Evaluation Method for Steel Pipe Pole Caps of Transmission Lines Based on Precast Pipe Piles

Abstract

This invention discloses an evaluation method for steel pipe pole foundations of transmission lines based on precast pipe piles, belonging to the field of evaluation technology for steel pipe pole foundations of transmission lines. The method includes the following steps: exciting elastic waves at the foundation-pipe pile interface using an acoustic wave emitting device, and receiving the acoustic wave signals by an ultrasonic array sensor on the top of the pipe pile; performing feature separation on the acoustic wave signals and evaluating the relative deviation of sound velocity based on the separation results; constructing an optimal sound velocity gradient interval based on the evaluation results using a dynamic optimization algorithm including a Gaussian mixture model, a density peak detection algorithm, and NSGA-II optimization; identifying gradient abrupt change regions within the interval and expanding the defect detection range based on these regions; constructing a defect feature map based on the expanded defect detection range; locating void defects according to the defect feature map and outputting a graded evaluation result. This method solves the problem that traditional methods, when evaluating by sound velocity, do not dynamically optimize the sound velocity gradient interval, resulting in a limited defect detection range and consequently reduced accuracy in foundation evaluation.

Description

Technical Field

[0001] This invention relates to the field of evaluation technology for steel pipe pole foundations of transmission lines, and more specifically, to a method for evaluating steel pipe pole foundations of transmission lines based on precast pipe piles. Background Technology

[0002] With the continuous development of infrastructure construction, especially in industries such as power and transportation, steel pipe piles and their related foundation structures are widely used in critical projects such as power transmission lines, communication towers, and bridges. Due to long-term external environmental pressures, climate change, and potential quality issues during construction, the safety and stability of these structures often face varying degrees of risk. In particular, the connection area between precast pipe piles and steel pipe pile foundations, as a critical load-bearing component, directly affects the stability and reliability of the entire structure.

[0003] In recent years, with the development of acoustic imaging and ultrasonic testing technologies, structural health monitoring methods based on acoustic signals have been widely applied in engineering practice. The propagation characteristics of acoustic signals can effectively reflect the internal defects of materials. By exciting elastic waves inside a structure and analyzing its reflected and scattered waves, defects such as cracks and voids that may exist in the structure can be accurately detected. However, existing acoustic detection methods still have some shortcomings, mainly in the processing of acoustic signals, the accuracy of defect localization, and the optimization of sound velocity inversion models.

[0004] For example, the invention patent announcement CN115586258A describes a method for assessing internal corrosion of substation grounding flat steel using spiral-guided ultrasonic waves. This method establishes a relationship between the ultrasonic wave propagation speed and the aging degree of the pantograph contactor by measuring the ultrasonic wave propagation speed within the contactor at different aging levels. This invention provides a scientific and rapid method for assessing the aging degree of the pantograph contactor, solving the inaccuracy problem caused by current manual visual assessment. It has positive theoretical guiding significance for assessing the aging degree of the pantograph contactor.

[0005] The above-disclosed technical solutions have at least the following technical problems:

[0006] Traditional assessment methods fail to accurately assess the relative deviation of sound velocity, resulting in limited detection accuracy. This is especially true in cases of complex structures or inhomogeneous materials, where they cannot accurately pinpoint potential defect locations. Furthermore, the lack of dynamic optimization of the sound velocity gradient range limits the scope and accuracy of defect detection, thereby reducing the accuracy of assessments for steel pipe pole foundations in transmission lines.

[0007] To address the above problems, this invention proposes a solution. Summary of the Invention

[0008] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide an evaluation method for steel pipe pole foundations of transmission lines based on precast pipe piles. This method expands the defect detection range by constructing an optimal sound velocity gradient interval, thereby addressing the problem that traditional methods do not dynamically optimize the sound velocity gradient interval when evaluating based on sound velocity, resulting in a limited range of defect detection and consequently reduced accuracy of foundation evaluation.

[0009] To achieve the above objectives, the present invention provides the following technical solution:

[0010] The evaluation method for steel pipe pole abutments of transmission lines based on precast pipe piles includes the following steps: elastic waves are excited at the abutment-pipe pile interface using an acoustic wave emitting device, and the acoustic wave signals are received by an ultrasonic array sensor on the top of the pipe pile; the acoustic wave signals are feature-separated, and the relative deviation of the sound velocity is evaluated based on the separation results; based on the evaluation results, an optimal sound velocity gradient interval is constructed using a dynamic optimization algorithm including a Gaussian mixture model, a density peak detection algorithm, and NSGA-II optimization; gradient abrupt change regions are identified within the interval, and the defect detection range is expanded based on these regions; a defect feature map is constructed based on the expanded defect detection range; void defects are located according to the defect feature map, and a graded evaluation result is output.

