A concrete damage positioning method based on piezoelectric impedance data probability imaging

Abstract

The application belongs to the technical field of concrete structure health monitoring, and discloses a concrete damage positioning method based on piezoelectric impedance data probability imaging, which comprises the following steps: (1) arranging a plurality of piezoelectric sensors on a concrete structure, collecting admittance signals of each piezoelectric sensor under different working conditions, and then obtaining conductance values before and after concrete damage; (2) calculating an RMSDk damage index by using the obtained conductance values, establishing a regression model of the damage index RMSDk and damage distance, and calculating a prediction value of the concrete damage distance under each working condition; (3) taking the damage distance prediction value and the regression fitting coefficient as basic parameters of probability distribution, establishing a probability distribution model of a single sensor, and obtaining a visual image of concrete damage positioning through the probability joint distribution of a plurality of sensors. The application solves the problem that a single sensor can only qualitatively analyze structural damage and cannot quantitatively determine the damage position.

Description

Technical Field

[0001] This invention belongs to the technical field of concrete structure health monitoring, and more specifically, relates to a method for locating concrete damage based on probabilistic imaging of piezoresistive impedance data. Background Technology

[0002] Modern civil engineering structures are becoming increasingly large-scale and complex. During their service life, these structures inevitably suffer various forms of damage. If this damage accumulates and is not detected and addressed promptly, even minor damage can rapidly spread and cause the entire structure to fail, resulting in adverse social impacts. Accurately monitoring and identifying early, minute damage to structures is a significant challenge in structural engineering. Traditional health monitoring technologies include X-ray inspection, ultrasonic testing, magnetic particle testing, and eddy current testing. However, these technologies have drawbacks: the testing instruments are expensive and complex; the tested components must be shut down during testing; significant manpower and resources are required; and real-time, online, and remote structural monitoring is not possible.

[0003] Piezoelectric impedance analysis (EMI) is a damage detection technique based on impedance analysis of smart piezoelectric materials. Its basic principle is as follows: structural damage, such as cracks, structural defects, or steel corrosion, causes changes in the structure's mechanical impedance. Utilizing the electromechanical coupling effect of piezoelectric materials, when an alternating electric field is applied to a piezoelectric element attached to the main structure, the piezoelectric element vibrates mechanically due to the inverse piezoelectric effect, simultaneously causing the main structure to vibrate. The vibration of the main structure, in turn, affects the vibration of the piezoelectric element, which, due to the direct piezoelectric effect, produces an electrical response, manifested as a change in impedance. The extracted impedance signal contains information about the structural damage. By comparing this impedance signal with that of an undamaged structure, the extent of damage can be determined, enabling health monitoring and damage diagnosis. Piezoelectric impedance analysis has many advantages, including sensitivity to localized structural damage, insensitivity to far-field effects, low cost, and online monitoring. However, current technologies are still limited to qualitative damage analysis, only able to determine the degree of structural damage, and cannot achieve precise damage localization.

[0004] In practical engineering applications, it is necessary not only to determine whether initial damage has occurred in a structure, but also to locate the damage. EMI technology identifies structural damage by measuring admittance signals and judging the shift in the admittance curve. Experiments show that after concrete develops cracks at different locations and of different sizes, the piezoresistive impedance signal changes, and the degree of damage causes a shift in the admittance curve. However, for a single sensor, the degree of damage is a composite function of damage distance and damage size. Different damage conditions with the same or similar damage indicators exist, thus affecting the effectiveness of EMI technology in identifying damage location. Therefore, solving the problem that a single sensor cannot simultaneously consider both damage distance and damage size, and establishing an effective multi-sensor damage location method, remains to be addressed. Summary of the Invention

[0005] To address the aforementioned deficiencies or improvement needs of existing technologies, this invention provides a concrete damage localization method based on piezoresistive impedance data probabilistic imaging. This method solves the problem that a single sensor can only qualitatively analyze structural damage but cannot quantitatively determine the damage location. It provides a multi-sensor joint localization method for concrete structure damage localization, thereby improving the localization accuracy.

[0006] To achieve the above objectives, according to one aspect of the present invention, a method for locating concrete damage based on probabilistic imaging of piezoresistive impedance data is provided, the method mainly comprising the following steps:

[0007] (1) Multiple piezoelectric sensors are arranged on the concrete structure, and excitation voltage is applied to the piezoelectric sensors under each damage condition to collect the admittance signal of each piezoelectric sensor under different conditions, thereby obtaining the conductivity value of the concrete before and after damage.

[0008] (2) The RMSDk damage index was calculated using the electrical conductivity values ​​before and after concrete damage. A regression model between the damage index RMSDk and the damage distance L was established to calculate the predicted value of the concrete damage distance under various working conditions.

