Transmission tower deformation monitoring method and device and computer program product

Abstract

The invention discloses a power transmission tower deformation monitoring method and device and a computer program product, and the method comprises the steps: obtaining a multi-angle collection image of a random speckle pattern on the surface of a power transmission tower; constructing an image sequence based on the multi-angle acquired images, reconstructing a three-dimensional displacement field on the surface of the power transmission tower by using a stereoscopic vision reconstruction algorithm, and exporting a strain tensor field; performing time sequence processing on the strain tensor field under a plurality of time nodes through a strain evolution analysis model of a continuous time sequence, and extracting a strain change trend of each preset structure part; carrying out space consistency mapping and weighted interpolation processing according to the strain change trends of the plurality of preset structure parts, and calculating to obtain the overall structural dependent variable of the power transmission tower; and S5, in combination with historical strain data and the current overall structural dependent variable, identifying an abnormal deformation area of the power transmission tower by comparing the spatio-temporal evolution mode of the strain tensor field. The method has the effect of improving the accuracy of deformation monitoring of the power transmission tower.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular to a method and device for monitoring deformation of a transmission tower, and a computer program product. Background Art

[0002] Transmission tower deformation primarily refers to geometric changes caused by various external forces or environmental factors, such as wind loads, icing loads, earthquakes, or material creep and fatigue during long-term operation. These changes manifest as bending, twisting, tilting, joint cracking, or loose connections in the tower columns or crossarms, leading to axial displacement, lateral offset, or height compression of the tower. This deformation not only affects the stability of transmission line conductor sag, electrical clearances, and electrical safety distances, but can also lead to a range of serious consequences, including insulation coordination failure, line tripping, or structural damage.

[0003] An existing method for monitoring transmission tower deformation involves attaching resistance strain gauges to key locations on the tower. The local strain values are inferred by measuring the changes in strain gauge resistance as the structure deforms. This method offers advantages such as simplicity, responsiveness, and low cost. However, due to its point-based monitoring nature, it cannot capture the global strain distribution of the tower structure and can only reflect local deformation information at a limited number of measurement points. Furthermore, in actual operation, it is susceptible to factors such as ambient temperature fluctuations, decreased bonding stability, and signal acquisition noise, resulting in insufficient long-term monitoring accuracy and reliability. Summary of the Invention

[0004] The technical problem to be solved by the embodiments of the present invention is to provide a method, device and computer program product for monitoring deformation of a transmission tower, so as to improve the accuracy of monitoring deformation of a transmission tower.

[0005] In order to solve the above technical problems, the present invention provides a method for monitoring deformation of a transmission tower, comprising:

[0006] Step S1, using multi-view digital image acquisition devices deployed at different preset structural parts of the transmission tower, synchronously acquiring multi-angle captured images of random speckle patterns on the surface of the transmission tower;

[0007] Step S2, constructing an image sequence based on the synchronously acquired multi-angle images, reconstructing a three-dimensional displacement field of the transmission tower surface using a stereo vision reconstruction algorithm, and deriving a strain tensor field based on the three-dimensional displacement field;

[0008] Step S3, performing time series processing on the strain tensor field at multiple time nodes through a continuous time series strain evolution analysis model to extract the strain change trend of each preset structural part;

[0009] Step S4, performing spatial consistency mapping and weighted interpolation processing according to the strain variation trends of the plurality of preset structural parts to calculate the overall structural strain of the transmission tower;

[0010] Step S5: combining the historical strain data with the current overall structural strain, and comparing the spatiotemporal evolution pattern of the strain tensor field to identify abnormal deformation areas of the transmission tower.

[0011] Preferably, the step S2 specifically includes:

[0012] The images collected at different angles at the same time are combined into an image sequence according to the spatial configuration relationship. Each set of image sequences corresponds to the surface state of the transmission tower at a time point.

[0013] Applying a stereo vision reconstruction algorithm to perform dense stereo matching on the image pairs with parallax in the image sequence, converting the matched disparity map into a three-dimensional point cloud through triangulation, and constructing a three-dimensional geometric model of the tower structure surface;

[0014] Obtaining the spatial displacement vectors of each of the preset structural parts at different times by three-dimensional point cloud registration and surface correspondence tracking at consecutive time points;

[0015] For any of the preset structural parts, based on the spatial displacement vectors at different times, a deformation field is constructed by spatial interpolation using least squares fitting for a local area in the three-dimensional point cloud, and a full-order strain tensor is derived from the displacement gradient to construct a local deformation gradient tensor of the neighborhood of the preset structural part in continuous space;

[0016] A strain tensor field of the transmission tower is constructed based on the local deformation gradient tensors of all the preset structural parts.

[0017] Preferably, for any of the preset structural parts, based on the spatial displacement vectors at different moments, a deformation field is constructed by spatial interpolation using least squares fitting for a local area in the three-dimensional point cloud, and a full-order strain tensor is derived from the displacement gradient, thereby constructing a local deformation gradient tensor of the neighborhood of the preset structural part in the continuous space, wherein the full-order strain tensor is expressed as follows:

[0018]

[0019] Among them, ε ij is the i-th row and j-th column component of the full-order strain tensor, u i is the displacement component of the preset structural part in the i-th direction during the deformation process, u j is the displacement component of the preset structural part in the jth direction during the deformation process, x iis the i-th coordinate component of the preset structural part in the undeformed state, x j is the jth coordinate component of the preset structural part in the undeformed state, represents the first-order partial derivative of the i-th displacement component with respect to the j-th coordinate component, represents the first-order partial derivative of the j-th displacement component with respect to the i-th coordinate component, is the geometric nonlinear term.

[0020] Preferably, the step S3 specifically includes:

[0021] defining a set of continuous discrete time nodes to form a strain data sequence containing time tags, wherein each of the discrete time nodes corresponds to an image captured by a multi-view digital image acquisition device at any moment of the preset structural part;

[0022] For each of the discrete time nodes, obtaining the strain tensor field of the preset structural part;

[0023] Normalizing and mapping the plurality of strain tensor fields in a unified spatial coordinate system;

[0024] For multiple strain tensor fields of normalized mapping, time series data are extracted, and the strain change rate is further extracted by fitting method.

