Bridge steel bar corrosion dynamic monitoring system based on fiber bragg grating sensing

Abstract

The invention provides a dynamic bridge steel bar corrosion monitoring system based on fiber bragg grating sensing, and relates to the technical field of bridge structure health monitoring and civil engineering safety, and the system comprises a fiber bragg grating sensor array which is arranged on the steel bar surface of a key part. The optical fiber grating sensor array is connected with the optical fiber demodulator, the optical fiber demodulator is connected with the data acquisition and communication unit, the optical fiber demodulator is used for resolving and obtaining data acquired by the optical fiber grating sensor array and sending the data to the data acquisition and communication unit, and the data acquisition and communication unit is connected with the optical fiber demodulator. The data acquisition and communication unit is connected with the data processing unit, and the data processing unit is used for dynamically monitoring bridge steel bar corrosion according to the acquired data. According to the method, millimeter-level quantitative diagnosis of steel bar corrosion is realized, the monitoring precision and timeliness are remarkably improved, and reliable technical support is provided for long-term operation and maintenance of a bridge structure.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of bridge structure health monitoring and civil engineering safety technology, in particular to a bridge steel bar corrosion dynamic monitoring system based on fiber grating sensing. BACKGROUND

[0002] Steel bar corrosion is one of the main factors leading to the durability decline, bearing capacity loss and even catastrophic failure of reinforced concrete bridges. Traditional steel bar corrosion monitoring methods such as half-cell potential method and linear polarization method have inherent shortcomings such as being susceptible to electromagnetic interference, being greatly affected by environment, being difficult to work stably for a long time, being able to provide only point information and being unable to quantitatively measure corrosion rate and corrosion products. Fiber Bragg grating sensors have outstanding advantages such as being resistant to electromagnetic interference, being corrosion-resistant, being small in size, being easy to distribute, being capable of absolute measurement and being good in long-term stability, and are very suitable for long-term monitoring in harsh environments such as bridges. However, existing technologies mostly remain in simple threshold judgment of physical quantity changes, and have shortcomings such as single sensing dimension, strong model dependency, low intelligent level and inability to quantitatively evaluate. For example, they usually rely on only a single parameter such as strain or temperature, and are difficult to accurately distinguish between corrosion strain and strain caused by load and temperature changes. Traditional decoupling methods rely on accurate physical models, but actual bridge structures are complex and have many environmental disturbances, and it is difficult to establish accurate physical models. They lack the ability to mine deep features of monitoring data, and are unable to accurately identify corrosion state, provide early warning and make trend prediction. They are difficult to directly and accurately convert sensor signals into key quantitative indicators such as steel bar corrosion depth and corrosion rate. In particular, existing monitoring methods have obvious deficiencies in data processing, and are unable to effectively combine advanced deep learning algorithms for multi-source signal decoupling, and lack effective automatic hyperparameter optimization mechanisms for model optimization, resulting in limited prediction accuracy and delayed warning response, and being unable to meet the urgent needs of modern bridge intelligent operation and maintenance. Therefore, there is an urgent need for a new steel bar corrosion monitoring method that can overcome the above shortcomings and realize intelligent, quantitative and dynamic real-time monitoring. Therefore, it is necessary to design a bridge steel bar corrosion dynamic monitoring system based on fiber grating sensing. SUMMARY

[0003] In order to overcome the deficiencies of the prior art, the purpose of the present application is to provide a bridge steel bar corrosion dynamic monitoring system based on fiber grating sensing.

[0004] To achieve the above purpose, the present application provides the following solutions: The application provides a bridge steel bar corrosion dynamic monitoring system based on a fiber grating sensor, which comprises a fiber grating sensor array, a fiber demodulator, a data acquisition and communication unit and a data processing unit, the fiber grating sensor array is arranged on the surface of steel bars at key positions and is used for collecting radial expansion pressure, point temperature and macro axial stress, the fiber grating sensor array is connected with the fiber demodulator, the fiber demodulator is connected with the data acquisition and communication unit, the fiber demodulator is used for calculating and obtaining the data collected by the fiber grating sensor array and sending the data to the data acquisition and communication unit, the data acquisition and communication unit is connected with the data processing unit, the data acquisition and communication unit is used for sending the obtained data to the data processing unit, and the data processing unit is used for performing bridge steel bar corrosion dynamic monitoring according to the obtained data. The key positions include water level fluctuation zones and splash zones of piers, positions below bridge deck drainage outlets and positions around expansion joints.

[0005] Preferably, the fiber grating sensor array comprises a fiber and radial expansion pressure sensors, point temperature sensors and macro axial stress sensors connected in series on the fiber, and the fiber is connected with the fiber demodulator.

