A method for monitoring concrete cracks based on an array piezoelectric single crystal sensor

Abstract

This invention discloses a concrete crack monitoring method based on an array of piezoelectric single-crystal sensors. Leveraging the excellent electromechanical coupling properties of piezoelectric single-crystal materials, an array-type sensor deployment system is constructed to achieve highly sensitive, multi-node synchronous acquisition of stress wave signals during concrete crack formation. Combining the stress wave arrival time difference and propagation energy attenuation characteristics, the crack location is calculated. Furthermore, a multi-source information fusion mechanism is introduced to dynamically correct and optimize the crack location results, achieving stable and reliable concrete crack positioning.

Description

Technical Field

[0001] This invention relates to the field of concrete, and more particularly to a method for monitoring concrete cracks based on an array of piezoelectric single-crystal sensors. Background Technology

[0002] Cracks inevitably develop in concrete structures during service, and their formation and propagation are usually accompanied by stress redistribution and transient stress wave release. Current technologies for monitoring concrete cracks mainly focus on strain monitoring, wave signal monitoring, and the deployment of multiple sensors.

[0003] (1) Crack monitoring technology based on strain measurement: In existing engineering practice, resistance strain gauges or fiber optic sensors are often deployed on or inside concrete structures to monitor the strain changes of concrete components over a long period of time. When cracks occur or propagate inside the concrete, the strain field will show a local abrupt change, which can be used to infer the existence and development trend of cracks. This type of technology usually focuses on macroscopic deformation response and is more suitable for monitoring cracks after they have developed to a certain scale. It has limited sensitivity to the early initiation stage of cracks.

[0004] (2) Crack monitoring technology based on piezoelectric materials: Utilizing the positive piezoelectric effect of piezoelectric materials to convert stress waves or vibration signals inside concrete into electrical signals is an important research direction in crack monitoring in recent years. In existing technologies, commonly used piezoelectric materials include piezoelectric ceramics and piezoelectric films, and sensors are mostly arranged in concrete structures by attachment or embedding. When cracks occur or propagate, the stress waves released inside the concrete are captured by the piezoelectric sensor, thereby achieving passive monitoring of crack events.

[0005] (3) Crack location technology based on wave propagation characteristics: In some studies, multiple sensors are deployed on the structure to collect stress wave signals when cracks are generated. The correspondence between crack location and signal characteristics is established by utilizing the time difference of wave arrival or the signal amplitude attenuation characteristics at different sensors, thereby realizing the inversion of crack location. This type of method usually relies on a limited number of sensor nodes and assumes that the wave velocity inside the concrete is uniform or approximately uniform.

[0006] Although the aforementioned existing technologies have achieved the monitoring of concrete cracks to a certain extent, they still have the following objective drawbacks in engineering applications, which are difficult to solve simultaneously under current technological conditions:

[0007] (1) Insufficient response capability to weak signals in the early stage of cracks: resistance strain gauges and fiber optic sensors mainly reflect the strain changes of concrete and are not sensitive to transient stress waves generated in the early stage of cracks; while existing piezoelectric ceramics or piezoelectric thin film materials have limited piezoelectric coefficients and electromechanical coupling performance. Under the background of complex engineering noise, the weak stress wave signals released in the early stage of cracks are easily submerged and difficult to reliably identify.

[0008] (2) Insufficient anti-interference capability of single positioning information source: Existing technologies usually locate cracks based only on time difference information or amplitude information. When there is noise interference in the signal or the quality of some sensor signals deteriorates, the positioning results are prone to large fluctuations, which makes it difficult to meet the requirements of stability and reliability for long-term monitoring of engineering structures.

[0009] (3) Lack of systematic fusion mechanism for multi-source information: Although some studies have attempted to introduce multiple signal features, they mostly use simple weighting or empirical correction methods, lacking dynamic information fusion methods for the continuous evolution of cracks, making it difficult to achieve continuous and smooth updates of crack positions on a time scale.

[0010] Existing concrete crack monitoring technologies have significant shortcomings in terms of sensor material performance, sensor deployment methods, and crack location information processing methods. On the one hand, traditional sensor materials have limited response capabilities to the weak stress wave signals released in the early stages of cracks, making it difficult to meet the needs of early crack monitoring. On the other hand, existing sensors are mostly deployed in a single-point or discrete manner, lacking a systematic array design, which limits the accuracy and stability of crack spatial location. At the same time, existing crack location methods usually rely on a single physical parameter, have insufficient anti-interference capabilities, and are difficult to obtain reliable results under the conditions of non-homogeneous concrete materials and complex working conditions. Summary of the Invention

[0011] The purpose of this invention is to provide a method for monitoring concrete cracks based on an array-type piezoelectric single-crystal sensor, which solves the aforementioned technical problems pointed out in the prior art.