[0025] Preferably, the step S4 specifically includes:

[0026] The strain variation trends of the plurality of preset structural parts are uniformly projected into the global coordinate system of the transmission tower through spatial consistency mapping, thereby forming strain data that is spatially distributed, continuous, and temporally sequential;

[0027] At each monitoring time node, the strain change trends of each preset structural part are jointly analyzed, and a mapping relationship from local strain to global strain is constructed using a method based on weighted spatial interpolation to obtain the overall strain of the structure. The weight parameters are set based on the force flow proportion of each part in the structural force system, the node importance coefficient and the density of the structural energy dissipation position.

[0028] Preferably, the step S5 specifically includes:

[0029] Constructing a complete spatiotemporal strain tensor database of the transmission tower over its structural history, the spatiotemporal strain tensor database covering strain evolution data of each preset structural part under multiple operating cycles, environmental changes, and load response conditions;

[0030] Pairing the strain evolution data with the overall structural strain at the corresponding time node for analysis, forming a normal operating mode of the structural response through spatial mapping and time series modeling, wherein the normal operating mode is expressed in a multidimensional feature space, including the mean of the principal strain, the rate of change, the direction of the maximum shear strain, the amplitude of the strain gradient mutation, and the coupling coefficient between the strain evolution data and the overall structural strain;

[0031] Comparing each of the strain tensor fields at the current time node with the strain tensor corresponding to the normal operating mode point by point, and measuring the degree of deviation of the strain tensor field of the preset structural part from the historical mode;

[0032] Initially screen the pre-set structural parts with deviations greater than or equal to the pre-set threshold and mark them as potential abnormal areas;

[0033] In combination with the overall structural strain of the transmission tower and the strain evolution data, analyzing whether there is non-cooperative change between the overall structural deformation and the local strain of the potential abnormal area;

[0034] If it is determined that there is a non-cooperative change between the overall structural deformation and the local strain of the potential abnormal area, the principal direction change of the strain tensor of the potential abnormal area is combined with the evolution of the tensor eigenvalues and eigenvectors to capture the sudden change of the local deformation mode and evaluate the influence of the potential abnormal area on the overall structural response;

[0035] The key potential deformation risk points with global structural impact are screened according to the impact degree.

[0036] Preferably, the step S1 specifically includes:

[0037] Arranging random speckle patterns at different predetermined structural locations, and installing a plurality of digital image acquisition devices in a multi-viewing manner around the random speckle patterns;

[0038] Image synchronization is achieved based on timing control or triggering mechanisms, and the initial image is obtained through edge computing nodes or wireless transmission modules in real time.

[0039] The acquisition system is calibrated to eliminate lens distortion errors, and time series filtering and image enhancement algorithms are introduced to process the initial image to obtain the multi-angle acquisition image.

[0040] The present invention also provides a transmission tower deformation monitoring device, comprising:

[0041] An acquisition module is configured to synchronously acquire multi-angle images of random speckle patterns on the surface of the transmission tower using multi-angle digital image acquisition devices arranged at different preset structural locations of the transmission tower;

[0042] A processing module is configured to construct an image sequence based on the synchronously acquired multi-angle images, reconstruct a three-dimensional displacement field on the surface of the transmission tower using a stereoscopic vision reconstruction algorithm, and derive a strain tensor field based on the three-dimensional displacement field; perform time series processing on the strain tensor field at multiple time nodes using a continuous time series strain evolution analysis model to extract strain change trends at each preset structural part; perform spatial consistency mapping and weighted interpolation processing based on the strain change trends of the multiple preset structural parts to calculate the overall structural strain of the transmission tower;

[0043] The output module is used to combine historical strain data with the current overall structural strain and identify abnormal deformation areas of the transmission tower by comparing the spatiotemporal evolution patterns of the strain tensor field.

[0044] The present invention also provides a transmission tower deformation monitoring device, comprising:

[0045] one or more processors;

[0046] Memory;

[0047] One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, and the one or more applications are configured to execute the transmission tower deformation monitoring method.

[0048] The present invention also provides a computer program product, comprising computer instructions, wherein the computer instructions instruct a computer device to execute operations corresponding to the method.

[0049] The present invention has the following beneficial effects: By deploying multi-view digital image acquisition equipment at key structural locations of transmission towers, combined with surface speckle patterns and stereoscopic reconstruction algorithms, it achieves high-resolution reconstruction of the tower's three-dimensional surface geometry. This technology overcomes the limitations of traditional point-based strain gauge measurement and simultaneously acquires the full-field displacement field and strain tensor field of the entire tower structure, enabling continuous deformation perception of the transmission tower from local points to the entire structure. Using multi-period point cloud registration technology in a unified spatial coordinate system, the present invention performs spatiotemporal mapping of strain data at different time points, forming a temporally continuous strain evolution database. This technology accurately captures subtle deformation gradient changes and enables dynamic tracking of complex operating conditions such as long-term structural degradation and sudden load response. By integrating structural force flow paths, node importance, and energy dissipation distribution parameters to set weight coefficients, the present invention constructs a global strain model that accurately reflects the overall deformation trend of the structure, providing a reliable quantitative indicator for structural safety assessment. By constructing a spatiotemporal strain tensor library for normal operating modes and combining point-by-point comparison analysis with tracking of the tensor's principal direction evolution, the present invention can accurately identify abnormal areas that deviate from threshold values. The innovative introduction of strain direction synergy analysis effectively distinguishes local anomalies from global effects, and can quantitatively assess the impact of potential risks on the overall performance of the structure. This invention realizes a complete closed loop from data acquisition, model construction to risk assessment: the front end obtains high-fidelity data through a multimodal sensor network, the middle platform uses a spatiotemporal coupling algorithm for intelligent analysis, and the back end establishes a diagnostic model based on physical mechanisms. It breaks through the limitations of traditional monitoring technology in spatial coverage, temporal continuity, and anomaly recognition capabilities, and provides a full life cycle health monitoring solution for transmission towers, significantly reducing operation and maintenance costs while improving the structural safety level. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1 The present invention is a flowchart of a method for monitoring deformation of a transmission tower according to a first embodiment of the present invention.