[0006] Preferably, the data processing unit is used for performing bridge steel bar corrosion dynamic monitoring according to the obtained data, and specifically, Step 1: radial expansion pressure, point temperature and macro axial stress data are obtained and pretreated; Step 2: a bridge steel bar corrosion dynamic monitoring model is constructed; Step 3: a data set is constructed; Step 4: the bridge steel bar corrosion dynamic monitoring model is trained based on the data set, and a trained bridge steel bar corrosion dynamic monitoring model is obtained; Step 5: radial expansion pressure, point temperature and macro axial stress data to be monitored are input into the trained bridge steel bar corrosion dynamic monitoring model, and a corrosion indicating factor is obtained; Step 6: the corrosion indicating factor is used to calculate the steel bar corrosion depth; Step 7: dynamic early warning is performed based on the steel bar corrosion depth.

[0007] Preferably, in Step 2, the bridge steel bar corrosion dynamic monitoring model is constructed, and specifically, The bridge steel bar corrosion dynamic monitoring model is constructed based on an improved CNN-LSTM network structure, wherein the specific structure of the improved CNN-LSTM network structure comprises an input layer, two CNN modules, two LSTM modules and a full connection layer, wherein the two CNN modules each comprise a convolution layer and a pooling layer, the convolution layer of the first CNN module comprises 32 filters, and the convolution kernel size is 3, and the convolution layer of the second CNN module comprises 64 filters, and the convolution kernel size is also 3. Each of the LSTM modules is 64 hidden units.

[0008] Preferably, in step 3, a data set is constructed, specifically: A plurality of steel bars identical to those in step 1 are made and cast in concrete test blocks; Based on the controllable accelerated corrosion environment, the test blocks are experimented; During the experiment, the radial expansion pressure, the temperature of the measuring point, the macroscopic axial stress data and the actual corrosion depth are collected; The radial expansion pressure, the temperature of the measuring point, the macroscopic axial stress data of each time period are aligned with the actual corrosion depth measured at the corresponding time, wherein each radial expansion pressure, the temperature of the measuring point, the macroscopic axial stress data obtains a corrosion indicator as a label; The data set is constructed based on the aligned data.

[0009] Preferably, in step 4, the bridge steel bar corrosion dynamic monitoring model is trained based on the data set, specifically: The data set is divided, wherein 70% of the data in the data set is divided into a training set, 15% of the data is divided into a validation set, and 15% of the data is divided into a test set; Based on the training set and the validation set, the IGWO algorithm is used to train the bridge steel bar corrosion dynamic monitoring model to obtain the optimal hyperparameters; Based on the optimal hyperparameters, the bridge steel bar corrosion dynamic monitoring model is reinitialized, the training set and the validation set are combined, and the initialized bridge steel bar corrosion dynamic monitoring model is finally trained based on the combined data set; The finally trained bridge steel bar corrosion dynamic monitoring model is evaluated based on the test set.

[0010] Preferably, in step 6, the steel bar corrosion depth is calculated based on the corrosion indicator, specifically: A physically guided Gaussian process regression model is constructed; The physically guided Gaussian process regression model is trained based on a preset corrosion indicator-steel bar corrosion depth data set; The corrosion indicator to be detected is input into the trained Gaussian process regression model to obtain the steel bar corrosion depth.

[0011] Preferably, in step 7, a dynamic early warning is carried out based on the steel bar corrosion depth, specifically: A four-level early warning is constructed, wherein if the steel bar corrosion depth is <0.1mm, it is judged as a first-level early warning, the data processing unit records the steel bar corrosion depth, and no other processing is performed; If 0.1mm≤steel bar corrosion depth<0.3mm, it is judged as a second-level early warning, the data processing unit sends it to the monitoring platform for staff to view, and the monitoring interval time of the steel bar is shortened and the monitoring frequency is improved; If 0.3mm≤steel bar corrosion depth<0.5mm, it is judged as a third-level early warning, the data processing unit sends it to the monitoring platform, the monitoring platform generates an operation and maintenance report, and sends it to the maintenance management department for maintenance; If the steel bar corrosion depth is ≥0.5mm, it is judged as a fourth-level early warning, the data processing unit sends it to the monitoring platform, and the monitoring platform notifies the corresponding department to take traffic control and immediately reinforce and repair the treatment.