[0012] This invention provides a method for monitoring concrete cracks based on an array-type piezoelectric single-crystal sensor, comprising the following steps:

[0013] Multiple piezoelectric single crystal sensors are arranged at predetermined intervals inside or on the surface of a concrete structure to form an array-type sensing system.

[0014] The array-type sensing system synchronously collects stress wave signals released during the generation or expansion of internal cracks in concrete in real time.

[0015] The initial estimated location of the crack is calculated based on the time difference between the stress wave signals received by each piezoelectric single crystal sensor in the array-type sensing system.

[0016] By combining stress wave signals with energy attenuation analysis, the initial estimated location of the crack is corrected to obtain the corrected estimated location.

[0017] The initial estimated location and the corrected estimated location are fused using a prediction-correction multi-source information fusion method to obtain the target crack location.

[0018] Preferably, the stress wave signal includes the signal arrival time, signal amplitude, and waveform characteristics.

[0019] Preferably, the initial estimated location of the crack is calculated based on the time difference between the stress wave signals received by each piezoelectric single-crystal sensor in the array-type sensing system, including the following steps:

[0020] Extract the signal arrival time of each piezoelectric single crystal sensor in the array sensing system that receives the stress wave signal, and calculate the signal arrival time difference between any two piezoelectric single crystal sensors based on the signal arrival time.

[0021] Based on the signal arrival time difference between any two piezoelectric single crystal sensors and the spatial relationship between each piezoelectric single crystal sensor, a localization model between the crack location and the time difference is established.

[0022] By solving the positioning model, the initial estimated location of the crack is obtained;

[0023] Solve the time-of-arrival mathematical model to obtain the initial estimated location of the crack.

[0024] Preferably, the positioning model is established based on the correspondence between the spatial distance difference between the crack location and each piezoelectric single crystal sensor and the signal arrival time difference.

[0025] Preferably, the initial estimated location of the crack is corrected by combining stress wave signals with energy attenuation analysis to obtain the corrected estimated location, including the following steps:

[0026] Extract the signal amplitude received by each piezoelectric single crystal sensor from the stress wave signal;

[0027] Starting from the initial estimated location of the crack, the corrected estimated location is obtained by combining the pre-established energy attenuation model with the analysis of the amplitude of each signal.

[0028] Preferably, the pre-established energy attenuation model is used to describe the functional relationship between the amplitude of the crack source signal and the distance between the crack location and each piezoelectric single crystal sensor. This functional relationship takes into account both geometric diffusion attenuation factors and material absorption attenuation factors. The initial estimation result of the crack location is corrected by this energy attenuation relationship model.

[0029] Preferably, starting from the initial estimated location of the crack, the crack location is updated through iterative optimization, so that the difference between the measured signal amplitude of each piezoelectric single crystal sensor and the theoretical signal amplitude obtained based on the energy attenuation relationship model is gradually reduced, thereby obtaining the corrected estimated location.

[0030] Preferably, the initial estimated position and the corrected estimated position are fused using a prediction-correction multi-source information fusion method to obtain the target crack position. The multi-source information fusion method is based on the recursive idea of ​​Kalman filtering and includes the following steps: constructing the observation data at the current moment based on the initial estimated position and the corrected estimated position to characterize the multi-source localization result of the crack position at the current moment.

[0031] Based on the target crack location estimation result of the previous moment, the crack location at the current moment is predicted by using the constant location assumption, and the predicted location at the current moment is obtained.

[0032] The predicted position at the current moment is compared with the observed data, and the predicted position is corrected and updated based on the difference between the two to obtain the updated target crack position at the current moment.

[0033] The updated target crack location is used as the input for the crack location prediction at the next time step, enabling a continuous and smooth recursive estimation of the crack location in the time dimension.

[0034] Preferably, the target crack location is a two-dimensional spatial coordinate, which includes the horizontal and vertical coordinates of the target crack location.

[0035] Preferably, while obtaining the updated target crack location at the current moment, the parameters used to characterize the uncertainty of crack location estimation are updated simultaneously, so as to adaptively adjust the weight relationship between the prediction results and the observation data in subsequent moments, thereby improving the stability and reliability of crack location results under noise interference and parameter uncertainty conditions.