[0052] Figure 2 It is a structural schematic diagram of a transmission tower deformation monitoring device according to a second embodiment of the present invention.

[0053] Figure 3 It is a structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The following descriptions of the embodiments refer to the accompanying drawings to illustrate specific embodiments in which the present invention may be implemented.

[0055] Please refer to Figure 1 As shown, the first embodiment of the present invention provides a method for monitoring deformation of a transmission tower, comprising:

[0056] Step S1, using multi-view digital image acquisition devices deployed at different preset structural parts of the transmission tower, synchronously acquiring multi-angle captured images of random speckle patterns on the surface of the transmission tower;

[0057] Step S2, constructing an image sequence based on the synchronously acquired multi-angle images, reconstructing a three-dimensional displacement field of the transmission tower surface using a stereo vision reconstruction algorithm, and deriving a strain tensor field based on the three-dimensional displacement field;

[0058] Step S3, performing time series processing on the strain tensor field at multiple time nodes through a continuous time series strain evolution analysis model to extract the strain change trend of each preset structural part;

[0059] Step S4, performing spatial consistency mapping and weighted interpolation processing according to the strain variation trends of the plurality of preset structural parts to calculate the overall structural strain of the transmission tower;

[0060] Step S5: combining the historical strain data with the current overall structural strain, and comparing the spatiotemporal evolution pattern of the strain tensor field to identify abnormal deformation areas of the transmission tower.

[0061] Specifically, a method for monitoring the deformation of a transmission tower according to an embodiment of the present invention is applied to a server. The server includes, but is not limited to, electronic devices such as mobile phones, tablet computers, wearable devices, and personal computers (PCs). It can also be a backend server that runs a method for monitoring the deformation of a transmission tower based on three-dimensional full-field strain measurement. The server can be implemented as a standalone server or a server cluster consisting of multiple servers.

[0062] Step S1 acquires a captured image of a random speckle pattern on the surface of the transmission tower using multi-view digital image acquisition devices arranged at different preset structural positions of the transmission tower.

[0063] In one possible implementation, multi-perspective digital image acquisition devices are deployed at different preset structural locations of a transmission tower to acquire images of random speckle patterns on the surface of the transmission tower. This method specifically includes: deploying random speckle patterns at different preset structural locations and installing multiple digital image acquisition devices in a multi-perspective manner around the random speckle patterns; synchronously acquiring images based on a timing control or triggering mechanism, and acquiring an initial image for real-time transmission via an edge computing node or wireless transmission module; and calibrating the acquisition system to eliminate lens distortion errors. Temporal filtering and image enhancement algorithms are then introduced to process the initial image to obtain the acquired image.

[0064] Specifically, when arranging random speckle patterns in different preset structural parts, we should first select locations that have a significant impact on the overall stability, such as the connection area between the tower column and the crossarm, the foundation anchor section or the lateral support structure, based on the force distribution of the tower, historical failure data and stability analysis of key nodes. We should use a high-contrast coating material with weather resistance and adhesion to spray or attach a randomly distributed black and white speckle pattern on its surface. The pattern should have sufficient texture complexity and spatial resolution to meet the displacement recognition requirements of the digital image correlation algorithm.

[0065] Multi-view image acquisition devices are then deployed around the pattern area. These devices need to be fixedly installed at positions where the three-dimensional structure is reasonably distributed to form a sufficient parallax baseline. The shooting perspectives should meet the geometric constraints of three-dimensional reconstruction, and calibration targets or reference points should be reserved for subsequent image alignment.

[0066] When achieving synchronous image acquisition based on timing control or triggering mechanisms, the image acquisition system should be configured with a unified clock synchronization module or use an external trigger signal to enable all imaging devices to capture the current frame image at the same time, ensuring that images from each perspective have temporal consistency to support time series analysis of dynamic deformation.

[0067] The collected image data can be locally compressed and encoded by edge computing nodes deployed on the tower body or base, and transmitted to the background processing center or cloud platform using wireless communication modules such as 4G, 5G or LoRa. During the transmission process, error detection and adaptive transmission protocols should be used to ensure the integrity and real-time nature of the image data, and the collection frequency can be automatically adjusted based on the communication status to achieve a balance between energy efficiency and data stability.

[0068] When eliminating lens distortion errors by calibrating the acquisition system, a high-precision calibration board and stereo calibration method should be used after installation to obtain the intrinsic and extrinsic parameters of each camera, including focal length, principal point position, distortion coefficient, and relative position matrix, and then a unified spatial coordinate system is constructed to achieve geometric correction and projection mapping between images.

[0069] In the image preprocessing stage, temporal filtering algorithms such as Kalman filtering or adaptive median filtering are introduced to remove background illumination disturbances and image jitter. At the same time, image enhancement techniques such as histogram equalization, multi-scale sharpening and noise suppression are applied to improve the quality of the original image, thereby generating acquired images with high signal-to-noise ratio and geometric consistency, providing a reliable input basis for subsequent strain field solution and three-dimensional reconstruction.

[0070] In step S2, an image sequence is formed based on the synchronously acquired images at different angles, and a three-dimensional displacement field of the tower structure surface is reconstructed using a stereo vision reconstruction algorithm, and the strain tensor field is further derived.

