a detachable connection, or an integral connection; it can refer to a mechanical connection or an electrical connection; it can refer to a direct connection or an indirect connection through an intermediate medium; it can refer to the internal communication of two elements or the interaction between two elements. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0061] Example 1

[0062] like Figure 1As shown, Embodiment 1 of the present invention provides an intelligent damage monitoring and control system for steel-UHPC composite bridge decks, applicable to all bridges using steel-UHPC composite bridge decks. The system includes a data acquisition module 4, a feature extraction module 5, a state recognition and prediction module 6, an active control module 7, and a feedback optimization module 8. Multiple monitoring areas 2 requiring close monitoring are designated on the bridge 1 using steel-UHPC composite bridge decks (this embodiment uses a cable-stayed bridge as an example). Multi-source sensors for monitoring strain changes, crack acoustic emission characteristics, and interface microenvironment humidity are deployed in these monitoring areas 2. These multi-source sensors are connected to the data acquisition module 4. The data acquisition module 4 collects the raw data from the multi-source sensors and transmits the processed data to the feature extraction module. The feature extraction module 5 receives the data. The data from the acquisition module is used to extract feature vectors and pass them to the state recognition and prediction module. The state recognition and prediction module 6 uses the feature vectors provided by the feature extraction module to perform state recognition and prediction, sends the prediction results to the active control module, and feeds back relevant information to the feedback optimization module to optimize the entire monitoring and control process. The active control module 7 receives the prediction results from the state recognition and prediction module, generates control commands and sends them to the intelligent actuators (including piezoelectric actuators and shape memory alloy wire actuators), and feeds back the execution status to the feedback optimization module 8. The feedback optimization module 8 collects the operating data and feedback information from each module, optimizes and adjusts the system parameters and control strategies, and feeds back the optimized parameters and strategies to each module to form a closed-loop adaptive control mechanism.

[0063] Furthermore, the multi-source sensors include a fiber optic grating sensor for monitoring strain changes, a piezoelectric ceramic sheet sensor for monitoring the acoustic emission characteristics of cracks, and a humidity-sensitive electrode sensor for monitoring the microenvironment humidity at the steel-UHPC interface. The data acquisition module 4 is installed in a location on the bridge deck that is easy to route and can aggregate data from various sensors, such as in the equipment room 9 below the bridge deck, and is connected to each sensor in the bridge deck monitoring area 2 via shielded cables. The feature extraction module 5 is integrated with the data acquisition module 4 and is located in the equipment room 9 below the bridge deck, facilitating real-time processing of the acquired data. The state recognition and prediction module 6 is integrated in the server or control system of the data processing center 10, performing in-depth analysis and processing of the data from the feature extraction module. The control unit of the active control module 7 is placed in the equipment room 9 below the bridge deck and is connected to actuators such as piezoelectric actuators and shape memory alloy wires, performing corresponding stress adjustment on the bridge deck according to instructions. The feedback optimization module 8 is integrated in the server or control system of the data processing center 10, interconnected with each module, and monitors and optimizes the operating status of the entire system.