[0036] Compared with the prior art, the embodiments of the present invention have at least the following technical advantages:

[0037] Analysis of the concrete crack monitoring method based on arrayed piezoelectric single-crystal sensors provided by this invention reveals that, in practical applications, based on the excellent electromechanical coupling performance of piezoelectric single-crystal materials, an arrayed sensor deployment system is constructed to achieve highly sensitive, multi-node synchronous acquisition of stress wave signals during the generation of concrete cracks. Furthermore, by combining the stress wave arrival time difference and propagation energy attenuation characteristics, the crack location is calculated inversely. A multi-source information fusion mechanism is further introduced to dynamically correct and optimize the crack location results, thereby achieving stable and reliable crack location without relying on high-precision prior material parameters. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the main process of a concrete crack monitoring method based on an array of piezoelectric single-crystal sensors. Detailed Implementation

[0039] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] The present invention will now be described in further detail with reference to specific embodiments and accompanying drawings.

[0041] Example 1

[0042] like Figure 1 As shown, Embodiment 1 of the present invention provides a method for monitoring concrete cracks based on an array-type piezoelectric single-crystal sensor, comprising the following steps:

[0043] Step S10: Arrange multiple piezoelectric single crystal sensors at predetermined intervals inside or on the surface of the concrete structure to form an array-type sensing system;

[0044] It should be noted that the piezoelectric single-crystal material used as the sensing element in the above embodiments of this application is because, compared with traditional piezoelectric ceramics, piezoelectric single-crystal materials have higher piezoelectric coefficients and electromechanical coupling coefficients, which can significantly improve the response sensitivity and signal-to-noise ratio to stress wave signals. The array layout allows each sensor to form a defined geometric relationship in space, providing the necessary spatial information basis for subsequent crack location inversion. Specifically, a grid layout can be used in the planar region: if the stress field distribution of the structure is relatively uniform, an orthogonal equidistant grid can be used to ensure spatial sampling consistency; if the local stress gradient is significant (such as holes, slots, stress concentration edges), an adaptive grid can be used, appropriately increasing the number of sensor nodes in high-stress areas. In the thickness direction, a layered layout can be used to distinguish the spatial depth characteristics of the crack. This array layout not only significantly improves the overall acquisition capability of stress wave signals, but also provides the necessary spatial constraints for crack location inversion, which is the structural basis for realizing crack localization.

[0045] Step S20: The array-type sensing system synchronously collects stress wave signals released during the generation or propagation of internal cracks in concrete in real time.

[0046] The stress wave signal includes the signal arrival time, signal amplitude (the signal amplitude is the maximum amplitude or root mean square amplitude of the stress wave signal received by each piezoelectric single crystal sensor), and waveform characteristics;

[0047] It should be noted that in the above embodiments of this application, when cracks are generated or expand inside the concrete, the transient stress wave released at the crack tip is synchronously received by piezoelectric single-crystal sensors deployed at different spatial locations, and the mechanical stress signal is converted into an electrical signal through the positive piezoelectric effect. Multi-channel time-domain data is acquired by the synchronous acquisition device, providing a data basis for subsequent crack location analysis.

[0048] Step S30: Calculate the initial estimated location of the crack based on the time difference between the stress wave signals received by each piezoelectric single crystal sensor in the array sensing system.

[0049] Step S40: By combining stress wave signals with energy attenuation analysis, the initial estimated location of the crack is corrected to obtain the corrected estimated location;

[0050] Step S50: The initial estimated position and the corrected estimated position are fused using a Kalman filter to obtain the target crack position. This involves fusing the initial estimated position and the corrected estimated position using a prediction-correction multi-source information fusion method to obtain the target crack position.

[0051] It should be noted that the above embodiments of this application are based on the excellent electromechanical coupling performance of piezoelectric single crystal materials. By constructing an array-type sensor deployment system, highly sensitive and multi-node synchronous acquisition of stress wave signals during the generation of concrete cracks is achieved. Combined with the stress wave arrival time difference and propagation energy attenuation characteristics, the crack location is inverted and calculated. Furthermore, a multi-source information fusion mechanism is introduced to dynamically correct and optimize the crack location results, thereby achieving stable and reliable location of concrete cracks without relying on high-precision prior material parameters.

[0052] Specifically, in step S30, the initial estimated location of the crack is calculated based on the time difference between the stress wave signals received by each piezoelectric single-crystal sensor in the array sensing system, including the following steps:

[0053] Step S31: Extract the signal arrival time of the stress wave signal received by each piezoelectric single crystal sensor in the array sensing system, and calculate the signal arrival time difference between any two piezoelectric single crystal sensors based on the signal arrival time.