[0011] In a preferred embodiment, the step of exciting elastic waves at the pile cap-pipe pile interface using an acoustic wave emitting device and receiving the acoustic wave signals by an ultrasonic array sensor on the top of the pipe pile specifically involves: deploying a multi-source ring acoustic wave emitting array on the surface of the pile cap and exciting pulse signals of different frequencies according to a preset timing sequence; synchronously activating the ultrasonic array sensor group on the top of the pipe pile to collect the time-domain sequences of reflected and scattered waves; dividing the time-domain sequences into time windows and extracting the acoustic wave signals based on a preset wave velocity inversion model.

[0012] In a preferred embodiment, the step of feature separation of the acoustic signal and evaluation of the relative deviation of sound speed based on the separation results specifically involves: denoising the acoustic signal using a spatial filtering algorithm to obtain a denoised signal; performing empirical mode decomposition on the denoised signal to obtain principal components; extracting the time-frequency feature vectors of the principal components and calculating the offset between the time-frequency feature vectors and a preset dispersion curve; calculating the relative deviation matrix of sound speed using a cross-correlation algorithm based on the offset; and evaluating the relative deviation of sound speed based on the relative deviation matrix to obtain a relative deviation evaluation matrix of sound speed.

[0013] In a preferred embodiment, the step of constructing the optimal sound velocity gradient interval based on the evaluation results using a dynamic optimization algorithm including a Gaussian mixture model, a density peak detection algorithm, and NSGA-II optimization is specifically as follows: The sound velocity relative deviation evaluation matrix is ​​input into the Gaussian mixture model, and a density peak detection algorithm is used to identify several gradient communities. A first sound velocity gradient topology map is constructed based on the spatial distribution of the gradient communities and the gradient change direction. First constraints are constructed based on the first sound velocity gradient topology map, including physical range constraints on the sound velocity gradient, confidence interval constraints within each Gaussian distribution community, and maximum rate of change constraints between adjacent sampling points. Based on the first constraints, random sampling is performed within a preset probability density distribution using the Monte Carlo method to obtain several first sound velocity gradient samples. A preset distributed community algorithm is used to perform NSGA-II optimization on the several first sound velocity gradient samples to obtain the optimal sound velocity gradient interval.

[0014] In a preferred embodiment, the step of using a preset distributed community algorithm to perform NSGA-II optimization on several first sound velocity gradient samples to obtain the optimal sound velocity gradient interval specifically involves: treating several first sound velocity gradient samples as individuals and grouping them according to preset community labels to obtain several first groups; using the NSGA-II algorithm to perform non-dominated sorting on each first group to obtain several Pareto front solution sets; performing cross-community migration on several Pareto front solution sets to obtain a first solution set distribution; dividing and merging the first groups based on the first solution set distribution to obtain a reconstructed community structure; and performing NSGA-II optimization again based on the reconstructed community structure to obtain the optimal sound velocity gradient interval.

[0015] In a preferred embodiment, the step of performing cross-community migration on several Pareto front solution sets to obtain a first solution set distribution specifically involves: calculating the crowding distance on several Pareto front solution sets, and filtering the several Pareto front solution sets according to the crowding distance to obtain several boundary solutions; calculating the gradient similarity matrix of several boundary solutions, and generating community migration instructions based on the gradient similarity matrix; and performing cross-community migration on several boundary solutions according to the community migration instructions to obtain the first solution set distribution.

[0016] In a preferred embodiment, the step of segmenting and merging the first population based on the first solution set distribution to obtain the reconstructed community structure specifically involves: counting the number of migrations in the first solution set distribution and making a merging judgment based on the number of migrations; if merging and reconstruction are to be performed, extracting the gradient principal direction vector of each community in the first solution set distribution; and segmenting and merging the communities based on the gradient principal direction vector to obtain the reconstructed community structure.

[0017] In a preferred embodiment, gradient abrupt change regions are identified within the interval, and the defect detection range is expanded based on these regions. Specifically, the pipe pile is constructed into a three-dimensional spatial grid according to axial layering and radial sectoring; within the optimal sound velocity gradient interval, the absolute value of the gradient change rate of adjacent spatial grids is calculated; gradient abrupt change is identified in the grid based on the absolute value of the gradient change rate to generate an initial defect seed point set; spatial expansion is performed along the axial and radial directions of the pipe pile with the preset seed point as the center to obtain the expanded region; the expanded region and the initial defect seed point set are merged to obtain the defect detection range.

[0018] In a preferred embodiment, the step of constructing a defect feature map based on the expanded defect detection range, locating voided defects based on the defect feature map, and outputting a graded evaluation result specifically involves: dividing the defect detection range into a three-dimensional voxel grid according to a preset resolution; performing multi-band acoustic energy attenuation imaging on each voxel to generate an initial defect energy map; constructing a defect feature tensor based on the initial defect energy map; extracting the voided region boundary features from the defect feature tensor using a three-dimensional convolutional neural network; and classifying the defects into three levels—minor, moderate, and severe—based on the voided region boundary features, and outputting a defect graded evaluation result that includes location, size, and level.