[0064] Furthermore, the data acquisition module 4 synchronously acquires multi-physical quantity signals sensed by multiple sources of sensors through shielded cable 31. After filtering and normalization, it obtains structural state feature vectors, forming a complete structural health monitoring database, and transmits the processed data to the feature extraction module 5. The feature extraction module 5 constructs a nonlinear interaction model to process the multi-source data acquired by the data acquisition module 4, extracts high-dimensional damage features, introduces a feature interaction enhancement mechanism, and captures the deep-level correlation characteristics between physical quantities through nonlinear mapping and feature interaction, improving the ability to identify minute damage changes, obtaining enhanced feature vectors, and transmitting the extracted enhanced feature vectors to the state recognition and prediction module 6. The state recognition and prediction module 6 establishes a dynamic health state recognition and evolution trend prediction model based on a gated recursive mechanism. According to the enhanced feature vectors provided by the feature extraction module, it dynamically identifies the structural health state and predicts the crack evolution trend, sends the prediction results to the active control module 7, and simultaneously feeds back relevant information to the feedback optimization module 8 to optimize the entire monitoring and control process. The active control module 7... Based on the intelligent identification and evolution prediction results, the local compressive stress required to actively regulate the cracks in the steel bridge deck and UHPC cracks is generated to generate intelligent actuator control commands. These commands drive the intelligent actuators to actively regulate the propagation and stress state of the cracks in the steel bridge deck and UHPC, delaying the further development of initial or existing damage to typical fatigue details of the steel bridge deck and UHPC. The intelligent actuators include piezoelectric actuators and shape memory alloy wires. After the regulation is executed, the feedback optimization module 8 corrects the model parameters in real time based on the latest observation data and updates the feature extraction, state recursion, and regulation planning modules. The regulation cost function is optimized through an adaptive learning mechanism, and various control parameters are dynamically adjusted to achieve continuous improvement in system performance and self-learning evolution. Through the cooperation of various modules, based on the data acquisition module 4 and feature extraction module 5, data support is provided for the state identification and prediction module 6. The results of the state identification and prediction module 6 guide the active control module 7 in regulation. The feedback optimization module 8 runs through the entire system, continuously optimizing the operation of each module to achieve intelligent damage monitoring and active control of the steel-UHPC composite bridge deck.

[0065] The intelligent damage monitoring and control system for steel-UHPC composite bridge decks provided by this invention can dynamically predict key indicators such as crack length and interface slip, with a prediction time window of 7–14 days and crack length prediction error controlled within ±2 mm. The active control module 7 applies high-frequency stress disturbance through a piezoelectric actuator in the intelligent actuator, reducing the fatigue detail stress amplitude by more than 15%, and the shape memory alloy wire controls the residual opening width after crack closure to be less than 0.05 mm. The feedback optimization module 8 supports real-time strategy updates after the control response, with a response cycle of less than 10 minutes, enabling continuous iterative control of the system and improving the long-term stability and adaptability of the structure.

[0066] Example 2

[0067] like Figure 2 As shown, Embodiment 2 of the present invention provides an intelligent damage monitoring and control method for steel-UHPC composite bridge decks, implemented using the aforementioned monitoring and control system, comprising:

[0068] S1: Fiber optic grating sensors, piezoelectric ceramic sheet sensors and humidity-sensitive electrode sensors are deployed in the critical stress area of ​​the steel-UHPC composite bridge deck.

[0069] The key stress areas of the steel-UHPC composite bridge deck in step S1 include the steel-UHPC interface, the interior of UHPC, and the typical crack locations of the steel bridge deck; the fiber optic grating sensor, piezoelectric ceramic sheet sensor, and humidity-sensitive electrode sensor are used to monitor strain changes, crack acoustic emission characteristics, and interface microenvironment humidity, respectively, and together constitute a multi-source sensor.

[0070] S2: Data acquisition module 4 synchronously acquires the multi-physical quantity signals monitored by each sensor in step S1 through shielded cable, including strain signal, acoustic emission signal and humidity signal, and performs filtering and normalization processing on the acquired data to remove noise interference and unify data dimensions, thereby obtaining the structural state feature vector at the current moment, forming a complete structural health monitoring database, and providing basic data support for subsequent damage identification and analysis.

[0071] Specifically, during the service life of the steel-UHPC composite bridge deck, the internal stress of the structure is complex, and the damage evolution process is diverse. In order to achieve a comprehensive perception of the structural state, the system is designed with a multi-source synchronous signal acquisition scheme, which integrates strain, acoustic emission, and humidity signals to form multi-dimensional damage identification basic data. After filtering and normalization, the structural state feature vector X at the current time t is obtained. t This will form a complete structural health monitoring database;

[0072] In step S2, the structural state feature vector at the current time t is calculated using equation (1):

[0073] X t =[x ε ,x AE ,x H ,x interface ,...] T (1)

[0074] Where, x ε Let x be the strain statistical eigenvector. AE x is the characteristic subvector of the acoustic radio frequency domain. H x is the feature vector of humidity change. interface It is the characteristic vector of interface slip and adhesion degradation.