[0054] Step S32: Establish a location model between crack location and time difference based on the signal arrival time difference between any two piezoelectric single crystal sensors, the spatial coordinates of each piezoelectric single crystal sensor, and the stress wave propagation speed;

[0055] Step S33: Solve the positioning model to obtain the initial estimated location of the crack;

[0056] The Time Difference of Arrival (TDOA) positioning model is a calculation formula model, which is expressed as follows:

[0057] ;

[0058] In the formula, The x-coordinate represents the initial estimated location of the crack. The ordinate represents the initial estimated location of the crack. Let x be the abscissa of the spatial position coordinates of the first piezoelectric single-crystal sensor among any two piezoelectric single-crystal sensors. Let be the ordinate of the spatial position coordinates of the first piezoelectric single-crystal sensor among any two piezoelectric single-crystal sensors. Let x be the abscissa of the spatial position coordinates of the second piezoelectric single-crystal sensor out of any two piezoelectric single-crystal sensors. Let be the ordinate of the spatial position coordinates of the second piezoelectric single-crystal sensor among any two piezoelectric single-crystal sensors. For the propagation speed of stress waves, This represents the signal arrival time difference between any two piezoelectric single-crystal sensors;

[0059] It should be noted that in the embodiments described above, when cracks occur inside the concrete, the propagation of these cracks releases transient stress wave signals. Sensors at different locations in the array experience time differences due to their varying distances. The mathematical principle behind this is the assumption that the signal source (crack) is located at coordinates... The two sensors are located at... and The speed of signal propagation is Then the time difference The calculation formula of the above-mentioned time-of-arrival mathematical model satisfies the time difference of three or more sensors. An equation system can be established to solve for the signal source position (i.e., the initial estimated position of the crack position). The solution of the time-of-arrival mathematical model in the above-mentioned embodiment of the application yields the initial estimated position of the crack position with high positioning accuracy, but it is highly dependent on wave velocity. Therefore, the embodiment of the application takes into account the energy attenuation of stress waves during propagation in concrete, and further establishes an energy attenuation model between the signal amplitude and the distance between the crack and the sensor. By analyzing the signal amplitude received by each sensor and combining the energy attenuation law, the initial estimation result of the crack position is corrected, thereby improving the positioning accuracy. For details, please refer to the above-mentioned step S40 and the subsequent steps S41 to S42.

[0060] Specifically, in step S40, the initial estimated location of the crack is corrected by combining the stress wave signal with energy attenuation analysis to obtain the corrected estimated location, including the following steps:

[0061] Step S41: Extract the signal amplitude received by each piezoelectric single crystal sensor from the stress wave signal;

[0062] Step S42: Starting from the initial estimated location of the crack, the corrected estimated location is calculated by combining the pre-established energy attenuation model with the amplitude analysis of each signal.

[0063] The pre-established energy decay model is a calculation formula model, expressed as follows:

[0064] ;

[0065] in, In the above formula, Let be the theoretical signal amplitude of the i-th piezoelectric single-crystal sensor; This represents the initial signal amplitude at the crack source. It is a natural constant. It is an exponentially decaying term with the natural constant as its base. This is the distance between the corrected estimated position and the spatial coordinates of the i-th piezoelectric single-crystal sensor. The material attenuation coefficient, The geometric attenuation coefficient, The x-coordinate of the corrected estimated position. The ordinate of the corrected estimated position. Let x be the abscissa of the spatial position coordinates of the i-th piezoelectric single crystal sensor. Let y be the ordinate of the spatial position coordinates of the i-th piezoelectric single crystal sensor;

[0066] Starting from the initial estimated location of the crack, the corrected estimated location is obtained by combining the pre-established energy attenuation model with the analysis and calculation of various signal amplitudes. This means that the initial estimated location of the crack is used as the starting point, and the signal amplitude (i.e., the signal amplitudes received by each piezoelectric single-crystal sensor extracted from the stress wave signal) and the theoretical signal amplitude (i.e., the amplitudes in the energy attenuation model) are calculated. The objective is to minimize the sum of squared differences between the two estimates. A gradient optimization algorithm is used (this gradient optimization process is common knowledge to those skilled in the art and will not be described in detail here). At the same time, the corrected estimated position is optimized (i.e., the above-mentioned position). , )) and the initial signal amplitude at the crack source (i.e., the above) ), to obtain the corrected estimated position;