[0012] According to the specific embodiments provided by the present application, the following technical effects are disclosed: The application provides a bridge steel bar corrosion dynamic monitoring system based on a fiber grating sensor, which comprises a fiber grating sensor array, a fiber demodulator, a data acquisition and communication unit and a data processing unit, the fiber grating sensor array is arranged on the surface of the steel bar at a key position and is used for collecting radial expansion pressure, point temperature and macro axial stress, the fiber grating sensor array is connected with the fiber demodulator, the fiber demodulator is connected with the data acquisition and communication unit, the fiber demodulator is used for calculating and obtaining the data collected by the fiber grating sensor array and sending the data to the data acquisition and communication unit, the data acquisition and communication unit is connected with the data processing unit, the data acquisition and communication unit is used for sending the obtained data to the data processing unit, and the data processing unit is used for performing bridge steel bar corrosion dynamic monitoring according to the obtained data. The fiber grating sensor array integrating radial expansion pressure, point temperature and macro axial stress sensing functions is arranged on the surface of the steel bar at the key position, and the improved CNN-LSTM deep learning model and the IGWO optimization algorithm are fused, so that real-time, online and quantitative diagnosis and prediction of the corrosion state of the bridge steel bar are realized, and the application has the advantages of high precision and anti-interference ability, quantitative monitoring, dynamic real-time and forward-looking prediction, intelligence and self-learning, strong engineering practicability and the like. Specifically, through the multi-physical field FBG sensing array and deep learning signal decoupling, the interference of temperature and load strain on the corrosion signal is effectively removed, and the accuracy and reliability of the monitoring are significantly improved; the direct and quantitative mapping from physical signals to key engineering parameters such as steel bar corrosion depth and rate is realized, which surpasses the traditional qualitative or semi-quantitative judgment; the improved grey wolf optimization algorithm is combined, the model super parameter is automatically set, and the model convergence speed and prediction accuracy are greatly improved; through the four-level early warning mechanism, the whole-process closed-loop management from daily monitoring to emergency disposal is realized; the whole system has self-learning ability, and as the monitoring data accumulates, the model can be continuously optimized through online learning to improve the adaptability and prediction accuracy under different bridges and environments; the sensing system has good durability, the method process is clear, and the sensing system is easy to integrate into the existing bridge health monitoring system to realize automatic and intelligent corrosion management. Compared with the prior art, the application can discover the steel bar corrosion problem earlier and more accurately, the early warning response time is advanced by about 40%, and the prediction accuracy is improved by more than 30%, which provides a scientific basis for the safe operation and life prediction of the bridge, effectively avoids the sudden structural damage caused by corrosion, and has great social and economic value. BRIEF DESCRIPTION OF DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0014] Figure 1 A use flowchart of a bridge steel bar corrosion dynamic monitoring system based on fiber grating sensing provided by the embodiment of the present application is shown in the figure; Figure 2 An improved CNN-LSTM network structure diagram is shown in the figure; Figure 3 An LSTM model structure diagram is shown in the figure. DETAILED DESCRIPTION

[0015] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0016] The purpose of the present application is to provide a bridge steel bar corrosion dynamic monitoring system based on fiber grating sensing, which realizes millimeter-level quantitative diagnosis of steel bar corrosion, significantly improves monitoring accuracy and timeliness, and provides reliable technical support for long-term operation and maintenance of bridge structures.

[0017] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0018] The present application provides a bridge steel bar corrosion dynamic monitoring system based on fiber grating sensing, comprising: a fiber grating sensor array, a fiber demodulator, a data acquisition and communication unit, and a data processing unit, the fiber grating sensor array is arranged on the surface of the steel bar at the key position, for collecting radial expansion pressure, point temperature and macro axial stress, the fiber grating sensor array is connected to the fiber demodulator, the fiber demodulator is connected to the data acquisition and communication unit, the fiber demodulator is used for calculating and obtaining the data collected by the fiber grating sensor array, and sending the data to the data acquisition and communication unit, the data acquisition and communication unit is connected to the data processing unit, the data acquisition and communication unit is used for sending the obtained data to the data processing unit, and the data processing unit is used for performing bridge steel bar corrosion dynamic monitoring according to the obtained data. The key positions include the water level fluctuation area and splash area of the pier, the area below the bridge deck drain and the side of the expansion joint, based on cost effectiveness, without each layout, only the key positions of the bridge most prone to corrosion are selected; The fiber grating sensor array comprises a fiber and a radial expansion pressure sensor, a point temperature sensor and a macro axial stress sensor connected in series on the fiber, and the fiber is connected to the fiber demodulator. The outer part of the optical fiber is provided with a flexible protective sleeve for providing physical protection. FBG P The radial expansion pressure sensor is encapsulated by a flexible material (such as polyurethane), which is highly sensitive to the radial expansion pressure caused by the reinforcement corrosion; FBG T The measuring point temperature sensor is encapsulated by a rigid material (such as a thin steel sheet) and directly pasted on the reinforcement for accurately measuring the measuring point temperature; FBG ε The strain is encapsulated by a rigid material and directly pasted on the reinforcement for monitoring the macro axial strain of the reinforcement; The optical fiber grating sensor array is firmly pasted on the surface of the reinforcement by using high-strength epoxy resin glue.