[0071] In one possible embodiment, an image sequence is formed based on synchronously acquired images from different angles, and a three-dimensional displacement field of the tower structure surface is reconstructed using a stereo vision reconstruction algorithm, and a strain tensor field is further derived. The method specifically includes: arranging images acquired from different angles at the same time into an image sequence according to a spatial configuration relationship, with each image sequence corresponding to the surface state of the transmission tower at a time point; applying the stereo vision reconstruction algorithm to perform dense stereo matching on image pairs with parallax in the image sequence, converting the matched disparity map into a three-dimensional point cloud through triangulation, thereby constructing a three-dimensional geometric model of the tower structure surface; obtaining the spatial displacement vectors of each preset structural part at different times by aligning the three-dimensional point clouds and tracking the surface correspondence relationship at consecutive time points; for any preset structural part, based on the spatial displacement vectors at different times, constructing a deformation field by spatial interpolation of a local area in the three-dimensional point cloud using least squares fitting, and deriving a full-order strain tensor from the displacement gradient, thereby constructing a local deformation gradient tensor of the neighborhood of the preset structural part in continuous space; and constructing the strain tensor field of the transmission tower based on the local deformation gradient tensors of all preset structural parts.

[0072] Specifically, when images captured at different angles at the same time are combined into an image sequence according to a spatial configuration relationship, the internal and external parameters of each image acquisition device should be jointly calibrated first to obtain its position and orientation information in a unified three-dimensional coordinate system, and the captured images should be grouped according to the timestamp of image acquisition so that each group of images corresponds to the surface state of the transmission tower at the same time.

[0073] Then, based on the calibration information, the group of images is organized into a structured image sequence according to the camera perspective order and parallax coverage relationship, providing geometric consistency guarantee for subsequent stereo matching and three-dimensional reconstruction. At the same time, time tags and image quality evaluation indicators can be introduced to dynamically manage the image sequence.

[0074] When applying a stereo vision reconstruction algorithm, a dense matching strategy is used to construct a cost volume for image pairs with parallax in the image sequence. Disparity maps are then obtained through cost aggregation and path optimization algorithms such as semi-global matching or a stereo matching network based on a deep convolutional network. Subsequently, the 3D spatial coordinates corresponding to each pixel are reconstructed based on the disparity values, camera parameters, and triangulation formulas to form a dense 3D point cloud model. The accuracy of this point cloud is determined by the accuracy of the disparity estimation and camera calibration. If necessary, point cloud filtering and edge enhancement strategies can be introduced to improve the quality of boundary reconstruction, ultimately resulting in a 3D geometric model of the tower structure surface.

[0075] When obtaining displacement vectors through 3D point cloud registration and surface correspondence tracking at consecutive time points, it is necessary to first perform initial rigid alignment on the 3D point clouds of consecutive time frames. Correspondence between points is established based on local feature descriptors such as FPFH or feature point matching. Then, fine registration is performed between the point clouds using iterative closest point (ICP) or non-rigid registration algorithms. After establishing one-to-one correspondence, the spatial displacement vectors of each preset structural component are extracted by calculating the spatial position change of the same point at different times. Kalman filtering or Bayesian filtering can be used to suppress measurement noise and improve trajectory smoothness, ensuring the stability and continuity of time series data.

[0076] For any pre-set structural part, based on the spatial displacement vector, a neighborhood model is constructed for the local area of its 3D point cloud. The least squares fitting algorithm is then used to calculate the displacement change rate of each point in the neighborhood, i.e., the deformation gradient tensor. This fitting process obtains the deformation gradient tensor by establishing a local coordinate system in the neighborhood and solving the partial derivatives of the displacement with respect to the coordinates. Furthermore, the strain state of the area is derived using the following full-order strain tensor expression:

[0077]

[0078] Among them, ε ij is the i-th row and j-th column component of the full-order strain tensor, which represents the strain in the j-th direction along the i-th direction. It is the basic quantity that describes the local deformation state of the material. i is the displacement component of the preset structural part in the i-th direction during the deformation process, u j is the displacement component of the preset structural part in the jth direction during the deformation process, x i is the i-th coordinate component of the preset structural part in the undeformed state, x j is the jth coordinate component of the preset structural part in the undeformed state, represents the first-order partial derivative of the i-th displacement component with respect to the j-th coordinate component, represents the first-order partial derivative of the j-th displacement component with respect to the i-th coordinate component, both reflecting the contribution of linear deformation; It is a geometric nonlinear term that takes into account the second-order variation under large deformation, which is important for capturing non-uniform deformation and rotation.

[0079] The above formula is a component expression of the full-order strain tensor, and its calculation principle is based on the change in the displacement field of a preset structural part from the initial state to the deformed state. The formula consists of linear strain terms and nonlinear geometric terms, and can accurately capture complex deformation behaviors including shear, stretching, compression, and torsion. Overall, the tensor reflects the deformation state of a small neighborhood of a point in three-dimensional space, has rotational invariance and symmetry, and can truly reflect the strain response of the actual structure under load. In the implementation process, by calculating the spatial gradient of the displacement field in high-density three-dimensional point cloud data, the εij value at each point can be solved, and then the strain tensor field of the entire structure can be constructed.

[0080] Based on the local deformation gradient tensors of all pre-defined structural components, a spatial grid mapping or interpolation model is established for the tower's overall 3D point cloud model. The local tensor results are mapped into a unified 3D coordinate system. A continuous representation of the strain tensor field is achieved through tensor field reconstruction algorithms, such as high-dimensional tensor interpolation or spatial mapping methods based on finite element meshes. Furthermore, the temporal dimension can be incorporated to construct a temporally and spatially distributed strain tensor evolution map, enabling full-field dynamic monitoring and high-resolution strain analysis of the structural response of key components of the transmission tower throughout its service life, providing a precise basis for structural health diagnosis and early damage identification.

[0081] Step S3 introduces a strain evolution analysis model based on continuous time series, and extracts the strain change trend of key structural positions by constructing the strain tensor field at multiple time nodes.