[0067] It should be noted that, in the above embodiments of this application, the energy decay model is based on the energy decay formula. (in, As initial energy, For the distance of transmission The energy after decay, the attenuation coefficient '(unit: Based on the energy absorption and geometric diffusion effect of materials, the actual signal amplitude measured by each piezoelectric single crystal sensor (i.e., the signal amplitude received by each piezoelectric single crystal sensor extracted from the stress wave signal) and the energy attenuation formula are related. Proportional, and, to better suit the complexity of concrete, introduced For the material attenuation coefficient and Geometric attenuation coefficient ( and All of them were pre-calibrated by conducting active excitation experiments on the same type of concrete specimens (which will not be repeated in this application). By describing the amplitude attenuation law of the crack source signal in space, the initial estimated location of the crack is corrected from the physical dimension to obtain the corrected estimated location.

[0068] Specifically, in step S50, the initial estimated position and the corrected estimated position are fused using a Kalman filter to obtain the target crack position, including the following steps:

[0069] Step S51: Construct the observation vector at the current time k based on the initial estimated position and the corrected estimated position. ;

[0070] Step S52: Read the previous target crack location state estimate from the previous time step (i.e., time step k-1). and the estimation error covariance matrix of the previous target crack location state estimation The predicted state value at time k is calculated based on the constant position model. Read the preset process noise covariance matrix Q, and calculate the state prediction error covariance matrix at the current time k based on the estimation error covariance matrix of the previous target crack position state estimation and the process noise covariance matrix Q. ;

[0071] Step S53: Read the preset observation matrix H and observation noise covariance matrix R, and predict the error covariance matrix based on the state at the current time k. Calculate the Kalman gain matrix at time k using the observation matrix H and the observation noise covariance matrix R. ;

[0072] Step S54: Based on the observation vector at the current time k The predicted state value at time k. and the Kalman gain matrix at the current time k Calculate the observation residuals Using the Kalman gain matrix The state correction is obtained by weighting the observation residuals. Add the state correction value to the state prediction value to obtain the updated target crack location state estimation vector at the current time (time k). The target crack location state estimation vector It contains two elements: the x-coordinate of the target crack location and the y-coordinate of the target crack location.

[0073] Step S55: Based on the Kalman gain matrix State prediction error covariance matrix And the identity matrix I is used to calculate the estimated error covariance matrix updated at time k. ;

[0074] Then, the target crack location state estimation vector and the updated estimated error covariance matrix Store it as input data for step S52 at the next time k+1.

[0075] It should be noted that in the above embodiments of this application, the Kalman filter fusion in the above embodiments of this application performs prior prediction based on a constant position model, constructs an observation vector by combining the initial estimated position and the corrected crack position, and then uses Kalman gain to adaptively weight the prediction residual. Through iterative updates of covariance, the filter can continuously adjust the uncertainty of state estimation and maintain convergence and robustness in engineering application scenarios such as noise disturbance, medium heterogeneity and environmental wave velocity drift, so as to achieve stable, high-confidence and continuous estimation of crack position.

[0076] First, in step S51, the observation vector at the current time is constructed. Clearly integrating the two types of information, "initial estimation results of crack location based on time difference of arrival" and "crack location results corrected based on energy attenuation model", and placing them in a single vector, provides a complete data foundation for calculating the difference between the predicted and actual values ​​(i.e., observation residuals) in subsequent steps.

[0077] Then, in step S52, the state prediction value and the covariance matrix of the prediction error are calculated as the prediction processing of the Kalman filter. The input is the system's optimal estimate of the state at the previous time step. and its uncertainty measure Constant position model ( This is based on the reasonable assumption that the location of the crack changes very little within an extremely short sampling interval. The process noise covariance matrix Q is a pre-defined small diagonal matrix (e.g., Its non-zero elements are used to quantify the imperfections of the constant-position model or the possible small, unpredictable variations in the crack location. Q is added to the estimation error covariance matrix of the previous time step. Up, get This reflects the additional state uncertainties introduced due to the passage of time and model simplification;

[0078] Further, in step S53, the Kalman gain matrix is ​​calculated. This is the core intermediate variable calculated in this step, which is an adaptive weight matrix. Its calculation formula incorporates the uncertainties from the prediction stage. The observation noise covariance matrix R is usually preset as a block diagonal matrix, such as... ,in Determined by the time delay estimation error, Due to the amplitude measurement error, the observation matrix H is fixed in this scheme. Its function is to linearly map the two-dimensional state vector (crack coordinates) to the four-dimensional observation space. The physical meaning is that both observation methods are direct measurements of the same state (x coordinates and y coordinates).