[0019] The main technical parameters of the optical fiber demodulator adopted by the present application include: 1. Channel number: support connecting multiple trunk optical cables at the same time (such as 4 channels, 8 channels or 16 channels) to meet the distributed monitoring requirements; 2. Scanning frequency: high-frequency scanning capability (such as ≥1Hz) to capture dynamic signals under the action of corrosion development and load; 3. Wavelength accuracy: high accuracy (such as ±1pm) to ensure reliable measurement of small strain and pressure changes; 4. Connection mode: physical connection with the trunk single-mode optical cable led out from the field junction box through the optical fiber flange / interface.

[0020] The data acquisition and communication unit adopted by the present application can be composed of an industrial computer or an embedded industrial computer, connected with the optical fiber demodulator through a network port / serial port, and provided with a 4G / 5G wireless communication module or an Ethernet interface for remote data transmission.

[0021] Since the optical fiber demodulator and the data acquisition and communication unit are relatively mature technologies in the prior art, the present application does not limit them, and only the functions are required.

[0022] As shown in Figure 1 The core of the present application is that the data processing unit performs bridge reinforcement corrosion dynamic monitoring based on the acquired data, which will be described in detail as follows: Step 1: acquiring radial expansion pressure, measuring point temperature and macro axial stress data, and preprocessing the same; Step 2: constructing a bridge reinforcement corrosion dynamic monitoring model; Step 3: constructing a data set; Step 4: training the bridge reinforcement corrosion dynamic monitoring model based on the data set to obtain a trained bridge reinforcement corrosion dynamic monitoring model; Step 5: input the radial expansion pressure, temperature and macroscopic axial stress data to be monitored into the trained bridge reinforcement corrosion dynamic monitoring model to obtain a corrosion indicator; Step 6: calculate the reinforcement corrosion depth based on the corrosion indicator; Step 7: perform dynamic early warning based on the reinforcement corrosion depth.

[0023] In step 1, the radial expansion pressure, temperature and macroscopic axial stress data are obtained and preprocessed, specifically: 1. Data cleaning and outlier processing Based on the existing technology algorithm, invalid and abnormal data points caused by sensor transient failure, signal transmission interference and the like are identified and processed; The invalid and abnormal data points are processed as follows: (1) Invalid value processing: directly eliminate or mark the missing signal (such as zero value or null value caused by demodulator lock loss); (2) Physical range filtering: based on the physical characteristics of the sensor, a reasonable threshold is set, for example, the wavelength drift of a certain FBG should not exceed ±5nm in theory, and data exceeding this range is considered abnormal and is eliminated; (3) Statistical anomaly detection: for more hidden anomalies, the isolation forest or Z-score method is used for detection, for example, the Z-score of each sensor data stream is calculated, and the data points with an absolute value greater than 3 are considered as extreme outliers and are processed; (4) Filling: for a small amount of data points that are eliminated, linear interpolation or the mean value of the previous and subsequent time data is used for filling to maintain the continuity of the time series.

[0024] 2. Data alignment and synchronization Used to ensure that data from different FBG sensors on the same reinforcement is strictly synchronized in time stamp; The optical fiber demodulator provides a unified time stamp for all channels, and during preprocessing, it is necessary to verify and ensure that the three data streams are aligned at each sampling time; if there is a slight misalignment due to data transmission, interpolation resampling is performed according to the time stamp to align all sensor data on the same time grid.

[0025] 3. Signal smoothing and denoising Used to filter out signal jitter caused by high-frequency electronic noise, environmental random vibration and the like, and retain low-frequency trend signals caused by corrosion, temperature, load and other real physical processes; The present application uses existing stable wavelet transform for denoising. Stable wavelet transform is an excellent non-stationary signal processing method, which can effectively separate the noise component and useful component in the signal, and maximally retain the edges and characteristics of the signal during reconstruction, which is superior to traditional low-pass filters; In addition, if the data noise is not serious, a moving average filter or a Savitzky-Golay filter can also be used, which can better preserve the local extreme value characteristics of the signal while smoothing.

[0026] 4. Data standardization The Robust Z-Score standardization is used to eliminate the dimensional differences between different physical quantities (pressure, temperature, strain) and scale the data to a range suitable for deep learning model training, so as to accelerate model convergence.