[0082] In one possible implementation, a strain evolution analysis model based on a continuous time series is introduced, and the strain change trend of key structural positions is extracted by constructing a strain tensor field at multiple time nodes. Specifically, the model includes: defining a set of continuous discrete time nodes to form a strain data sequence containing time tags, wherein each discrete time node corresponds to an image captured by a multi-view digital image acquisition device at any moment of a preset structural part; for each discrete time node, the strain tensor field of the preset structural part is obtained; the multiple strain tensor fields are normalized and mapped in a unified spatial coordinate system; for the multiple normalized mapped strain tensor fields, time series data is extracted, and the strain change rate is further extracted through a fitting method.

[0083] Specifically, when defining a set of continuous discrete time nodes, the image acquisition interval should be set based on the operating characteristics of the transmission tower structure, the external load variation cycle, and the sampling capacity of the acquisition system. For example, continuous acquisition should be performed in units of minutes, hours, or days to ensure that the response evolution of the structure under normal and abnormal conditions is covered. Each discrete time node corresponds to a synchronous multi-view image acquisition action. After synchronous processing, the image is input into the stereo reconstruction module, thus corresponding to the deformation state of the tower structure at that moment. To ensure sequence integrity, each time node is uniquely time-tagged and its acquisition parameters, environmental conditions, and image quality information are archived to construct a time-driven strain data index system.

[0084] At each discrete time node, stereo matching and 3D reconstruction are performed based on the multi-view image sequence to obtain a dense 3D point cloud of the preset structural parts. Based on this point cloud, a strain tensor field is constructed through displacement solution, deformation gradient calculation, and tensor derivation. This tensor field includes six components, representing the normal strain and shear strain response of the structure in the three principal axes, and comprehensively reflects the microscopic deformation state of the local structural unit in any direction in space. Corresponding local strain tensor subfields can be constructed for multiple key parts and uniformly recorded in a multidimensional tensor database with time as the principal axis, which serves as the basic data source for strain evolution analysis.

[0085] When normalizing and mapping multiple strain tensor fields in a unified spatial coordinate system, a global spatial coordinate system should first be established based on the tower's three-dimensional model. The tensor fields corresponding to each time node should be registered and aligned using the coordinate transformation matrix used during reconstruction to eliminate the effects of spatial position offsets caused by acquisition angle and parallax fluctuations. Subsequently, the tensor values at each time node are amplitude normalized and scaled to eliminate fluctuations caused by non-structural factors such as illumination and environmental interference, resulting in a continuous tensor time series with constant structural spatial position and consistent tensor definitions. If necessary, local reference points or a rigid skeleton can be introduced as an alignment basis to improve spatial mapping accuracy.

[0086] For multiple strain tensor fields mapped from normalized maps, representative tensor components at key locations of the structure, such as the maximum principal strain or shear strain index, are selected and their variation at consecutive time nodes is extracted to form a strain time series. This series can be smoothed and trend-modeled using fitting methods such as polynomial regression, spline interpolation, or weighted sliding average. The first-order derivative is then calculated to obtain the strain change rate, identifying changes in the deformation strength of the structure over a specific time period. Furthermore, the amplitude of the rate change can be compared to set thresholds to identify strain mutation points or potential fatigue concentration areas, providing core evolutionary characteristic indicators for subsequent anomaly identification, lifespan assessment, and condition warning.

[0087] Step S4 analyzes the overall structural strain of the transmission tower based on the strain variation trends of multiple preset structural parts. Specifically, this includes: uniformly projecting the strain variation trends of multiple preset structural parts into the global coordinate system of the transmission tower through spatial consistency mapping to form strain data that is spatially distributed, continuous, and temporally sequential; jointly analyzing the strain variation trends of each preset structural part at each monitoring time node, and constructing a mapping relationship from local strain to global strain using a weighted spatial interpolation method to obtain the overall structural strain. The weight parameters are set based on the force flow proportion of each part in the structural force system, the node importance coefficient, and the density of structural energy dissipation locations.

[0088] Specifically, when projecting the strain variation trends of multiple pre-set structural parts into the global coordinate system of the transmission tower through spatial consistency mapping, a unified global spatial coordinate system must first be established based on a high-precision three-dimensional geometric model of the tower. The position of each pre-set structural part within this coordinate system is then spatially calibrated so that the corresponding strain data has clear spatial position attributes. Subsequently, the strain tensors or representative principal strain indicators obtained at different times are aligned through coordinate transformation and rigid registration algorithms. The local strain trends are mapped into a unified structural framework and organized into a spatiotemporal tensor data structure based on continuous time series. This results in a multi-part strain data matrix that is continuously distributed in space and dynamically evolves in time, providing a basis for overall structural response analysis.

[0089] When jointly analyzing the strain change trends of each preset structural part at each monitoring time node, it is necessary to extract the representative strain value or its derivative information of each part at a specific time node and construct an interpolation network based on the spatial distribution density. Using weighted spatial interpolation methods such as Gaussian kernel interpolation or radial basis function interpolation algorithms, all local strain variables are expanded to the entire domain of the structure, and a mapping relationship from local discrete points to continuous fields is constructed. The formula for the constructed overall structural strain variable is as follows:

[0090]

[0091] in, represents the overall strain of the structure at the tth time node, N is the total number of preset structural parts, w n is the weighted coefficient of the nth preset structural part, The model is essentially a digital modeling of the macroscopic deformation state of the overall structure by synthesizing the temporal and spatial trends of multiple physically representative local strain changes.

[0092] Determining the weight parameters is crucial in this process. The results of the structural static analysis should be combined with the assessment of the force flow transmission path to quantify the contribution of each monitored location to the overall mechanical system. Node importance coefficients, such as connectivity, path centrality, and strain sensitivity under historical loads, should also be incorporated. Furthermore, the energy dissipation location density—the proportion of energy dissipated by each location during the load evolution of the structure—is considered, and the weights are further modified to ensure that the interpolation results accurately reflect the true deformation-dominant regions and the core mechanical response areas of the tower structure.