[0009] (3) The predicted damage distance obtained from the regression model and regression fit coefficient r 2 These two parameters serve as the basic parameters of the probability distribution. A probability distribution model for a single sensor is established, and then a visualized image of concrete damage location is obtained through the joint probability distribution of multiple sensors, thereby realizing the location of concrete damage.

[0010] Furthermore, the piezoelectric sensor is a surface-mount piezoelectric sensor, specifically a piezoelectric ceramic PZT.

[0011] Furthermore, the piezoelectric sensor is bonded to the concrete structure using epoxy resin.

[0012] Furthermore, the excitation voltage is 0.2V to 1V; the damage form is a crack on the concrete surface.

[0013] Furthermore, the formula for calculating the RMSDk damage index is as follows:

[0014]

[0015] in, This represents the admittance value of measurement point i under operating condition k. Let i be the admittance value of the corresponding measurement point under operating condition k-1, where i is the measurement point number and N is the total number of measurement points.

[0016] Furthermore, the formula for calculating the damage distance L is as follows:

[0017]

[0018] in, and The x and y axes represent the piezoelectric sensor, respectively. and The horizontal and vertical coordinate vectors of the crack are separated, and the min function is used to calculate the shortest distance from the piezoelectric sensor to the crack.

[0019] Furthermore, the expression for the regression model is:

[0020] Y = b1X1 + b2X2 + b3

[0021] Where Y represents the damage index RMSDk, X1 represents the damage distance L, X2 is the damage condition number, and b1, b2 and b3 are all undetermined parameters of the regression model.

[0022] Furthermore, the regression fit coefficient r 2 The calculation formula is:

[0023]

[0024] Where i represents the damage condition number, This represents the predicted damage distance obtained through a regression model. This represents the average damage distance under all operating conditions.

[0025] Furthermore, the expression for the probability distribution model is:

[0026]

[0027]

[0028] Where i represents the piezoelectric sensor number, j represents the damage condition number, and xi and y i σ represents the x-axis and y-axis of the piezoelectric sensor, respectively. 2 r is the fitting parameter of the regression model 2 L i,j Let R represent the distance between piezoelectric sensor i and crack j, R represent the distance between a point on the damage plane and the piezoelectric sensor, and M represent the set of sensors participating in the joint probability distribution localization.

[0029] In summary, compared with the prior art, the concrete damage localization method based on piezoresistive impedance data probabilistic imaging provided by this invention mainly has the following advantages:

[0030] Beneficial effects:

[0031] 1. This invention involves attaching multiple piezoelectric sensors to a concrete surface. When an excitation voltage is applied to the piezoelectric sensors, they resonate with the concrete. The admittance signals from each piezoelectric sensor are collected using this resonance. This enables real-time detection of early, minor damage to the concrete surface and allows for cross-verification of monitoring results to eliminate false alarms and ensure the reliability of damage identification results.

[0032] 2. By collecting the admittance signal of the sensor, a multiple linear regression model is used to eliminate the cumulative damage effect, thereby constructing a more accurate functional expression between the damage index RMSDk and the damage distance L, so as to realize damage localization.

[0033] 3. The present invention has high measurement accuracy and convenient method. Traditional piezoresistive impedance technology cannot handle the cumulative effect of damage well, that is, the damage that appears first will affect the identification effect of the damage that appears later.

[0034] 4. This invention visualizes the probability of damage at each point on the concrete surface through a probability distribution model. It not only allows for the free selection of sensors participating in joint positioning, but also provides excellent identification results and a clear and easy-to-understand structural representation. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the reinforced concrete slab constructed according to an embodiment of the present invention;

[0036] Figure 2 This is an admittance curve of the piezoelectric sensor PZT5 constructed according to the embodiments of the present invention under various operating conditions;

[0037] Figure 3 This is a partial enlarged view of the admittance curves of the piezoelectric sensor PZT5 constructed according to the embodiments of the present invention under various operating conditions;

[0038] Figure 4It is a regression model of the damage index RMSDk and damage distance L of the piezoelectric sensor PZT4 constructed according to the embodiments of the present invention;

[0039] Figure 5 This is a visualization of the damage probability distribution of the piezoelectric sensor PZT1 constructed according to the embodiments of the present invention;

[0040] Figure 6 This is a visualization of the damage probability distribution of the piezoelectric sensor PZT4 constructed according to the embodiments of the present invention;

[0041] Figure 7 This is a visualization of the damage probability distribution of the piezoelectric sensor PZT8 constructed according to the embodiments of the present invention;

[0042] Figure 8 This is a visualization image of the joint damage probability distribution of piezoelectric sensors PZT1, PZT4, and PZT8 constructed according to embodiments of the present invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0044] This invention provides a concrete damage localization method based on piezoresistive impedance data probabilistic imaging. It uses a multiple linear regression model to calculate the predicted value of the damage location and combines a joint probability distribution model to visualize the damage identification of concrete structures. This provides a feasible solution to the shortcomings of traditional EMI technology, which can only qualitatively analyze the degree of damage.