[0027] In step 2, a bridge reinforcement corrosion dynamic monitoring model is constructed, specifically: The bridge reinforcement corrosion dynamic monitoring model is constructed based on the improved CNN-LSTM network structure; wherein the input of the model is three-dimensional time series data, i.e. radial expansion pressure, temperature of measuring point and macroscopic axial stress data; and the output is a corrosion indicator. As shown in the formula (1), the improved CNN-LSTM network structure includes an input layer, two CNN modules, two LSTM modules and a full connection layer, wherein the input layer is used to receive three-dimensional time series data, and the full connection regression layer is used to map the LSTM output to the final corrosion indicator; the two CNN modules and the two LSTM modules are described in detail: Figure 2 Each CNN module includes a convolution layer, an activation function layer and a pooling layer, wherein the first layer of convolution layer includes 32 filters, and the convolution kernel size is 3; the second layer of convolution layer includes 64 filters, and the convolution kernel size is 3; the design of the two convolution networks can effectively capture the relevant characteristics and improve the learning ability of the features, providing high-quality feature input for the subsequent time series modeling of the LSTM network. The LSTM used is a special recurrent neural network (RNN) designed to capture long-term dependencies in sequence data, effectively solving the gradient vanishing problem of traditional RNN in processing long sequences by introducing memory cells and gating mechanisms, and the relevant formulas are as follows: (1) (2) (3) (4) (5) In the formula, x t , h t－1 are the input and output of the LSTM model at t and t-1 time; c t , c t－1 ​The cell states at time t and t-1, respectively; i t , f t , o t The input gate, the forget gate and the output gate at time t, respectively; σ() is a sigmoid function; and is an element-wise multiplication, and W and b are parameters of the model (different subscripts represent different parameters); The structure of the LSTM model is shown in Figure 3 Two LSTM modules are provided, each having 64 hidden units, and a multi-layer LSTM structure is used to improve the time series modeling effect and enhance the generalization ability of the model.

[0028] In step 3, the data set is constructed, specifically: A plurality of reinforcing bars identical to those in step 1 are made and are cast in the concrete test block; The test block is subjected to an experiment based on a controllable accelerated corrosion environment; During the experiment, the radial expansion pressure, the temperature of the measuring point and the macroscopic axial stress data are collected, and at predetermined, non-equidistant key time points (for example, corrosion 0 hours, 100 hours, 300 hours, 600 hours, …), one or a group of test pieces are taken out for destructive testing to obtain the actual corrosion depth, specifically: The concrete is broken, and the corrosion depth (unit: mm) on the surface of the reinforcing bar is measured directly using a vernier caliper or a laser scanning microscope; After removing the corrosion products on the surface of the reinforcing bar, the mass loss of the reinforcing bar is measured using a precision balance, and then the average cross-sectional loss rate or the corrosion depth of the reinforcing bar is converted through a formula; Next, the corrosion depth is used to label the continuous FBG data, specifically including: 1. Define the reference point: The corrosion depth D0 at the starting time (t0, 0 hours) of the test is defined as 0; The corrosion depth D n at the end time (t n , for example, 1000 hours) of the test is defined as 1; A linear mapping relationship from [D0, D n ] to [0, 1] is established; 2. Linear interpolation to assign true values: For any FBG sampling time t between two destructive testing time points t i and t j , it is assumed that the corrosion depth is linearly increasing; The linear interpolation formula is used to calculate the corrosion indicator true value Y true (t) corresponding to the time t as follows: Y true(t) = (D(t) - D0) / (D n -D0), where D(t) = D i + (t - t i ) (D j -D i ) / (t j -t i ); The entire corrosion process is normalized to a scale of 0 to 1, 0 represents brand new, 1 represents severe corrosion at the end of the test, and each time point in between is assigned a value between 0 and 1 according to its rust progress from the starting point and the ending point; Finally, the data set is obtained.

[0029] In step 4, the bridge reinforcement corrosion dynamic monitoring model is trained based on the data set, specifically: The data set is divided, wherein 70% of the data in the data set is divided into a training set, 15% of the data is divided into a validation set, and 15% of the data is divided into a test set; Based on the training set and the validation set, the IGWO algorithm is used to train the bridge reinforcement corrosion dynamic monitoring model to obtain the optimal hyperparameters; Based on the optimal hyperparameters, the bridge reinforcement corrosion dynamic monitoring model is reinitialized, the training set and the validation set are merged, and the initialized bridge reinforcement corrosion dynamic monitoring model is finally trained based on the merged data set; The finally trained bridge reinforcement corrosion dynamic monitoring model is evaluated based on the test set.