[0093] Finally, by weighting the normalized strain trends of all parts and their corresponding structural function weights, a time-series curve representing the overall strain of the tower structure can be constructed. This overall strain can be used to reflect the changing trend of the structure's macroscopic stress state and provide a quantitative basis for fatigue assessment, risk warning, and maintenance strategies.

[0094] Step S5 combines the historical strain data with the current overall structural strain, and identifies abnormal deformation areas of the transmission tower by comparing the spatiotemporal evolution patterns of the strain tensor field.

[0095] In one possible implementation, the historical strain data and the overall strain of the structure are combined to identify abnormal deformation areas, specifically including: constructing a complete strain tensor spatiotemporal database of the transmission tower structure in the historical period, the strain tensor spatiotemporal database covering the strain evolution data of each preset structural part under multiple operating cycles, environmental changes and load response conditions; pairing the strain evolution data with the overall strain of the structure at the corresponding time node for analysis, and forming a normal operating mode of the structural response through spatial mapping and time series modeling. The normal operating mode is expressed in a multi-dimensional feature space, including the mean of the principal strain, the rate of change, the direction of the maximum shear strain, the amplitude of the strain gradient mutation, and the coupling coefficient between the strain evolution data and the overall strain of the structure; and comparing each strain tensor field at the current time node with the strain tensor field at the current time node. The strain tensors corresponding to the normal operating mode are compared point by point to measure the degree of deviation of the strain tensor field of the preset structural parts from the historical pattern; the preset structural parts with a deviation greater than or equal to the preset threshold are preliminarily screened and marked as potential abnormal areas; the overall structural strain and strain evolution data of the transmission tower are combined to analyze whether there are non-cooperative changes between the overall structural deformation and the local strain of the potential abnormal area; if it is determined that there are non-cooperative changes between the overall structural deformation and the local strain of the potential abnormal area, the main direction change of the strain tensor in the potential abnormal area is combined with the evolution of the tensor eigenvalue and eigenvector to capture the sudden change of the local deformation mode and evaluate the degree of influence of the potential abnormal area on the overall structural response; key potential deformation risk points with global structural influence are screened according to the degree of influence.

[0096] Specifically, to construct a complete spatiotemporal database of strain tensors over the transmission tower's structural history, three-dimensional strain tensor data for each predefined structural location over multiple time periods must be extracted based on long-term, continuously acquired, multi-view images and their reconstruction results. These data are then organized into a high-dimensional spatiotemporal tensor database, organized chronologically and spatially, along with the structural location. This database should cover a wide range of operating conditions, including normal operating conditions, extreme weather conditions, temperature fluctuations, wind load variations, and historical maintenance cycles, ensuring temporal integrity and environmental diversity. Each data point should include the tensor value, principal strain direction, eigenvalue combination, local coordinates, and corresponding acquisition time tag, forming a unified data management structure to provide computability support for subsequent comparison and modeling.

[0097] When pairing strain evolution data with the overall structural strain at corresponding time points, a one-to-one correspondence between the strain tensor field and the overall strain value should be established for each historical time point. Local strain characteristics should be mapped to the overall response index space through spatial interpolation or regional reduction. Based on this, a multidimensional feature space is constructed, including the local principal strain mean, strain growth rate, magnitude of change in the maximum shear strain direction, and the correlation coefficient between the degree of local abrupt change in strain gradient and the change in the overall strain value. Through the combined statistical modeling of these indicators, methods such as Gaussian mixture models, autoencoders, or principal component analysis are used to learn the structural response pattern under normal conditions, forming a standard response domain that can be used for anomaly identification.

[0098] When comparing each strain tensor field at the current time node with the normal operating mode, the current tensor data should first be aligned with the historical model, including tensor coordinate mapping, unit normalization, and directional consistency processing. Subsequently, using methods such as Mahalanobis distance, tensor similarity measurement, or dynamic time series comparison, the degree of deviation from the normal mode in terms of principal strains, eigenvalue combinations, and orientation angles for each structural component is calculated. This results in a point-by-point map of abnormal deviation indicators across the entire tower, which provides spatial distribution information and time-dependent characteristics.

[0099] When initially screening for structural parts with deviations greater than or equal to a preset threshold, statistical or engineering experience thresholds should be set based on the deviation index map, such as the upper limit of the confidence interval, a percentage threshold exceeding two standard deviations from the mean, or a historical maximum. Structural parts that meet the conditions are marked and aggregated into potential abnormal areas based on their adjacent relationships in the structural topology, achieving preliminary identification from isolated abnormal points to regional deformation aggregates, which are then recorded and visualized.

[0100] When analyzing whether there are non-synergistic changes between the local strain and the overall deformation in the abnormal region, combined with the overall strain evolution data, the strain evolution trend in the abnormal region should be cross-referenced with the overall strain curve for the same time period to determine whether there is a spatial-temporal offset, such as a sudden increase in local strain while the overall strain changes gradually, or persistent high local strain during overall recovery. Such non-synergistic relationships can be quantified by calculating the difference between the local and global strain derivatives or the sudden change in the correlation coefficient, providing a basis for further determination of the abnormal mechanism.

[0101] If non-synergistic changes are confirmed, further analysis of the principal direction changes of the strain tensor in the potential abnormal region is necessary. By tracking the evolution of the tensor eigenvalues and corresponding eigenvectors at multiple time points in this region, it is possible to determine whether there has been a sudden change in deformation mode, such as a jump in the principal strain direction, a rotation of the tensor's principal axis, or an increase in shear strain dominance. A function of the tensor's direction change angle can be constructed to assess its degree of deviation from the structural geometry. Furthermore, the influence coefficient on the overall stiffness field, force transmission chain, and dynamic response can be calculated by combining the relationship between the abnormal region's location and the force path.

[0102] Ultimately, when screening key potential deformation risk points based on impact, a comprehensive assessment index is established. This weighted superposition of deviation amplitude, non-cooperativity, directional mutation, and global response sensitivity is used to output a risk score for each potential abnormal area. By setting risk level thresholds, core deformation areas with high system impact are screened as key target points for subsequent structural repair, reinforcement, or monitoring priorities. Their evolution trajectory and strain history can be automatically annotated in the system for decision-making reference.