[0075] S3: Feature extraction module 5 constructs a nonlinear interaction model. Based on the structural health monitoring data obtained in step S2, a feature interaction enhancement mechanism is introduced. Through nonlinear mapping and feature interaction, the deep-level correlation characteristics between physical quantities are captured, further enhancing feature expression and improving the ability to identify minute damage changes. The enhanced feature vector is obtained so as to more accurately reflect the damage state of the structure.

[0076] The feature enhancement in step S3 is achieved through a primary mapping and a secondary interaction, specifically as follows:

[0077]

[0078] Among them, z t Let W1 be the enhanced feature vector at time t, W2 be the first-order mapping weight matrix, and b be the second-order feature interaction weight matrix. e Let σ be the feature enhancement bias vector, σ(·) be the nonlinear activation function, and o represent the Hadamard product.

[0079] Step S3 introduces a feature interaction enhancement mechanism for the multi-source data collected by the data acquisition module; firstly, a mapping is performed, and the original feature vector is linearly transformed using the first mapping weight matrix W1 to extract the main feature components;

[0080] Then, a secondary feature interaction is performed. Through operations such as the secondary feature interaction weight matrix W2, the nonlinear activation function σ(·), and the Hadamard product, the deep-level correlation characteristics between physical quantities are captured, further enhancing the feature expression and improving the ability to identify minute damage changes. The enhanced feature vector is obtained so as to more accurately reflect the damage state of the structure.

[0081] S4: The state identification and prediction module 6 establishes a dynamic health state identification and evolution trend prediction model based on a gated recursive mechanism. Based on the enhanced feature vectors provided by the feature extraction module, it dynamically identifies the structural health state and predicts the crack evolution trend. The prediction results are sent to the active control module 7, and relevant information is fed back to the feedback optimization module 8 to optimize the entire monitoring and control process. Specifically, this includes:

[0082] S41: Establish a dynamic health status identification and evolution trend prediction model based on a gated recursive mechanism;

[0083] S42: Calculate the updated weights by combining the health status of the previous time step with the current characteristics;

[0084] The updated weight g t Calculated using equation (3):

[0085] g t=σ(Wgh) t-1 +V g z t +b g (3)

[0086] Among them, g t The gating vectors are used to adjust the fusion ratio of historical information and new features; Wg and V g b is the weight matrix for the gating mechanism, which is applied to the historical state vector and the current feature sub-vector, respectively; g Here, σ is the bias term of the gated unit; σ(·) is the Sigmoid activation function; h t-1 This is the health state vector from the previous time step;

[0087] The updated weight g t The weight matrices Wg and V of the gating mechanism g Gating unit bias term b g Together with the Sigmoid activation function σ(·), it is used to adjust the fusion ratio of historical information and new features, and avoid misjudgment due to small fluctuations;

[0088] S43: Based on the updated weights, the historical state and current characteristics are weighted to calculate the updated health status at the current moment, dynamically reflecting the damage accumulation process and providing support for continuous tracking of crack and fissure development.

[0089] The updated health status at the current moment is represented by equation (4):

[0090]

[0091] Among them, h t h is the updated health state vector at the current moment. t-1 W is the health state vector from the previous time step, recursively obtained from the initial state h0; h and V h b is a recursive weight matrix for health status, corresponding to historical status and current features respectively; h The bias term is updated to reflect the health status; tanh(·) is the hyperbolic tangent activation function;

[0092] S44: Based on the updated health status at the current moment, the damage vector at future moments is predicted using the weight vector and bias term of the damage prediction output, including the steel plate crack length, UHPC crack length and interface slip, thereby quantifying the damage development trend and providing a basis for subsequent active control.

[0093] The damage vector at the future moment is represented by equation (5):

[0094]

[0095] Among them, S t+1 a represents the damage vector at a future time step; t+1 To predict the typical fatigue detail crack length of a steel bridge deck, l t+1 δ is the predicted UHPC crack length. t+1 W is the predicted interface slip. O，S W O，U W O，I These are the weight vectors for the steel structure, UHPC, and interface damage prediction outputs, respectively; b O，S b O，U b O，I These correspond to the bias terms in the prediction outputs for steel structures, UHPC, and interface damage, respectively.