[0019] The technical effects and advantages of this invention regarding the evaluation method for transmission line steel pipe pole caps based on precast pipe piles are as follows:

[0020] 1. This invention excites elastic waves at the interface between the pile cap and the pipe pile using an acoustic wave emitting device. Reflected and scattered wave signals are received by an ultrasonic array sensor at the top of the pipe pile. By separating the features of these acoustic signals and removing noise, the relative deviation of the sound velocity can be accurately assessed, providing high-quality signal data for subsequent defect detection. The key technology in this process is based on the extraction of the time-frequency feature vector of the acoustic signal, calculating the relative deviation matrix of the sound velocity using a cross-correlation algorithm, and obtaining an accurate assessment result of the relative deviation of the sound velocity through precise evaluation, laying the foundation for subsequent defect localization. Based on the assessment results, a dynamic optimization algorithm is used to construct the optimal sound velocity gradient interval. By combining a Gaussian mixture model with a density peak detection algorithm, abrupt changes in the sound velocity gradient can be effectively identified, thereby accurately defining the optimal sound velocity gradient interval. This optimal gradient interval not only optimizes the sound wave propagation path but also ensures higher accuracy and sensitivity in detecting structural defects within this interval. By optimizing the sound velocity gradient, interference from external noise can be reduced, the accuracy of defect localization can be improved, and more reliable assessment results can be provided.

[0021] 2. By constructing an optimal sound velocity gradient range, this invention further expands the defect detection range. Through the identification and expansion of gradient abrupt change regions, it can cover more potential defect areas, especially in complex structural environments, enabling more comprehensive detection of potential structural defects such as cracks and voids. By establishing a defect feature map within this expanded region and combining it with a three-dimensional convolutional neural network to extract defect features, this invention achieves efficient defect localization and hierarchical evaluation, significantly improving detection accuracy and efficiency. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating the evaluation method for transmission line steel pipe pole caps based on precast pipe piles according to the present invention.

[0023] Figure 2 This is a heat map of the relative deviation matrix of sound speed in this invention;

[0024] Figure 3 This is the first sound velocity gradient topology diagram of the present invention;

[0025] Figure 4 This is a three-dimensional defect energy map of the present invention. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0027] Example 1, Figure 1 The present invention provides an evaluation method for steel pipe pole caps of transmission lines based on precast pipe piles, comprising the following steps:

[0028] S1, elastic waves are excited at the pile cap-pipe pile interface by an acoustic wave emitting device, and the acoustic wave signals are received by an ultrasonic array sensor at the top of the pipe pile.

[0029] In this example, an elastic wave is excited at the pile cap-pipe pile interface by an acoustic wave emitting device, and the acoustic wave signal is received by an ultrasonic array sensor at the top of the pipe pile. Specifically:

[0030] A multi-source ring acoustic wave emission array is arranged on the surface of the pier to excite pulse signals of different frequencies according to a preset timing sequence;

[0031] The ultrasonic array sensor group at the top of the pipe pile is activated simultaneously to collect the time-domain sequences of reflected and scattered waves;

[0032] The time-domain sequence is divided into time windows, and the acoustic signal is extracted based on a preset wave velocity inversion model.

[0033] It should be noted that acoustic signals refer to the wave signals generated by elastic waves excited by an acoustic wave emitting device, which propagate through the interface between the pile cap and the pipe pile, and then undergo reflection and scattering. These acoustic signals contain key information about the internal properties of the medium (such as the pile cap and pipe pile structure), such as the material density, defect location, and structural integrity of the interface. By receiving these acoustic signals through an ultrasonic array sensor group, characteristics such as wave propagation time and frequency changes can be captured, which can be further used to analyze and evaluate the health status of the pipe pile. The wave velocity inversion model refers to a model that calculates wave velocity changes by mathematically modeling the propagation characteristics of acoustic signals (such as propagation time and frequency). This model is based on a certain inversion algorithm and combines time-domain data of actual acoustic signals to perform wave velocity inversion calculations. By comparing the difference between the actual measured wave velocity and the preset value in the theoretical model, the relative deviation of the sound velocity of the medium can be obtained, thereby reflecting the physical state of the pipe pile and the pile cap and identifying potential defect areas. This method uses changes in wave velocity to identify defects such as voids and cracks that may exist in the structure, thus providing an effective basis for defect location and evaluation.

[0034] Furthermore, the acoustic signal acquisition process specifically involves deploying a multi-source ring-shaped acoustic wave transmitting array on the surface of the pile cap, exciting pulse signals of different frequencies according to a preset timing sequence to generate elastic waves, and simultaneously activating the ultrasonic array sensor group at the top of the pipe pile to receive the time-domain sequences of reflected and scattered waves. Subsequently, the acquired time-domain signals are processed by time window segmentation, and effective acoustic signal features are extracted based on a pre-established wave velocity inversion model. This achieves high-precision excitation and signal capture of the elastic waves at the pile cap-pipe pile interface, providing a data foundation for subsequent sound velocity analysis. This scheme, through the time-sequential excitation and synchronous reception design of the multi-source ring array, combined with time window segmentation and the wave velocity inversion model, effectively improves the spatiotemporal resolution and signal-to-noise ratio of the interface acoustic wave signal.