[0030] Based on the training set and the validation set, the IGWO algorithm is used to train the bridge reinforcement corrosion dynamic monitoring model to obtain the optimal hyperparameters, specifically: 1. Define the search space: determine the hyperparameters to be optimized and their ranges, for example: learning rate (Learning Rate), number of LSTM units (Number of LSTM Units), number of CNN filters (Number of CNN Filters), dropout rate (Dropout Rate), batch size (Batch Size); 2. IGWO optimization: Encode a set of hyperparameters as the position of a "gray wolf"; Fitness calculation: for each wolf (each combination of hyperparameters): (1) Instantiate a model: use the current combination of hyperparameters to create a new, untrained improved CNN-LSTM model; (2) Brief training: use the training set to train this model for a small number of rounds; (3) Evaluation: Compute the RMSE of the model on the validation set as the performance indicator, where fitness = -RMSE; The wolf pack updates the position according to the fitness (a, b, d wolf) and searches the next generation; 3, Output: After optimization, output the global optimal hyperparameter combination.

[0031] Here, the IGWO algorithm needs to be introduced, specifically: The grey wolf optimization (GWO) algorithm is generated in the cooperation and competition relationship between grey wolf individuals, which uses the action of grey wolves to search for the optimal solution of the problem. The GWO algorithm process covers multiple links such as grey wolf social grade division, hunting prey, attacking prey, and searching for prey. The IGWO algorithm is improved based on the GWO algorithm, so this place does not introduce the GWO algorithm in detail. In order to solve the problems of lack of population diversity, imbalance between development and exploration, and premature convergence of the GWO algorithm, the dimension learning-based hunting (DLH) search strategy is introduced based on the GWO algorithm, which enhances the balance between local search and global search, and the individual hunting of wolves is learned by neighbors, thereby maintaining diversity.

[0032] In step 6, the steel bar corrosion depth is calculated based on the corrosion indicator factor, specifically: Calculating the steel bar corrosion depth by the corrosion indicator factor does not need to use complex physical formulas that require accurate parameters, but uses a hybrid method of "data-driven as the main, physical mechanism as the framework". Gaussian process regression is a probabilistic non-parametric model that can not only give the predicted value, but also give the uncertainty range (confidence interval) of the prediction, which is crucial for risk assessment; First, data preparation is needed, consistent with the above data set. The data here comes from the laboratory, the difference is that the data here is a mapping pair of corrosion indicator factor-steel bar corrosion depth; Next, Gaussian process regression (GPR) is used to learn the complex nonlinear relationship from the corrosion indicator factor to the steel bar corrosion depth; In order to ensure that the model prediction conforms to the physical law (for example, the corrosion depth cannot be negative, and it is usually monotonically increasing), prior knowledge is introduced in the kernel function selection of GPR, such as using a kernel function that guarantees non-negative output, or adding a monotonicity constraint in the loss function. This step is crucial to ensure that the data-driven model learns within the framework of physical laws and improves the reliability of extrapolation; The laboratory data set is used to train the GPR model; The corrosion indicator factor to be monitored is input into the trained GPR model to obtain the steel bar corrosion depth.

[0033] In step 7, a dynamic early warning is carried out based on the steel bar corrosion depth, specifically: A four-level early warning is constructed, wherein if the steel bar corrosion depth is <0.1mm, it is judged as a first-level early warning, the data processing unit records the steel bar corrosion depth, and no other processing is performed; If 0.1mm≤steel bar corrosion depth<0.3mm, it is judged as a second-level early warning, the data processing unit sends it to the monitoring platform for the staff to check, and the monitoring interval time of the steel bar is shortened to improve the monitoring frequency; If 0.3mm≤steel bar corrosion depth<0.5mm, it is judged as a third-level early warning, the data processing unit sends it to the monitoring platform, the monitoring platform generates an operation and maintenance report and sends it to the maintenance management department for maintenance; If the steel bar corrosion depth is ≥0.5mm, it is judged as a fourth-level early warning, the data processing unit sends it to the monitoring platform, and the monitoring platform notifies the corresponding department to take traffic control and immediately reinforce and repair the treatment.