[0079] Furthermore, in step S54, the state estimation is updated to obtain the final crack location (i.e., the target crack location). This is an update operation of the Kalman filter. Specifically, the observation residuals are first calculated. This represents the deviation between the current actual observation and the observation predicted based on historical data. Then, the Kalman gain matrix calculated in step S53 is used. The residual is weighted, and the Kalman gain matrix is ​​obtained. The dynamic process determines what percentage of the observed residuals should be used to correct the predicted values. Finally, the weighted correction is added to the state predicted values. The above yields the optimal state estimate for the current moment after information fusion optimization, i.e., the updated target crack location state estimate vector. ;

[0080] The final crack location state estimation vector output by step S54 It is the optimal estimate of the system state (i.e., crack location) at this moment by Kalman filtering. The structure of this vector contains the x-coordinate and y-coordinate of the target crack location. In specific implementations, such as in software code or hardware logic, the values ​​of the x-coordinate and y-coordinate of the target crack location can be directly read from the corresponding memory address or register of this vector for display, storage or triggering of early warning, without the need for additional coordinate extraction or format conversion calculations.

[0081] Finally, in step S55, the estimated error covariance matrix is ​​updated. After the state update, the measure of the uncertainty in the state estimate must be updated synchronously using the formula... Calculated This reflects the estimation of the final crack location and state after incorporating new observational information. An increase in confidence (i.e., a reduction in uncertainty) in the updated This will be fed back into the next round of recursion, step S52, thus forming a continuous adaptive estimation closed loop;

[0082] For example, before the initial calculation (k=1) at system startup, initialization is required. and For example, it can be Set the coordinates of the center point of the monitoring area, and Set as a large diagonal matrix (e.g.) This indicates that the initial location of the crack is unknown and highly uncertain. Subsequently, whenever a new result from steps S30 and S40 is generated (i.e., a new crack event is detected), steps S51 to S55 are executed sequentially. It is assumed that at some moment, due to signal interference, the initial estimated location of the crack in step S30 has a large error, i.e., its corresponding observation noise. If a larger value is assigned during the preset, then the calculation is performed in step S53. At that time, the confidence weight of the initial estimated location observation component of the crack location will be automatically reduced. In the state update of step S54, the final result... It will rely more on the energy decay results of step S40 to achieve suppression of unreliable information and robust fusion.

[0083] It should also be noted that in the above embodiments of this application, the uncertainty of the wave velocity is not affected by the concrete wave velocity parameters that are known in advance (the concrete wave velocity parameters are the stress wave propagation speed that appears in step S32 above), but by the coordinated constraint of array redundancy information and multi-source positioning results.

[0084] In the array deployment, one or more pairs of sensors with known geometric distances can be selected as reference channels. By comparing the arrival time difference of the same crack event on different sensors, the apparent propagation speed corresponding to the current event can be deduced in real time (the apparent propagation speed is also the stress wave propagation speed mentioned above).

[0085] The apparent propagation velocity is used to correct the TDOA location calculation for the current crack event, rather than as a global fixed parameter, thereby achieving an adaptive response to changes in wave velocity (which is also the stress wave propagation velocity mentioned above).

[0086] In the process of multi-source information fusion and location update, the embodiments of this application do not treat wave velocity as a state variable that must be explicitly modeled and recursively applied. Instead, the uncertainty of wave velocity is reflected through the uncertainty of the observation results. Specifically, the positioning results based on the time difference of arrival have corresponding uncertainty weights in the fusion process, while the positioning results based on energy attenuation characteristics provide distance constraints that are independent of wave velocity, thereby compensating for wave velocity errors.

[0087] Based on the excellent electromechanical coupling properties of piezoelectric single crystal materials, the above-described embodiments of this application construct an array-type piezoelectric single crystal sensor deployment system with a defined spatial geometric relationship. By synchronously acquiring stress wave signals released during the generation of concrete cracks through multiple nodes, and combining various physical feature information for collaborative positioning, the stable and reliable determination of the location of concrete cracks can be achieved without relying on high-precision prior material parameters.