[0035] S2, perform feature separation on the acoustic signal, and evaluate the relative deviation of the sound velocity based on the separation results;

[0036] In this example, feature separation is performed on the acoustic signal, and the relative deviation of the sound velocity is evaluated based on the separation results, specifically as follows:

[0037] A spatial filtering algorithm is used to reduce the noise of the acoustic signal, resulting in a denoised signal.

[0038] Empirical mode decomposition is performed on the denoised signal to obtain the principal components;

[0039] Extract the time-frequency eigenvectors of the principal components and calculate the offset between the time-frequency eigenvectors and the preset dispersion curve;

[0040] Based on the offset, the relative deviation matrix of sound speed is calculated using a cross-correlation algorithm;

[0041] The relative deviation of sound speed is evaluated based on the relative deviation matrix of sound speed, and the relative deviation of sound speed evaluation matrix is ​​obtained.

[0042] In this example, the specific calculation formula for the spatial domain filtering algorithm is as follows:

[0043]

[0044] in, For noise reduction signal, Let be the weighting coefficient of the i-th ultrasonic array sensor. The original time-domain signal received by the i-th ultrasonic array sensor. This represents the number of ultrasonic array sensors.

[0045] In this example, the formula for calculating the relative sound speed deviation matrix is ​​as follows:

[0046]

[0047] in, Let be the relative deviation of the sound velocity between the i-th and j-th ultrasonic array sensors. For time-frequency feature vectors, Let be the offset between the time-frequency feature vector corresponding to the i-th ultrasonic array sensor and the preset dispersion curve. Let be the offset between the time-frequency feature vector corresponding to the j-th ultrasonic array sensor and the preset dispersion curve. The number of principal components.

[0048] It should be noted that spatial filtering algorithms are techniques for removing noise signals, particularly suitable for processing high-frequency noise components in acoustic signals. In this process, the acoustic signal is first received by a sensor array and converted into a digital signal, which is then processed using spatial filtering algorithms. Spatial filtering analyzes the spatial distribution of the acoustic signal and uses specific mathematical models (such as high-pass, low-pass, or band-pass filters) to suppress noise components. By filtering out noise signals and preserving the original reflected and scattered wave components, the signal becomes clearer, facilitating subsequent feature extraction and analysis.

[0049] Furthermore, dispersion curves describe the relationship between frequency and wave speed as a wave propagates through a medium. In sound wave propagation, waves of different frequencies may propagate at different speeds. By demonstrating the relationship between frequency and wave speed, dispersion curves reflect the elastic properties, density, and structural characteristics of the medium. Studying dispersion curves allows us to understand the distribution of the medium and identify structural defects. For example, a significant change in wave speed may indicate the presence of defects such as cracks or voids in the medium.

[0050] Finally, after calculating the relative deviation matrix of sound velocity, the next step is to evaluate these deviations. The evaluation process involves analyzing the relative deviation values ​​of sound velocity and scoring each region according to a pre-defined standard or model. For example, if the deviation values ​​of some regions are large, it may indicate the presence of defects (such as cracks or voids) in those regions. Through a comprehensive evaluation of all deviation values, a relative deviation evaluation matrix of sound velocity is generated, which is used to represent the health status of each detected region, such as... Figure 2 As shown, the heatmap matrix visually reflects the deviation distribution between signals received by different sensors. Abnormal areas (red / dark blue) indicate possible abnormal material density or connection defects. This assessment matrix can help engineers identify potential abnormal areas in pipe piles or pile caps, providing data support for subsequent defect location and assessment.

[0051] S3. Based on the evaluation results, the optimal sound velocity gradient range is constructed using a dynamic optimization algorithm that includes Gaussian mixture model, density peak detection algorithm and NSGA-II optimization.

[0052] In this example, based on the evaluation results, a dynamic optimization algorithm including Gaussian mixture model, density peak detection algorithm, and NSGA-II optimization is used to construct the optimal sound velocity gradient interval, specifically:

[0053] The relative deviation evaluation matrix of sound speed is input into the Gaussian mixture model, and several gradient communities are identified by the density peak detection algorithm. The first sound speed gradient topology map is constructed based on the spatial distribution of the gradient communities and the gradient change direction.

[0054] The first constraint conditions are constructed based on the first sound velocity gradient topology graph, including the physical range constraint of the sound velocity gradient, the confidence interval constraint within each Gaussian distribution community, and the maximum rate of change constraint between adjacent sampling points.

[0055] Based on the first constraint, the Monte Carlo method is used to randomly sample within the preset probability density distribution to obtain several first sound speed gradient samples.