[0096] S5: Based on the intelligent identification and evolution prediction results, the active control module 7 derives the local compressive stress required to actively regulate the cracks in the steel bridge deck and UHPC cracks. The system generates the intelligent actuator control command to apply the local compressive stress field in real time, delaying the further development of the initial damage or existing damage of typical fatigue details in the steel bridge deck and UHPC, thereby actively regulating crack propagation and stress state.

[0097] For crack propagation control of steel bridge decks, the local compressive stress to be applied by the piezoelectric actuator in the intelligent actuator is calculated based on parameters such as crack propagation threshold, maximum principal stress at fatigue details, and crack geometric correction coefficient; the stress calculation formula for crack propagation control of steel bridge decks is expressed by equation (6):

[0098]

[0099] Where, σ pzt Applying local compressive stress to the piezoelectric actuator, σ max The maximum principal stress at the fatigue detail is (ΔK). th is the crack propagation threshold value, and Y is the crack geometry correction coefficient;

[0100] For UHPC crack control, the local compressive stress to be applied to the shape memory alloy in the smart actuator is determined based on factors such as the UHPC tensile stress and energy release rate control target value; the UHPC crack control stress calculation is expressed by equation (7):

[0101]

[0102] Where, σ sma The local compressive stress applied to the shape memory alloy, σ t For UHPC tensile stress, G target The target value for energy release rate control;

[0103] S6: After the control is executed, the feedback optimization module 8 corrects the model parameters in real time based on the latest observation data; updates the feature extraction module, state recursion module (i.e., state identification and prediction module 6), and control planning module (i.e., active control module 7) to adapt to changes in structural state and actual damage evolution; optimizes the control cost function through an adaptive learning mechanism, dynamically adjusts various control parameters, and achieves continuous improvement and self-learning evolution of system performance; enabling the entire intelligent damage monitoring and control system to continuously optimize monitoring and control effects, and better ensure the safety and reliability of the steel-UHPC composite bridge deck during its service life; specifically including:

[0104] S61. Monitoring system response: After active control is executed, the system continuously collects structural response data through sensors, covering multiple physical quantities such as strain, acoustic emission signals and humidity, as feedback information input into subsequent processes.

[0105] S62. Collect feedback data: Feedback optimization module 8 collects new sensor data, as well as the operating data of feature extraction module 5 and state recognition and prediction module 6, to understand whether feature extraction is effective and whether state recognition and prediction are accurate.

[0106] S63. Analyze feedback data: Use methods such as deviation analysis to find the deviation between the predicted damage trend and the actual monitoring data; a large deviation indicates that the model prediction is inaccurate and the parameters need to be adjusted; if the deviation is small or stable, the model is relatively accurate, but it may still need to be optimized to improve the accuracy.

[0107] S64. Evaluate model performance: Establish an indicator system, such as prediction accuracy and feature extraction efficiency; calculate indicator values ​​based on feedback data to determine whether the model performance still meets the monitoring and control requirements.

[0108] S65. Adjust model parameters: Based on the feedback data analysis results, use machine learning methods to adjust the parameters of the feature extraction module to optimize the feature extraction effect; use the model update algorithm to correct the weight matrix and bias terms of the state recognition and prediction module.

[0109] S66, Optimize control strategy

[0110] Based on the latest structural health status, optimize the control strategy of the active control module, such as adjusting the control signals of the piezoelectric actuator and shape memory alloy wire, to ensure the control effect;

[0111] S67, Update System Model

[0112] The adjusted model parameters and optimized control strategies are integrated into the monitoring and control system, and the system model is updated so that it can more accurately reflect the current health status and damage evolution trend of the structure.

[0113] S68. Verify the optimization effect

[0114] Run the monitoring and control system again in a real or simulated environment to verify the performance of the optimized system; compare the results before and after optimization to evaluate whether the optimization improves the accuracy of damage monitoring and the effect of active control.

[0115] S69. Continuous monitoring and adjustment: After verification, the system continues to monitor bridge damage, collects feedback data regularly, and evaluates model performance. If performance deteriorates, the above steps are repeated for adjustment and optimization to ensure that the system is always in the best operating condition.