[0088] In this embodiment, piezoelectric single-crystal materials are used as sensing elements, and multiple piezoelectric single-crystal sensors are arranged at predetermined intervals and in an array format inside or on the surface of the concrete structure, so that each sensor forms a clear and describable geometric relationship in space. Specifically, a grid layout can be used in planar areas: if the stress field distribution of the structure is relatively uniform, an orthogonal equidistant grid can be used to ensure spatial sampling consistency; if the local stress gradient is significant (such as holes, slots, stress concentration edges), an adaptive grid can be used, and sensor nodes can be appropriately increased in high-stress areas. In the thickness direction, a layered layout can be used to distinguish the spatial depth characteristics of cracks. This array-style layout not only significantly improves the overall acquisition capability of stress wave signals, but also provides the necessary spatial constraints for crack location inversion, which is the structural basis for realizing crack localization.

[0089] On the other hand, when cracks form or propagate within the concrete, the transient stress waves released at the crack tips can be synchronously received by piezoelectric single-crystal sensors at different spatial locations in the array, and the mechanical signals are converted into electrical signals through the positive piezoelectric effect. Through multi-node synchronous acquisition, multi-dimensional information such as the arrival time, amplitude, and waveform of the stress waves can be obtained simultaneously, providing a complete data foundation for subsequent crack location analysis.

[0090] On the other hand, the embodiments of this application use the crack location results obtained based on time difference of arrival and energy attenuation model as different information sources for joint utilization. By introducing a multi-source information fusion mechanism, the crack location is dynamically corrected and optimized, so that the location results remain stable and continuous under engineering conditions such as noise interference, material heterogeneity and parameter uncertainty.

[0091] On the other hand, to adapt to the early-age hydration heat inside concrete, the scouring effect of pouring and mixing machinery, and the long-term service environment, this study adopts a cement-based composite material encapsulation strategy for the piezoelectric single-crystal sensor. Specifically, the encapsulation process involves wrapping the piezoelectric single-crystal wafer with a cement-based composite material to form a small cylindrical unit with a diameter of approximately 15 mm, which is then cured using standard curing methods. Furthermore, to minimize the relative positional error of the multi-node array during the embedding stage, acrylic positioning templates are used to prefabricate the installation holes, allowing the spatial geometry of the array to be formed during the pouring stage, avoiding array geometric deviations caused by manual positioning errors.

[0092] The sensor embedding location can be determined using finite element simulation, prioritizing the placement of array nodes along potential crack initiation zones. Alternatively, service history data from similar structures can be used to prioritize array placement in common cracking areas (such as expansion joint transition zones, negative bending moment zones in the base plate, and the edges of holes).

[0093] On the other hand, to ensure the array has sufficient inversion capability, the distance between sensors must not be too large. The sensor spacing must satisfy the spatial sampling theorem, i.e., the spacing must be less than half the minimum detectable crack length to avoid signal aliasing. If the sensor spacing is too large, high-frequency spatial signals (such as short-wavelength stress waves generated by small-scale cracks) will be misidentified as low-frequency signals, leading to positioning errors.

[0094] In practical engineering, the array layout can follow the following principles: (1) Prioritize covering stress concentration areas (such as mid-span locations and near constraint edges); (2) Layout can be layered along the thickness direction to obtain crack depth information; (3) Avoid passing through densely reinforced steel zones to reduce electromagnetic shielding and stress wave reflection.

[0095] In terms of layout strategy, the array can adopt a combination of "grid layout + layered layout" to simultaneously meet the requirements of planar coverage and thickness direction resolution.

[0096] Specifically, a gridded layout can be used in planar areas: if the structural stress field distribution is relatively uniform, an orthogonal equidistant grid can be used to ensure spatial sampling consistency; if the local stress gradient is significant (such as in holes, slots, or stress concentration edges), an adaptive grid can be used, appropriately increasing the number of sensor nodes in high-stress areas. In the thickness direction, a layered layout can be used to distinguish the spatial depth characteristics of cracks. Through this combined layout of "planar gridding + thickness layering," the array can simultaneously sense the appearance, depth, and penetration process of cracks from multiple directions, thereby achieving phased crack monitoring capabilities.

[0097] In summary, the concrete crack monitoring method based on arrayed piezoelectric single-crystal sensors proposed in this invention improves the overall acquisition capability of stress wave signals through the structured deployment of arrayed piezoelectric single-crystal sensors, and provides the necessary spatial constraints for crack location inversion.

[0098] Based on the stress wave synchronous acquisition mechanism of array multi-node, multi-dimensional information such as arrival time, amplitude and waveform of stress waves can be obtained at the same time, providing a complete data foundation for subsequent crack location analysis.