[0056] The optimal sound velocity gradient range is obtained by using a pre-defined distributed community algorithm to optimize several first sound velocity gradient samples using NSGA-II.

[0057] In this example, the specific calculation formula for the Gaussian mixture model is as follows:

[0058]

[0059] in, Let x be the probability density of the observed data point. Let be the weights of the t-th Gaussian distribution. Let be the mean of the t-th Gaussian distribution. Let be the covariance matrix of the t-th Gaussian distribution. Let be the probability density function of the t-th Gaussian distribution. The number of Gaussian distributions represents the number of gradient communities to be identified.

[0060] It should be noted that the density peak detection algorithm is a clustering method based on the local density and distance of data points. In this step, the algorithm analyzes the density distribution of different regions in the relative deviation matrix of sound speed. By calculating the local density of each data point (i.e., the density of the surrounding area) and the distance to the farthest point from that point, the density peak of each data point is determined. Based on these density peaks, the density peak detection algorithm can effectively identify multiple gradient communities (i.e., regions that exhibit significant differences in sound speed gradient changes). Using this method, a first sound speed gradient topology map can be constructed, showing the structure of the sound speed gradient distribution and the relationships between different gradient regions.

[0061] Furthermore, the first constraint refers to the restriction imposed on the range and pattern of sound speed gradient variation when constructing the optimal sound speed gradient interval. Specifically, the constraint may include limiting the maximum or minimum value of the sound speed gradient, its rate of change, or requiring the gradient to remain consistent within certain specific regions. The mathematical form of the first constraint can be expressed as the following joint constraint:

[0062]

[0063] in, Let x be the sound speed gradient at position x. and These are the minimum and maximum values ​​of the sound velocity gradient, respectively (determined by the physical properties of the sound propagation medium, such as in seawater). , ), and denoted as the mean and covariance matrix of the t-th Gaussian distribution (obtained through Gaussian mixture model clustering), and k is the confidence coefficient (usually k=2, corresponding to the 95% confidence interval). This is the maximum allowable rate of change (the gradient difference threshold per unit distance). adjacent sampling points , The spatial distance between two points.

[0064] These constraints ensure that the optimization process does not deviate from actual physical phenomena and avoids constructing unreasonable sound speed gradient ranges. For example, if the sound speed changes drastically within a certain region, it may be necessary to limit the rate of change of the gradient. and the gradient range within the community ( This ensures that the trend of change in this range conforms to the actual physical constraints.

[0065] In this example, a pre-defined distributed community algorithm is used to perform NSGA-II optimization on several first sound velocity gradient samples to obtain the optimal sound velocity gradient range, specifically:

[0066] Several first sound velocity gradient samples are treated as individuals and grouped according to preset community labels to obtain several first groups;

[0067] The NSGA-II algorithm is used to perform non-dominated sorting on each first group to obtain several Pareto front solution sets;

[0068] By performing cross-community migration on several Pareto front solution sets, the first solution set distribution is obtained;

[0069] Based on the distribution of the first solution set, the first group is segmented and merged to obtain the reconstructed community structure;

[0070] Based on the reconstructed community structure, NSGA-II optimization was performed again to obtain the optimal sound speed gradient range.

[0071] For example, suppose that during the evaluation of a steel pipe pile cap for a transmission line, elastic waves are first excited at the cap-pile interface using an acoustic wave transmitter, and acoustic signals from the cap to the pile area are received by an ultrasonic array sensor. After processing, the relative deviation matrix of sound velocity in this area is obtained, and it is divided into several gradient communities using a Gaussian mixture model (GMM). These communities are identified using a density peak detection algorithm, and a first sound velocity gradient topology map is constructed based on the spatial distribution and gradient change direction of the gradient communities. Specifically, each gradient community is treated as a topology node, and the node attributes include the community center coordinates, average gradient value, and gradient principal direction vector. If the absolute value of the difference in average gradient values ​​between two community spatial regions is less than a preset threshold, and the similarity of the gradient principal direction vectors is greater than a preset threshold, then an edge is established between these two gradient communities, such as... Figure 3 As shown, different colors represent multiple sound velocity gradient communities automatically divided by the Gaussian mixture model, demonstrating the spatial distribution pattern of gradient changes, which helps to identify potential heterogeneous structural regions.

[0072] Next, based on the structure of these gradient communities, a preset first constraint condition was used to define the range of sound speed gradient variation. For example, it was assumed that the set sound speed gradient intervals were [1800, 2200] m / s, [1800, 2200] m / s, and [1800, 2200] m / s, and that the maximum rate of gradient change within this interval must not exceed 100 m / s / km. Based on these constraints, random sampling was performed within this probability distribution using the Monte Carlo method, resulting in several sound speed gradient samples ranging from 1800 m / s to 2200 m / s.