[0116] The intelligent damage monitoring and control system and method for steel-UHPC composite bridge decks of this invention, by constructing a multi-source intelligent monitoring and active control closed-loop system, integrates information from multiple physical quantities such as strain, acoustic emission, and humidity, to achieve real-time identification and trend prediction of key damage elements of steel-UHPC composite bridge decks. The system can formulate response strategies in the early stages of damage initiation, actively regulate the local stress state of the structure, and dynamically intervene in crack propagation paths and interface slip behavior, significantly improving safety and maintainability during service. Compared with traditional methods, this invention achieves more accurate damage monitoring and more effective active control, possessing greater technical advantages and application value.

[0117] Example 3

[0118] Embodiment 3 of the present invention provides an electronic device, Figure 3 This is a schematic diagram of the electronic device in this embodiment, as shown below. Figure 3 As shown, the electronic device 1000 in this embodiment may include: a processor 1001, a network interface 1004, and a memory 1005. Furthermore, the electronic device 1000 may also include: a user interface 1003, and at least one communication bus 1002. The communication bus 1002 is used to implement communication between these components. The user interface 1003 may include a display screen and a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be a high-speed RAM memory or a non-volatile memory, such as at least one disk storage device. Optionally, the memory 1005 may also be at least one storage device located remotely from the aforementioned processor 1001. Figure 3 As shown, the memory 1005, which is a computer-readable storage medium, may include an operating system, a network communication module, a user interface module, and a device control application.

[0119] like Figure 3In the electronic device 1000 shown, the network interface 1004 provides network communication functions; the user interface 1003 is mainly used to provide an input interface for users; and the processor 1001 can be used to call the device control application stored in the memory 1005 to implement the above-mentioned intelligent damage monitoring and control method for steel-UHPC composite bridge panels.

[0120] It should be understood that in some feasible implementations, the processor 1001 described above may be a central processing unit (CPU), which may also be other general-purpose processors, DSPs, ASICs, FPGAs, or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor. The memory may include read-only memory and random access memory, and provides instructions and data to the processor. A portion of the memory may also include non-volatile random access memory. For example, the memory may also store device type information.

[0121] In specific implementation, the aforementioned electronic device 1000 can perform the above-described actions through its built-in functional modules. Figure 2 The implementation methods provided for each step are detailed in the above-mentioned implementation methods, and will not be repeated here.

[0122] The electronic device provided in this embodiment constructs a multi-source intelligent monitoring and active control closed-loop system, integrating information from multiple physical quantities such as strain, acoustic emission, and humidity, to achieve real-time identification and trend prediction of key damage elements in steel-UHPC composite bridge decks. The system can formulate response strategies in the early stages of damage initiation, actively regulate the local stress state of the structure, and dynamically intervene in crack propagation paths and interface slip behavior, significantly improving safety and maintainability during service. Compared to traditional methods, this invention achieves more accurate damage monitoring and more effective active control, possessing greater technical advantages and application value.

[0123] Example 4

[0124] Embodiment 4 of this application also provides a computer-readable storage medium storing a computer program that is executed by a processor to implement... Figure 2 The methods provided in each step are detailed in the implementation methods provided in the above steps, and will not be repeated here.

[0125] The computer-readable storage medium provided in Embodiment 4 constructs a multi-source intelligent monitoring and active control closed-loop system, integrating information from multiple physical quantities such as strain, acoustic emission, and humidity, to achieve real-time identification and trend prediction of key damage elements in steel-UHPC composite bridge decks. The system can formulate response strategies in the early stages of damage initiation, actively regulate the local stress state of the structure, and dynamically intervene in crack propagation paths and interface slip behavior, significantly improving safety and maintainability during service. Compared to traditional methods, this invention achieves more accurate damage monitoring and more effective active control, possessing greater technical advantages and application value.