[0045] The positioning method mainly includes the following steps:

[0046] Step 1: Arrange multiple piezoelectric sensors on the concrete structure, and apply an excitation voltage to the piezoelectric sensors under each damage condition to collect the admittance signal of each piezoelectric sensor under different conditions, thereby obtaining the conductivity value of the concrete before and after damage.

[0047] The piezoelectric sensor is a surface-mount piezoelectric sensor, specifically a piezoelectric ceramic PZT; the piezoelectric sensor is bonded to the concrete structure using epoxy resin. The excitation voltage is 0.2V to 1V; the damage mode is surface cracks in the concrete.

[0048] Step two: Calculate the RMSDk damage index using the electrical conductivity values ​​of the concrete before and after damage, establish a regression model between the damage index RMSDk and the damage distance L, and use this model to calculate the predicted value of the concrete damage distance under various working conditions.

[0049] The formula for calculating the RMSDk damage index is as follows:

[0050]

[0051] in, This represents the admittance value of measurement point i under operating condition k. Let i be the admittance value of the corresponding measurement point under operating condition k-1, where i is the measurement point number and N is the total number of measurement points.

[0052] The formula for calculating the damage distance L is:

[0053]

[0054] in, and The x and y axes represent the piezoelectric sensor, respectively. and The horizontal and vertical coordinate vectors of the crack are separated, and the min function is used to calculate the shortest distance from the piezoelectric sensor to the crack.

[0055] The expression for the regression model is:

[0056] Y = b1X1 + b2X2 + b3

[0057] Where Y represents the damage index RMSDk, X1 represents the damage distance L, X2 is the damage condition number, and b1, b2, and b3 are undetermined parameters of the regression model. This regression model is a multiple linear regression model. By introducing the variable X2, the cumulative damage effect is eliminated, thereby better fitting the relationship between the damage distance L and the damage index RMSDk.

[0058] Step 3: Calculate the predicted damage distance values ​​obtained from the regression model. and regression fit coefficient r 2 These two parameters serve as the basic parameters of the probability distribution. A probability distribution model for a single sensor is established, and then a visualized image of concrete damage location is obtained through the joint probability distribution of multiple sensors, thereby realizing the location of concrete damage.

[0059] regression fit coefficient r 2 The calculation formula is:

[0060]

[0061] Where i represents the damage condition number, This represents the predicted damage distance obtained through a regression model. This represents the average damage distance under all operating conditions.

[0062] The expression for the probability distribution model is:

[0063]

[0064]

[0065] Where i represents the piezoelectric sensor number, j represents the damage condition number, and x i and y i σ represents the x-axis and y-axis of the piezoelectric sensor, respectively. 2 r is the fitting parameter of the regression model 2 L i,j Let R represent the distance between piezoelectric sensor i and crack j, R represent the distance between a point on the damage plane and the piezoelectric sensor, and M represent the set of sensors participating in the joint probability distribution localization.

[0066] In another embodiment, a method for locating concrete damage using piezoresistive impedance data probabilistic imaging employs multiple sensors in conjunction for damage localization, specifically including the following steps:

[0067] (1) As Figure 1 As shown, after the piezoelectric ceramic PZT is soldered with wires, it is then pasted onto the concrete surface with epoxy resin in the designated positions. After the epoxy resin has solidified, the damaged locations are marked.

[0068] (2) Figure 1 As shown, a cut 40mm long, 2mm wide, and 5mm deep is created at a designated location using the working conditions in sequence, with each working condition spaced 30 minutes apart.

[0069] (3) The admittance signals under various damaged conditions and undamaged conditions were collected using an Agilent 4294A impedance meter.

[0070] (4) Based on the admittance value obtained in step (3), obtain the RMSDk damage index, establish a regression model between the damage index RMSDk and the damage distance L, and use this model to calculate the predicted value of the concrete damage distance under each working condition. This allows for a rough location of concrete damage; subsequently, the predicted damage distance values ​​obtained from the regression model are used. and regression fit coefficient r 2 These two parameters serve as the basic parameters of the probability distribution. A probability distribution model for a single sensor is established, and then a visualized image of concrete damage location is obtained through the joint probability distribution of multiple sensors, thereby achieving precise location of concrete damage.

[0071] RMSDk is the square root of the ratio of the square of the signal deviation measured before and after damage to the square of the signal before damage in the concrete structure. This statistical index measures the degree of change in the electrical signal of the PZT (Piezoelectric ceramics) sensor before and after the admittance signal change, and is obtained according to the following formula:

[0072]

[0073] in, This represents the admittance value of measurement point i under operating condition k. Let i be the admittance value of the corresponding measurement point under operating condition k-1, where i is the measurement point number and N is the total number of measurement points.