[0073] Subsequently, the NSGA-II algorithm was used to perform multi-objective optimization on these sound velocity gradient samples, with the optimization objectives being to minimize the volatility of gradient changes and enhance the uniformity of the gradient distribution. Specifically, the NSGA-II algorithm performs non-dominated sorting on each sound velocity gradient sample, dividing the samples into multiple populations, and performs multiple rounds of crossover, mutation, and selection based on the Pareto front solution set, ultimately obtaining a set of optimal sound velocity gradient intervals. These optimal solutions represent the equilibrium state of the sound velocity gradient distribution, while satisfying constraints such as maximum gradient change rate and minimum volatility.

[0074] Ultimately, after optimization, an optimal sound velocity gradient range was obtained: [1850, 2150] m / s, [1850, 2150] m / s, and [1850, 2150] m / s. This range represents the area within the steel pipe pile cap region of the transmission line where the sound velocity variation is most stable and representative. Using this optimal range, engineers can accurately identify potential defect areas at the cap-pile interface, such as cracks or voids, providing crucial decision support for further structural reinforcement or repair.

[0075] In this example, cross-community migration is performed on several Pareto front solutions to obtain the first solution set distribution, specifically:

[0076] The crowding distance is calculated for several Pareto front solution sets, and the crowding distance is used to filter several Pareto front solution sets to obtain several boundary solutions;

[0077] Calculate the gradient similarity matrix of several boundary solutions, and generate community migration instructions based on the gradient similarity matrix;

[0078] According to the community migration instruction, several boundary solutions are migrated across communities to obtain the distribution of the first solution set.

[0079] It should be noted that, firstly, a crowding distance is defined for the samples in each Pareto front solution set. This is an indicator of the density of sample solutions in the target space. Assume the calculated crowding distances are as follows:

[0080]

[0081] A larger crowding distance indicates that the solution is more sparse in the target space. Based on the crowding distance, several boundary solutions are selected. These solutions are located at the "edge" of the solution set, that is, the points furthest from other solutions. For example, suppose 10 boundary solutions are selected from the solution set, specifically: By calculating the gradient similarity between these boundary solutions, community migration instructions are generated. Gradient similarity is defined as a similarity measure between boundary solutions, and the specific calculation formula is as follows:

[0082]

[0083] in, Let be the gradient similarity between the m-th boundary solution and the n-th boundary solution. For the m-th boundary solution, For the nth boundary solution, and These are the maximum and minimum values ​​of the boundary solutions, respectively.

[0084] It should be noted that, based on the similarity matrix, community migration instructions are generated, indicating the migration paths of boundary solutions in the solution space. These migration instructions enable solutions with low similarity (i.e., relatively distant solutions) to exchange information, facilitating a comprehensive exploration of the solution space.

[0085] Then, based on the generated migration instructions, the cross-community migration operation is performed. This process includes the following steps:

[0086] Cross-community migration: Boundary solutions are migrated across communities according to migration instructions, that is, boundary solutions with low similarity are exchanged to different communities for further optimization and merging.

[0087] Update the solution set: After the migration operation, a new solution set is obtained. These solutions are more evenly distributed, and the solution space is explored more fully.

[0088] In this example, the first population is segmented and merged based on the distribution of the first solution set to obtain the reconstructed community structure, specifically:

[0089] Count the number of migrations in the distribution of the first solution set, and make a merging judgment based on the number of migrations;

[0090] If merging and reconstruction are performed, extract the gradient principal direction vector of each community in the first solution set distribution;

[0091] The community is segmented and merged based on the gradient principal direction vector to obtain the reconstructed community structure.

[0092] It should be noted that, firstly, the distribution of the first solution set obtained through cross-community migration optimization is analyzed, and the migration count of each solution within the solution set is counted. The migration count reflects the changes of each solution in the solution space. If a solution migrates multiple times in multiple optimization iterations, it indicates that the solution's position in the solution space is unstable and may require further adjustment or reassignment to other communities. By statistically analyzing the migration count, the stability and importance of solutions can be determined, and based on this information, a decision can be made on whether to merge communities.

[0093] Next, if the clusters of solutions with a high number of migrations exhibit similar gradient change trends (i.e., their principal gradient directions are similar), the communities can be segmented or merged by extracting the "principal gradient direction vectors" (i.e., the main directions of gradient change). The principal gradient direction vector is a mathematical description representing the dominant direction of gradient change in the solution space, and can be obtained by calculating the gradient vector of a solution (e.g., based on the gradient rate of change or other optimization metrics). If the principal gradient directions of multiple solutions are similar, the communities containing these solutions can be merged, allowing them to be optimized within the same community; conversely, if the principal gradient directions differ significantly, these solutions can be segmented to form new community structures.

[0094] Through the process of segmentation and merging, a new, reconstructed community structure is obtained, resulting in a more reasonable distribution of solutions and better meeting the optimization objective. This reconstruction process helps to improve the diversity of the solution set, avoids excessive concentration in a certain local region, and thus makes the optimization of the entire sound velocity gradient range more comprehensive and balanced.