[0034] In order to verify the effectiveness and advancement of the present application, the bridge pier monitoring of a certain coastal large-scale sea-crossing bridge is taken as an implementation case. The bridge has been in operation for 8 years and is in a marine corrosion environment. The pier part is seriously eroded by chloride ions and is a high-risk area of corrosion. We selected a total of 20 steel bars in the water level change area and the splash area of 4 key piers (Nos. P3-P6) of the bridge, 5 in each pier, at different height positions, to comprehensively monitor the corrosion development; 1. System deployment and data collection The fiber grating sensor array is arranged on the steel bar, and is led to the monitoring station house at the top of the pier through armored cable. The monitoring system is continuously operated for 12 months, the collection frequency is set to 2Hz, about 63 million effective data records are obtained, the fiber demodulator adopts a 16-channel type, the wavelength accuracy is ±1pm, and the monitoring requirements are fully met; 2. Data processing and model training Firstly, the original data collected is preprocessed, including data cleaning, abnormal value processing, signal alignment, stable wavelet transform denoising and Robust Z-Score standardization. During the 12-month monitoring period, an accelerated corrosion test was simultaneously carried out in the laboratory, 30 groups of same test pieces were prepared, destructive detection was carried out at 5 key time points (0, 400, 800, 1200, 1600 hours), actual corrosion depth data was obtained, and a corrosion indicator true value data set was established; The training process of the improved CNN-LSTM model adopts the IGWO algorithm for hyperparameter optimization. The optimized parameters include learning rate, LSTM hidden unit number, CNN filter number, dropout rate, and batch size. After 200 iterations, the IGWO algorithm converges, and the optimal hyperparameter combination is found. The final model performs excellently on the test set, and all indicators are significantly better than those of the comparative models. 3. Performance comparison and analysis To objectively evaluate the technical effect of the present application, the method of the present application is compared with traditional LSTM, GRU, non-optimized CNN-LSTM, and the commonly used half-cell potential method in engineering. The comparison results are shown in Tables 1, 2, and 3.

[0035] From the comparison results, it can be seen that the method of the present application is significantly better than the traditional methods in terms of prediction accuracy, early warning timeliness, and reliability, which is specifically shown in the following aspects: 1. Prediction accuracy is improved: compared with the traditional LSTM, the RMSE is reduced from 0.048 mm to 0.019 mm, and the accuracy is improved by about 60%; compared with the non-optimized CNN-LSTM, the accuracy is improved by about 38%; 2. Early warning response is advanced: the method of the present application can effectively warn when the corrosion depth reaches 0.25 mm, which is about 30% earlier than the traditional method, leaving sufficient time for maintenance decision-making; 3. False alarm rate is greatly reduced: through IGWO optimization and multi-source signal decoupling, the false alarm rate is reduced from 23.5% of the traditional method to 4.3%, greatly reducing unnecessary maintenance costs.

[0036] In practical application, the present application system successfully monitors the abnormal corrosion development of P4 bridge pier No. 3 steel bar at the 9th month. The system triggers a level 3 warning on the 12th day of the 9th month, showing that the steel bar corrosion depth reaches 0.32 mm. After on-site verification, it is confirmed that the concrete cover at this position has micro cracks, which is highly consistent with the prediction result. The maintenance department timely takes surface sealing and corrosion prevention treatment, effectively controls the corrosion development, and avoids possible structural damage. In contrast, the same area monitored by the half-cell potential method detects abnormalities until the 11th month, at which time the corrosion depth has reached 0.45 mm, significantly increasing the treatment cost and difficulty.

[0037] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be mutually referred to.

[0038] The principles and implementation manners of the present application are described by using specific examples in the present application, and the above examples are only used to help understand the method of the present application and its core idea; meanwhile, for the general technical personnel in the art, the specific implementation manners and application ranges will be changed according to the idea of the present application. In conclusion, the content of the present specification should not be understood as the limitation of the present application.

Claims

1. A dynamic monitoring system for steel reinforcement corrosion in bridges based on fiber optic grating sensing, characterized in that, include: The system comprises a fiber Bragg grating sensor array, a fiber optic demodulator, a data acquisition and communication unit, and a data processing unit. The fiber Bragg grating sensor array is installed on the surface of the reinforcing steel at key locations to collect radial expansion pressure, measuring point temperature, and macroscopic axial stress. The fiber Bragg grating sensor array is connected to the fiber optic demodulator, which is connected to the data acquisition and communication unit. The fiber optic demodulator is used to process the data collected by the fiber Bragg grating sensor array and send it to the data acquisition and communication unit. The data acquisition and communication unit is connected to the data processing unit, which is used to send the acquired data to the data processing unit. The data processing unit is used to perform dynamic monitoring of bridge steel reinforcement corrosion based on the acquired data. The key areas include the water level fluctuation zone and splash zone of the bridge pier, the area below the drainage outlet on the bridge deck, and the area around the expansion joint.

2. The bridge steel reinforcement corrosion dynamic monitoring system based on fiber optic grating sensing according to claim 1, characterized in that, The fiber optic grating sensor array includes an optical fiber and a radial expansion pressure sensor, a measuring point temperature sensor, and a macroscopic axial stress sensor connected in series on the optical fiber. The optical fiber is connected to the optical fiber demodulator.