[0103] Please refer to Figure 2 As shown, the second embodiment of the present invention provides a transmission tower deformation monitoring device, including an acquisition module 201, a processing module 202 and an output module 203, wherein:

[0104] The acquisition module 201 is used to synchronously acquire multi-angle images of random speckle patterns on the surface of the transmission tower based on multi-view digital image acquisition devices deployed at different preset structural parts of the transmission tower;

[0105] The processing module 202 is used to construct an image sequence based on the synchronously acquired multi-angle images, reconstruct the three-dimensional displacement field of the transmission tower surface using a stereoscopic vision reconstruction algorithm, and derive a strain tensor field based on the three-dimensional displacement field; perform time series processing on the strain tensor field at multiple time nodes using a continuous time series strain evolution analysis model to extract the strain change trend of each preset structural part; perform spatial consistency mapping and weighted interpolation processing based on the strain change trends of the multiple preset structural parts to calculate the overall structural strain of the transmission tower;

[0106] The output module 203 is used to combine the historical strain data with the current overall structural strain, and identify abnormal deformation areas of the transmission tower by comparing the spatiotemporal evolution patterns of the strain tensor field.

[0107] In a possible implementation, the processing module 202 is configured to group the images collected at different angles at the same time into an image sequence according to a spatial configuration relationship, and each image sequence corresponds to the surface state of the transmission tower at a time point.

[0108] The processing module 202 is used to apply a stereo vision reconstruction algorithm to perform dense stereo matching on image pairs with parallax in the image sequence, and convert the matched disparity map into a three-dimensional point cloud through triangulation, thereby constructing a three-dimensional geometric model of the tower structure surface.

[0109] The acquisition module 201 is used to obtain the spatial displacement vectors of each preset structural part at different times through three-dimensional point cloud registration and surface correspondence tracking at consecutive time points.

[0110] The processing module 202 is used to construct a deformation field for any preset structural part based on the spatial displacement vectors at different times by using least squares fitting spatial interpolation for the local area in the three-dimensional point cloud, and derive the full-order strain tensor from the displacement gradient, thereby constructing the local deformation gradient tensor of the neighborhood of the preset structural part in the continuous space.

[0111] The processing module 202 is used to construct a strain tensor field of the transmission tower according to the local deformation gradient tensors of all preset structural parts.

[0112] In one possible implementation, the processing module 202 is configured to construct a deformation field for any predetermined structural part by spatially interpolating the local area in the three-dimensional point cloud using least squares fitting based on the spatial displacement vectors at different times, and derive a full-order strain tensor from the displacement gradient, thereby constructing a local deformation gradient tensor in the neighborhood of the predetermined structural part in the continuous space, where the full-order strain tensor is expressed as follows:

[0113]

[0114] Among them, ε ij is the i-th row and j-th column component of the full-order strain tensor, which represents the strain in the j-th direction along the i-th direction. It is the basic quantity that describes the local deformation state of the material. i is the displacement component of the preset structural part in the i-th direction during the deformation process, u j is the displacement component of the preset structural part in the jth direction during the deformation process, x i is the i-th coordinate component of the preset structural part in the undeformed state, x j For the preset structural parts,

[0115] Deformation and rotation are of great significance.

[0116] In one possible implementation, the processing module 202 is used to define a set of continuous discrete time nodes to form a strain data sequence containing time tags, wherein each discrete time node corresponds to an image captured by a multi-view digital image acquisition device at any moment of a preset structural part.

[0117] The acquisition module 201 is used to acquire the strain tensor field of a preset structural part for each discrete time node.

[0118] The processing module 202 is used to perform normalized mapping on the multiple strain tensor fields in a unified spatial coordinate system.

[0119] The processing module 202 is used to extract time series data from the multiple strain tensor fields of the normalized mapping, and further extract the strain change rate through a fitting method.

[0120] In a possible implementation, the processing module 202 is configured to uniformly project the strain variation trends of multiple preset structural parts into the global coordinate system of the transmission tower through spatial consistency mapping, thereby forming strain data that is spatially distributed, continuous, and temporally sequential.

[0121] The processing module 202 is used to jointly analyze the strain change trends of each preset structural part at each monitoring time node, and adopt a method based on weighted spatial interpolation to construct a mapping relationship from local strain to global strain, so as to obtain the overall strain of the structure. Among them, the weight parameters are set according to the force flow proportion of each part in the structural force system, the node importance coefficient and the density of the structural energy dissipation position.

[0122] In one possible implementation, the processing module 202 is used to construct a complete strain tensor spatiotemporal database of the transmission tower within its structural history period. The strain tensor spatiotemporal database covers strain evolution data of each preset structural part under multiple operating cycles, environmental changes, and load response conditions.

[0123] The processing module 202 is used to pair the strain evolution data with the overall structural strain at the corresponding time node for analysis, and form a normal operating mode of the structural response through spatial mapping and time series modeling. The normal operating mode is expressed in a multi-dimensional feature space, including the mean of the principal strain, the rate of change, the direction of the maximum shear strain, the amplitude of the strain gradient mutation, and the coupling coefficient between the strain evolution data and the overall structural strain.

[0124] The processing module 202 is used to compare each strain tensor field at the current time node with the strain tensor corresponding to the normal operation mode point by point, and measure the degree of deviation of the strain tensor field of the preset structural part from the historical mode.

[0125] The processing module 202 is used to preliminarily screen predetermined structural parts with a deviation greater than or equal to a predetermined threshold value and mark them as potential abnormal areas.

[0126] The processing module 202 is used to analyze whether there is non-cooperative change between the overall structural deformation and the local strain in the potential abnormal area by combining the overall structural strain and the strain evolution data of the transmission tower.