[0095] S4, Identify gradient abrupt change regions in the interval, and expand the defect detection range based on the regions;

[0096] In this example, gradient abrupt change regions are identified within the specified interval, and the defect detection range is expanded based on these regions. Specifically:

[0097] The pipe piles are constructed into a three-dimensional spatial grid according to axial layering and radial sector division;

[0098] Within the optimal sound velocity gradient range, calculate the absolute value of the gradient change rate between adjacent spatial grids;

[0099] Gradient mutation identification is performed on the mesh based on the absolute value of the gradient change rate to generate an initial set of defect seed points;

[0100] With the preset seed point as the center, the space is expanded along the axial and radial directions of the pipe pile to obtain the expanded area;

[0101] The defect detection range is obtained by merging the extended region with the initial defect seed point set.

[0102] It should be noted that, firstly, the pipe pile structure is divided into layers with an axial length of 0.5 meters and sectors with a radial length of 30 degrees, constructing a three-dimensional spatial grid system, forming a total of 240 independent grid units in 20 axial layers and 12 radial sectors. Within the optimal sound velocity gradient range (assuming it is between 0.05 and 0.12 km / s), the absolute value of the gradient change rate between each grid unit and its adjacent units is calculated. When this value exceeds 0.08 km / s / m, it is identified as a gradient abrupt change region, thereby identifying 15 initial defect seed points. With each seed point as the center, the area is expanded spatially by 1 meter vertically and 60 degrees horizontally. Finally, the expanded area is merged with the initial seed points to form a defect detection range containing 42 grid units. This range covers the possible void defect areas on the pipe pile surface, providing a precise spatial positioning basis for the subsequent construction of defect feature maps.

[0103] S5. Construct a defect feature map based on the expanded defect detection range, locate void defects according to the defect feature map, and output the graded evaluation results.

[0104] In this example, a defect feature map is constructed based on the expanded defect detection range. The location of voids is then determined based on the defect feature map, and a graded evaluation result is output. Specifically:

[0105] The defect detection area is divided into a three-dimensional voxel grid according to a preset resolution;

[0106] Multi-band acoustic energy attenuation imaging is performed on each voxel to generate an initial defect energy map, such as... Figure 4 As shown, the energy intensity distribution in the defect space region is illustrated. The red areas represent locations with significant attenuation, where structural defects such as voids and cracks may exist.

[0107] Based on the initial defect energy map, a defect feature tensor is constructed.

[0108] A three-dimensional convolutional neural network will be used to extract the boundary features of the voided region in the defect feature tensor.

[0109] Based on the boundary characteristics of the voided area, defects are classified into three levels: minor, moderate, and severe, and the output includes defect classification assessment results containing location, size, and level.

[0110] It should be noted that, firstly, the expanded defect detection range is divided into a 5cm×5cm×5cm three-dimensional voxel grid. Each voxel unit is used for acoustic energy attenuation imaging in three frequency bands: 10kHz, 20kHz, and 30kHz. An initial defect energy map is generated by calculating the signal attenuation coefficient of each frequency band. Based on this map, a defect feature tensor containing three dimensions, namely acoustic attenuation rate, energy distribution gradient, and dispersion characteristics, is constructed. A pre-trained three-dimensional convolutional neural network (containing 5 convolutional layers and 3 fully connected layers) is used to extract defect boundary features. After the network output is classified by Softmax, the precise boundary coordinates of the voided area and the defect level score (divided into three levels: slight, moderate, and severe) are obtained. Finally, a graded evaluation report containing the defect location, size, and severity is output.

[0111] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0112] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0113] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0114] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0115] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0116] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An evaluation method for steel pipe pole caps of transmission lines based on precast pipe piles, characterized in that, Includes the following steps: Elastic waves are excited at the pile cap-pipe pile interface by an acoustic wave emitting device, and the acoustic wave signals are received by an ultrasonic array sensor at the top of the pipe pile. The acoustic signal is subjected to feature separation, and the relative deviation of the sound velocity is evaluated based on the separation results; Based on the evaluation results, the optimal sound velocity gradient range was constructed using a dynamic optimization algorithm that includes Gaussian mixture model, density peak detection algorithm and NSGA-II optimization. Identify gradient abrupt regions within the specified interval, and expand the defect detection range based on these regions. Based on the expanded defect detection range, a defect feature map is constructed. Based on the defect feature map, the void defect is located and a graded evaluation result is output. Based on the evaluation results, an optimal sound velocity gradient interval is constructed using a dynamic optimization algorithm that includes a Gaussian mixture model, a density peak detection algorithm, and NSGA-II optimization. Specifically: The relative deviation evaluation matrix of sound speed is input into the Gaussian mixture model, and several gradient communities are identified by the density peak detection algorithm. The first sound speed gradient topology map is constructed based on the spatial distribution of the gradient communities and the gradient change direction. The first constraint conditions are constructed based on the first sound velocity gradient topology graph, including the physical range constraint of the sound velocity gradient, the confidence interval constraint within each Gaussian distribution community, and the maximum rate of change constraint between adjacent sampling points. Based on the first constraint, the Monte Carlo method is used to randomly sample within the preset probability density distribution to obtain several first sound speed gradient samples. A pre-defined distributed community algorithm is used to perform NSGA-II optimization on several first sound velocity gradient samples to obtain the optimal sound velocity gradient interval. Specifically, the first sound velocity gradient samples are treated as individuals and grouped according to pre-defined community labels to obtain several first groups; the NSGA-II algorithm is used to perform non-dominated sorting on each first group to obtain several Pareto front solution sets; cross-community migration is performed on several Pareto front solution sets to obtain the first solution set distribution; the first groups are segmented and merged based on the first solution set distribution to obtain the reconstructed community structure; and NSGA-II optimization is performed again based on the reconstructed community structure to obtain the optimal sound velocity gradient interval.