[0126] Any references to memory, storage, database, or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0127] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for intelligent damage monitoring and control of steel-UHPC composite bridge deck, characterized in that: Includes the following steps: S1: Fiber optic grating sensors, piezoelectric ceramic sheet sensors and humidity-sensitive electrode sensors are deployed in the critical stress area of ​​the steel-UHPC composite bridge deck. S2: The data acquisition module (4) synchronously acquires the multi-physical quantity signals monitored by each sensor in step S1 through the shielded cable, and performs filtering and normalization processing on the acquired data to obtain the structural state feature vector at the current moment, forming a complete structural health monitoring database. S3: Feature extraction module (5) constructs a nonlinear interaction model. For the structural health monitoring data obtained in step S2, a feature interaction enhancement mechanism is introduced. Through nonlinear mapping and feature interaction, the deep-level correlation characteristics between physical quantities are captured, and the enhanced feature vector is obtained. S4: State recognition and prediction module (6) Establish a dynamic health state recognition and evolution trend prediction model based on gated recursion mechanism. Based on the enhanced feature vector, dynamically identify the structural health state and predict the crack evolution trend. S5: The active control module (7) derives the local compressive stress required to actively control the cracks in the steel bridge deck and the UHPC cracks based on the intelligent identification and evolution prediction results. The system generates the control command of the intelligent actuator and applies the local compressive stress field in real time. S6: After the regulation is executed, the feedback optimization module (8) corrects the model parameters in real time based on the latest observation data; updates the feature extraction module (5), the state recognition and prediction module (6) and the active control module (7); optimizes the regulation cost function through the adaptive learning mechanism, dynamically adjusts various control parameters, and realizes the continuous improvement of system performance and self-learning evolution.

2. The intelligent damage monitoring and control method for steel-UHPC composite bridge decks according to claim 1, characterized in that: The key stress areas of the steel-UHPC composite bridge deck mentioned in step S1 include the steel-UHPC interface, the interior of the UHPC, and the typical crack locations of the steel bridge deck.

3. The intelligent damage monitoring and control method for steel-UHPC composite bridge decks according to claim 1, characterized in that: In step S2, the structural state feature vector at the current moment is calculated using equation (1): X t =[x ε ,x AE ,x H ,x interface ,…] T (1) Among them, X t x is the structural state feature vector at the current time t; ε Let x be the strain statistical eigenvector. AE x is the characteristic subvector of the acoustic radio frequency domain. H x is the feature vector of humidity change. interface It is the characteristic vector of interface slip and adhesion degradation.

4. The intelligent damage monitoring and control method for steel-UHPC composite bridge decks according to claim 3, characterized in that: The feature enhancement in step S3 is achieved through a primary mapping and a secondary interaction, specifically as follows: Among them, z t Let W1 be the enhanced feature vector at time t, W2 be the first-order mapping weight matrix, and b be the second-order feature interaction weight matrix. e Let σ be the feature enhancement bias vector, σ(·) be the nonlinear activation function, and o represent the Hadamard product.

5. The intelligent damage monitoring and control method for steel-UHPC composite bridge decks according to claim 4, characterized in that: Step S4 includes: S41: Establish a dynamic health status identification and evolution trend prediction model based on a gated recursive mechanism; S42: Calculate the updated weights by combining the health status of the previous moment with the current features; S43: Based on the updated weights, the historical state and current characteristics are weighted to calculate the updated health status at the current moment, dynamically reflecting the damage accumulation process and providing support for continuous tracking of crack and fissure development. S44: Based on the updated health status at the current moment, use the weight vector and bias term of the damage prediction output to predict the damage vector at future moments, including the steel plate crack length, UHPC crack length and interface slip.

6. The intelligent damage monitoring and control method for steel-UHPC composite bridge decks according to claim 5, characterized in that: The updated weight g described in step S42 t Calculated using equation (3): g t =σ(Wgh t-1 +V g With t +b g (3) Among them, g t W is a gating vector used to adjust the fusion ratio of historical information and new features; g and V g b is the weight matrix for the gating mechanism, which is applied to the historical state vector and the current feature sub-vector, respectively; g Here, σ is the bias term of the gated unit; σ(·) is the Sigmoid activation function; h t-1 This is the health state vector from the previous moment.