3. The bridge steel reinforcement corrosion dynamic monitoring system based on fiber optic grating sensing according to claim 2, characterized in that, The data processing unit is used to perform dynamic monitoring of bridge steel reinforcement corrosion based on the acquired data, specifically: Step 1: Obtain radial expansion pressure, measuring point temperature, and macroscopic axial stress data, and preprocess them; Step 2: Construct a dynamic monitoring model for steel reinforcement corrosion in bridges; Step 3: Build the dataset; Step 4: Train the dynamic monitoring model for bridge steel reinforcement corrosion based on the dataset to obtain the trained dynamic monitoring model for bridge steel reinforcement corrosion. Step 5: Input the radial expansion pressure, measuring point temperature and macroscopic axial stress data to be monitored into the trained bridge steel reinforcement corrosion dynamic monitoring model to obtain corrosion indicator factors. Step 6: Calculate the corrosion depth of the reinforcing steel based on the corrosion indicator factor; Step 7: Dynamic early warning based on the depth of steel corrosion.

4. The bridge steel reinforcement corrosion dynamic monitoring system based on fiber optic grating sensing according to claim 3, characterized in that, In step 2, a dynamic monitoring model for steel reinforcement corrosion of the bridge is constructed, specifically as follows: A dynamic monitoring model for steel bar corrosion of bridges is constructed based on an improved CNN-LSTM network structure. The specific structure of the improved CNN-LSTM network structure includes an input layer, two CNN modules, two LSTM modules, and a fully connected layer. Each of the two CNN modules includes a convolutional layer and a pooling layer. The convolutional layer of the first CNN module includes 32 filters and a kernel size of 3, while the convolutional layer of the second CNN module includes 64 filters and a kernel size of 3. Each of the LSTM modules has 64 hidden units.

5. The bridge steel reinforcement corrosion dynamic monitoring system based on fiber optic grating sensing according to claim 4, characterized in that, Step 3 involves constructing the dataset, specifically as follows: Make multiple steel bars identical to those in step 1 and cast them into concrete test blocks; Experiments were conducted on the test blocks based on a controllable accelerated corrosion environment; During the experiment, radial expansion pressure, measuring point temperature, macroscopic axial stress data, and actual corrosion depth were collected. The radial expansion pressure, measuring point temperature, and macroscopic axial stress data for each time period are aligned with the actual corrosion depth measured at the corresponding time. Each segment of radial expansion pressure, measuring point temperature, and macroscopic axial stress data is assigned a corrosion indicator factor as a label. The dataset is constructed based on the aligned data.

6. The bridge steel reinforcement corrosion dynamic monitoring system based on fiber optic grating sensing according to claim 5, characterized in that, Step 4: Train the dynamic monitoring model for bridge steel reinforcement corrosion based on the dataset, specifically as follows: The dataset is divided into three parts: 70% as the training set, 15% as the validation set, and 15% as the test set. Based on the training and validation sets, the IGWO algorithm was used to train the dynamic monitoring model for steel corrosion of bridges to obtain the optimal hyperparameters; The dynamic monitoring model for bridge steel corrosion was re-initialized based on the optimal hyperparameters. The training set and validation set were merged, and the initialized dynamic monitoring model for bridge steel corrosion was finally trained based on the merged dataset. The final trained dynamic monitoring model for bridge steel corrosion was evaluated based on the test set.

7. The bridge steel reinforcement corrosion dynamic monitoring system based on fiber optic grating sensing according to claim 6, characterized in that, In step 6, the corrosion depth of the reinforcing steel is calculated based on the corrosion indicator factor, specifically as follows: Construct a physical-guided Gaussian process regression model; It is trained based on a pre-defined rust indicator factor - steel bar rust depth dataset; The corrosion indicator factor to be detected is input into the trained Gaussian process regression model to obtain the corrosion depth of the steel bars.

8. The dynamic monitoring system for bridge steel reinforcement corrosion based on fiber optic grating sensing according to claim 7, characterized in that, In step 7, dynamic early warning is provided based on the depth of steel reinforcement corrosion, specifically as follows: A four-level early warning system is established. If the depth of steel bar corrosion is less than 0.1 mm, it is judged as a level one early warning. The data processing unit records the depth of steel bar corrosion without performing any other processing. If the corrosion depth of the steel bar is less than 0.3 mm and 0.1 mm, it is judged as a level two warning. The data processing unit sends it to the monitoring platform for staff to view and shortens the monitoring interval of this steel bar to increase the monitoring frequency. If the steel bar corrosion depth is less than 0.5mm and less than 0.3mm, it is judged as a level three warning. The data processing unit sends it to the monitoring platform, the monitoring platform generates an operation and maintenance report, and sends it to the maintenance management department for maintenance. If the depth of steel bar corrosion is ≥0.5mm, it is judged as a level four warning. The data processing unit sends it to the monitoring platform, and the monitoring platform notifies the relevant departments to take traffic control measures and immediately carry out reinforcement and repair.

Images