2. The evaluation method for transmission line steel pipe pole caps based on precast pipe piles according to claim 1, characterized in that, The process involves exciting elastic waves at the pile cap-pipe pile interface using a sound wave emitting device, with the sound wave signals received by an ultrasonic array sensor at the top of the pipe pile. Specifically: A multi-source ring acoustic wave emission array is arranged on the surface of the pier to excite pulse signals of different frequencies according to a preset timing sequence; The ultrasonic array sensor group at the top of the pipe pile is activated simultaneously to collect the time-domain sequences of reflected and scattered waves; The time-domain sequence is divided into time windows, and the acoustic signal is extracted based on a preset wave velocity inversion model.

3. The evaluation method for transmission line steel pipe pole caps based on precast pipe piles according to claim 2, characterized in that, The process of performing feature separation on the acoustic signal and evaluating the relative deviation of the sound velocity based on the separation results specifically includes: A spatial filtering algorithm is used to reduce the noise of the acoustic signal, resulting in a denoised signal. Empirical mode decomposition is performed on the denoised signal to obtain the principal components; Extract the time-frequency eigenvectors of the principal components and calculate the offset between the time-frequency eigenvectors and the preset dispersion curve; Based on the offset, the relative deviation matrix of sound speed is calculated using a cross-correlation algorithm; The relative deviation of sound speed is evaluated based on the relative deviation matrix of sound speed, and the relative deviation of sound speed evaluation matrix is ​​obtained.

4. The evaluation method for transmission line steel pipe pole caps based on precast pipe piles according to claim 3, characterized in that, The process of performing cross-community migration on several Pareto front solutions to obtain the first solution set distribution is as follows: The crowding distance is calculated for several Pareto front solution sets, and the crowding distance is used to filter several Pareto front solution sets to obtain several boundary solutions; Calculate the gradient similarity matrix of several boundary solutions, and generate community migration instructions based on the gradient similarity matrix; According to the community migration instruction, several boundary solutions are migrated across communities to obtain the distribution of the first solution set.

5. The evaluation method for transmission line steel pipe pole caps based on precast pipe piles according to claim 4, characterized in that, The process of segmenting and merging the first group based on the distribution of the first solution set to obtain the reconstructed community structure is as follows: Count the number of migrations in the distribution of the first solution set, and make a merging judgment based on the number of migrations; If merging and reconstruction are performed, extract the gradient principal direction vector of each community in the first solution set distribution; The community is segmented and merged based on the gradient principal direction vector to obtain the reconstructed community structure.

6. The evaluation method for transmission line steel pipe pole caps based on precast pipe piles according to claim 5, characterized in that, Identify gradient abrupt change regions within the defined interval, and expand the defect detection range based on these regions, specifically as follows: The pipe piles are constructed into a three-dimensional spatial grid according to axial layering and radial sectoring. Within the optimal sound velocity gradient range, calculate the absolute value of the gradient change rate between adjacent spatial grids; Gradient mutation identification is performed on the mesh based on the absolute value of the gradient change rate to generate an initial set of defect seed points; Centered on the preset seed point, the spatial expansion is carried out along the axial and radial directions of the pipe pile to obtain the expanded area. The defect detection range is obtained by merging the extended region with the initial defect seed point set.

7. The evaluation method for transmission line steel pipe pole caps based on precast pipe piles according to claim 6, characterized in that, The process involves constructing a defect feature map based on the expanded defect detection range, locating void defects based on the defect feature map, and outputting a graded evaluation result. Specifically: The defect detection area is divided into a three-dimensional voxel grid according to a preset resolution; Multi-band acoustic energy attenuation imaging is performed on each voxel to generate an initial defect energy map; Based on the initial defect energy map, a defect feature tensor is constructed by integrating acoustic attenuation rate, energy distribution gradient and dispersion characteristics. A three-dimensional convolutional neural network will be used to extract the boundary features of the voided region in the defect feature tensor. Based on the boundary characteristics of the voided area, defects are classified into three levels: minor, moderate, and severe, and the output includes defect classification assessment results containing location, size, and level.

Images