7. The intelligent damage monitoring and control method for steel-UHPC composite bridge decks according to claim 6, characterized in that: The updated health status at the current moment, as described in step S43, is represented by equation (4): Among them, h t h is the updated health state vector at the current moment. t-1 W is the health state vector from the previous time step, recursively obtained from the initial state h0; h and V h b is a recursive weight matrix for health status, corresponding to historical status and current features respectively; h Update the bias term for the health status; tanh(·) is the hyperbolic tangent activation function.

8. A method for intelligent damage monitoring and control of steel-UHPC composite bridge decks according to any one of claims 5-7, characterized in that: The damage vector at future time points mentioned in step S44 is represented by equation (5): Among them, S t+1 a represents the damage vector at a future time step; t+1 To predict the typical fatigue detail crack length of a steel bridge deck, l t+1 δ is the predicted UHPC crack length. t+1 W is the predicted interface slip. O，S W O，U W O，I These are the weight vectors for the steel structure, UHPC, and interface damage prediction outputs, respectively; b O，S b O，U b O，I These correspond to the bias terms in the prediction outputs for steel structures, UHPC, and interface damage, respectively.

9. A method for intelligent damage monitoring and control of steel-UHPC composite bridge decks according to any one of claims 1-7, characterized in that: The formula for calculating the control stress for crack propagation in the steel bridge deck in step S5 is expressed by equation (6): Where, σ pzt Applying local compressive stress to the piezoelectric actuator, σ max The maximum principal stress at the fatigue detail is (ΔK). th is the crack propagation threshold value, and Y is the crack geometry correction coefficient; The calculation of crack control stress in UHPC is expressed by equation (7): Where, σ sma The local compressive stress applied to the shape memory alloy, σ t For UHPC tensile stress, G target The target value is set for the energy release rate control.

10. An intelligent damage monitoring and control system for steel-UHPC composite bridge decks, characterized in that, The method for implementing the intelligent damage monitoring and control of steel-UHPC composite bridge deck as described in any one of claims 1-9 includes: a data acquisition module (4), a feature extraction module (5), a state recognition and prediction module (6), an active control module (7), and a feedback optimization module (8); The bridge (1) with steel-UHPC composite bridge deck is marked with multiple monitoring areas (2) that require special attention. Multi-source sensors for monitoring strain changes, crack acoustic emission characteristics and interface micro-environment humidity are installed in the monitoring areas (2). The multi-source sensors (3) are connected to the data acquisition module (4). The multi-source sensors (3) include a fiber optic grating sensor for monitoring strain changes, a piezoelectric ceramic sheet sensor for monitoring crack acoustic emission characteristics, and a humidity-sensitive electrode sensor for monitoring the micro-environment humidity of the steel-UHPC interface. The data acquisition module (4) synchronously acquires multi-physical quantity signals sensed by multi-source sensors (3) through shielded cable (3), and after filtering and normalization, forms a complete structural health monitoring database, and transmits the processed data to the feature extraction module (5); The feature extraction module (5) constructs a nonlinear interaction model, processes the multi-source data collected by the data acquisition module (4), introduces a feature interaction enhancement mechanism, captures the deep-level correlation characteristics between physical quantities through nonlinear mapping and feature interaction, obtains enhanced feature vectors, and passes the extracted enhanced feature vectors to the state recognition and prediction module (6). The state recognition and prediction module (6) establishes a dynamic health state recognition and evolution trend prediction model based on the gated recursion mechanism. According to the enhanced feature vector provided by the feature extraction module, it dynamically identifies the structural health state and predicts the crack evolution trend. The prediction results are sent to the active control module (7), and relevant information is fed back to the feedback optimization module (8). The active control module (7) actively regulates the local compressive stress required to be applied to the steel bridge deck cracks and UHPC cracks based on the intelligent identification and evolution prediction results, generates intelligent actuator control commands, and drives the intelligent actuator to actively regulate the expansion and stress state of the steel bridge deck cracks and UHPC cracks. After the regulation is executed, the feedback optimization module (8) corrects the model parameters in real time based on the latest observation data, updates the feature extraction, state recursion and regulation planning modules; optimizes the regulation cost function through the adaptive learning mechanism, dynamically adjusts various control parameters, and feeds back the optimized parameters and strategies to each module to form a closed-loop adaptive control mechanism.

Images