[0099] By using crack initial location processing based on time difference of arrival, the crack location is accurately located. Furthermore, crack location correction based on energy decay model introduces geometric decay and material decay to locate the crack location from a physical dimension.

[0100] The fusion of multi-source observation results ensures that the positioning results remain stable and continuous even under engineering conditions such as noise interference, material heterogeneity, and parameter uncertainty.

[0101] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; those skilled in the art can modify the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring concrete cracks based on an array-type piezoelectric single-crystal sensor, characterized in that, The following steps are included: Multiple piezoelectric single crystal sensors are arranged at predetermined intervals inside or on the surface of a concrete structure to form an array-type sensing system. The array-type sensing system synchronously collects stress wave signals released during the generation or expansion of internal cracks in concrete in real time. The initial estimated location of the crack is calculated based on the time difference between the stress wave signals received by each piezoelectric single crystal sensor in the array-type sensing system. By combining stress wave signals with energy attenuation analysis, the initial estimated location of the crack is corrected to obtain the corrected estimated location. The initial estimated location and the corrected estimated location are fused using a prediction-correction multi-source information fusion method to obtain the target crack location.

2. The method for monitoring concrete cracks based on an array-type piezoelectric single-crystal sensor according to claim 1, characterized in that, Stress wave signals include signal arrival time, signal amplitude, and waveform characteristics.

3. The method for monitoring concrete cracks based on an array-type piezoelectric single-crystal sensor according to claim 2, characterized in that, Based on the time difference between the stress wave signals received by each piezoelectric single-crystal sensor in the array-type sensing system, the initial estimated location of the crack is calculated, including the following steps: Extract the signal arrival time of each piezoelectric single crystal sensor in the array sensing system that receives the stress wave signal, and calculate the signal arrival time difference between any two piezoelectric single crystal sensors based on the signal arrival time. A crack location model is established based on the signal arrival time difference between any two piezoelectric single crystal sensors, the spatial coordinates of each piezoelectric single crystal sensor, and the stress wave propagation speed. Solve the crack location model to obtain the initial estimated location of the crack.

4. The method for monitoring concrete cracks based on an array-type piezoelectric single-crystal sensor according to claim 3, characterized in that, The crack location model is established based on the spatial distance relationship between the crack location and each piezoelectric single crystal sensor. The difference between the distance from the crack location to the first piezoelectric single crystal sensor and the distance to the second piezoelectric single crystal sensor is proportional to the signal arrival time difference.

5. A method for monitoring concrete cracks based on an array-type piezoelectric single-crystal sensor according to claim 4, characterized in that, The initial estimated location of the crack is corrected by combining stress wave signals with energy attenuation analysis, resulting in a corrected estimated location. The steps include the following: Extract the signal amplitude received by each piezoelectric single crystal sensor from the stress wave signal; Starting from the initial estimated location of the crack, the corrected estimated location is obtained by combining the pre-established signal energy attenuation relationship model with the analysis of the amplitude of each signal.

6. The method for monitoring concrete cracks based on an array-type piezoelectric single-crystal sensor according to claim 5, characterized in that, The signal energy attenuation model is used to describe the functional relationship between the signal amplitude of the crack source and the distance between the crack location and each piezoelectric single crystal sensor. The functional relationship includes a geometric diffusion attenuation term and a material absorption attenuation term.

7. A method for monitoring concrete cracks based on an array-type piezoelectric single-crystal sensor according to claim 6, characterized in that, Starting from the initial estimated location of the crack, and combining a pre-established signal energy attenuation model with the signal amplitudes received by each piezoelectric single-crystal sensor, the crack location is iteratively corrected to obtain the corrected estimated location. Specifically, by comparing the difference between the measured signal amplitudes of each piezoelectric single-crystal sensor and the theoretical signal amplitudes obtained based on the signal energy attenuation model, the crack location is updated multiple times to gradually reduce the difference, thereby obtaining the corrected estimated location.

8. A method for monitoring concrete cracks based on an array-type piezoelectric single-crystal sensor according to claim 7, characterized in that, The process of fusing the initial estimated location with the corrected estimated location to obtain the target crack location includes the following steps: The observation data for the current moment is constructed based on the initial estimated position and the corrected estimated position; the crack position for the current moment is predicted based on the target crack position estimation result of the previous moment, and the predicted position for the current moment is obtained; the predicted position is corrected and updated based on the difference between the predicted position and the observation data for the current moment, and the target crack position for the current moment is obtained; the updated target crack position is used as the input for the crack position prediction for the next moment, so as to realize the continuous dynamic estimation of the crack position.

Images