[0074] The present invention will now be described in further detail with reference to specific embodiments.

[0075] like Figure 1 As shown, in this embodiment, the specimen is a reinforced concrete block with dimensions of 1000mm × 500mm × 60mm. A cutting machine was used to cut slits on its surface, and an Agilent 4294A impedance meter was used to collect admittance signals under different damage conditions.

[0076] The method for locating concrete damage using an adhesive piezoelectric sensor provided by this invention involves cutting a concrete test block. For example... Figure 2 and Figure 3 The figure shows the admittance curves and magnified views of PZT under various operating conditions.

[0077] from Figure 4 It can be seen that after eliminating the damage accumulation effect, the damage index RMSDk decreases with the increase of damage distance.

[0078] from Figure 5 , Figure 6 and Figure 7 Damage identification results for piezoelectric sensors PZT1, PZT4, and PZT8 can be obtained. In the figure, the black bars represent cracks, and the yellow ring area is the area with the highest probability of damage. It can be seen that the damage identification effect of a single sensor is not good, and it cannot distinguish the direction of the damage location.

[0079] from Figure 8 The damage identification results obtained by the joint localization of piezoelectric sensors PZT1, PZT4 and PZT8 can be obtained. The red dot in the figure is the point with the maximum damage probability, which is located near the crack. Obviously, the identification effect of this method is excellent, and the damage location of the concrete is accurately located with an error of less than 1mm.

[0080] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for locating concrete damage based on probabilistic imaging of piezoresistive impedance data, characterized in that, The method includes the following steps: (1) Multiple piezoelectric sensors are arranged on the concrete structure, and excitation voltage is applied to the piezoelectric sensors under each damage condition to collect the admittance signal of each piezoelectric sensor under different conditions, thereby obtaining the conductivity value of the concrete before and after damage. (2) Calculate the RMSDk damage index using the electrical conductivity values ​​before and after concrete damage, and establish the relationship between the damage index RMSDk and the damage distance. L A regression model was used to calculate the predicted value of concrete damage distance under various working conditions. ; (3) The predicted damage distance obtained from the regression model and regression fit coefficients r 2 These two parameters serve as the basic parameters of the probability distribution. A probability distribution model for a single sensor is established, and then a visualized image of concrete damage location is obtained through the joint probability distribution of multiple sensors, thereby realizing the location of concrete damage.

2. The concrete damage localization method based on piezoresistive impedance data probabilistic imaging as described in claim 1, characterized in that: The piezoelectric sensor is a surface-mount piezoelectric sensor, specifically a piezoelectric ceramic PZT.

3. The concrete damage localization method based on piezoresistive impedance data probabilistic imaging as described in claim 2, characterized in that: The piezoelectric sensor was bonded to the concrete structure using epoxy resin.

4. The concrete damage localization method based on piezoresistive impedance data probabilistic imaging as described in claim 1, characterized in that: The excitation voltage is 0.2V to 1V; the damage form is cracks on the concrete surface.

5. The concrete damage localization method based on piezoresistive impedance data probabilistic imaging as described in claim 1, characterized in that: The formula for calculating the RMSDk damage index is as follows: in, Indicates working conditions k Measurement points g The admittance value, For the corresponding measurement points under operating conditions k- The admittance value is 1, g is the number of the measuring point, and N is the total number of measuring points.

6. The concrete damage localization method based on piezoresistive impedance data probabilistic imaging as described in claim 5, characterized in that: The damage distance L The calculation formula is: in, and The x and y axes represent the piezoelectric sensor, respectively. and The horizontal and vertical coordinate vectors of the crack are separated, and the min function is used to calculate the shortest distance from the piezoelectric sensor to the crack.

7. The concrete damage localization method based on piezoresistive impedance data probabilistic imaging as described in claim 6, characterized in that: The expression for the regression model is: in, Y The damage index RMSDk represents... Indicates the distance of damage L , The number is the damage condition number. , and All of these are undetermined parameters for the regression model.

8. The concrete damage localization method based on piezoresistive impedance data probabilistic imaging as described in any one of claims 1-7, characterized in that: regression fit coefficients r 2 The calculation formula is: in, i The number indicating the damage condition. This represents the predicted damage distance obtained through a regression model. This represents the average damage distance under all operating conditions.

9. The concrete damage localization method based on piezoresistive impedance data probabilistic imaging as described in claim 8, characterized in that: The expression for the probability distribution model is: in, h This indicates the serial number of the piezoelectric sensor. j Indicates the crack number. and The x and y axes represent the piezoelectric sensor, respectively. The fitting parameters of the regression model r 2 , Indicates piezoelectric sensor h With cracks j The distance between them R The distance between a point on the damage plane and the piezoelectric sensor is represented by M, which represents the set of sensors participating in the joint probability distribution localization.

Images