[0127] The processing module 202 is used to, if it is determined that there are non-cooperative changes between the overall structural deformation and the local strain of the potential abnormal area, combine the changes in the main directions of the strain tensor of the potential abnormal area, and capture the sudden changes in the local deformation mode through the evolution of the tensor eigenvalues and eigenvectors to evaluate the degree of influence of the potential abnormal area on the overall structural response.

[0128] The processing module 202 is used to screen key potential deformation risk points with global structural impact according to the degree of impact.

[0129] In a possible implementation, the processing module 202 is configured to arrange random speckle patterns at different preset structural locations, and install a plurality of digital image acquisition devices around the random speckle patterns in a multi-viewing manner.

[0130] The acquisition module 201 is used to achieve synchronous image acquisition according to a timing control or trigger mechanism, and obtain the initial image transmitted in real time through the edge computing node or wireless transmission module.

[0131] The output module 203 is used to eliminate lens distortion errors by calibrating the acquisition system, and introduce time series filtering and image enhancement algorithms to process the initial image to obtain the acquired image.

[0132] It should be noted that the above embodiments provide devices that implement their functions using only the division of the above functional modules as examples. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the device and method embodiments provided in the above embodiments are based on the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0133] Corresponding to the transmission tower deformation monitoring method described in the first embodiment of the present invention, the third embodiment of the present invention further provides a transmission tower deformation monitoring device, comprising:

[0134] one or more processors;

[0135] Memory;

[0136] One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, and the one or more applications are configured to execute the transmission tower deformation monitoring method described in the aforementioned embodiment 1 of the present invention.

[0137] Corresponding to the transmission tower deformation monitoring method described in the aforementioned embodiment 1 of the present invention, embodiment 4 of the present invention further provides a computer program product, including computer instructions, which instruct a computer device to perform operations corresponding to the transmission tower deformation monitoring method described in the aforementioned embodiment 1 of the present invention.

[0138] Preferably, the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor, or the processor may be any conventional processor. The processor is the control center of the device, and various parts of the device are connected using various interfaces and lines.

[0139] The memory mainly includes a program storage area and a data storage area, wherein the program storage area can store an operating system, an application program required for at least one function, etc., and the data storage area can store related data, etc. In addition, the memory can be a high-speed random access memory, or a non-volatile memory, such as a plug-in hard disk, a smart memory card (SmartMedia Card, SMC), a secure digital (Secure Digital, SD) card, and a flash card, etc., or the memory can also be other volatile solid-state storage devices.

[0140] It should be noted that the above-mentioned device may include but is not limited to a processor and a memory, which can be understood by those skilled in the art.

[0141] like Figure 3 As shown, an embodiment of the present invention further provides an electronic device, including: at least one processor 301 , at least one communication bus 302 , a user interface 303 , a network interface 304 , and at least one memory 305 .

[0142] The communication bus 302 is used to implement the connection and communication between these components.

[0143] The user interface 303 may include a display screen (Display) and a camera (Camera). Optionally, the user interface 303 may also include a standard wired interface and a wireless interface.

[0144] The network interface 304 may optionally include a standard wired interface or a wireless interface (such as a WI-FI interface).

[0145] The processor 301 may include one or more processing cores. The processor 301 utilizes various interfaces and lines to connect various parts of the entire server. By running or executing instructions, programs, code sets, or instruction sets stored in the memory 305, and calling data stored in the memory 305, the processor 301 performs various server functions and processes data. Optionally, the processor 301 may be implemented in at least one hardware form of a digital signal processing (DSP), a field-programmable gate array (FPGA), or a programmable logic array (PLA). The processor 301 may integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. The CPU primarily processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display; and the modem is used to handle wireless communications. It is understood that the modem may not be integrated into the processor 301 and may be implemented separately on a single chip.

[0146] Among them, the memory 305 may include a random access memory (RAM) or a read-only memory (Read-Only Memory). Optionally, the memory 305 includes a non-transitory computer-readable storage medium. The memory 305 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 305 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area may store data involved in the above-mentioned various method embodiments, etc. The memory 305 may also be optionally at least one storage device located away from the aforementioned processor 301. Refer to Figure 3 , as a computer storage medium, the memory 305 may include an operating system, a network communication module, a user interface module, and a transmission tower deformation monitoring application.

[0147] exist Figure 3 In the electronic device shown, the user interface 303 is mainly used to provide an input interface for the user and obtain the data input by the user; and the processor 301 can be used to call the transmission tower deformation monitoring application stored in the memory 305. When executed by one or more processors 301, the electronic device executes one or more of the methods described in the above embodiments. It should be noted that for the aforementioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described order of actions, because according to the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0148] For the working principle and process of the above embodiment, please refer to the description of the above embodiment of the present invention, which will not be repeated here.

[0149] As can be seen from the above description, compared with the prior art, the present invention offers the following advantages: By deploying multi-view digital image acquisition equipment at key structural locations of transmission towers, combined with surface speckle patterns and stereoscopic reconstruction algorithms, the present invention achieves high-resolution reconstruction of the tower's three-dimensional surface geometry. This technology overcomes the limitations of traditional point-based strain gauge measurement and simultaneously acquires the full-field displacement field and strain tensor field of the entire tower structure, enabling continuous deformation perception of the transmission tower from local points to the entire structure. Using multi-period point cloud registration technology within a unified spatial coordinate system, the present invention spatially and temporally maps strain data from different time points, forming a strain evolution database with temporal continuity. This technology accurately captures subtle deformation gradient changes, enabling dynamic tracking of complex operating conditions such as long-term structural degradation and sudden load response. By integrating structural force flow paths, node importance, and energy dissipation distribution parameters to set weight coefficients, the present invention constructs a global strain model that accurately reflects the overall deformation trend of the structure, providing a reliable quantitative indicator for structural safety assessment. The present invention constructs a spatiotemporal strain tensor library of normal operating modes, combines point-by-point comparison analysis with tensor main direction evolution tracking, and can accurately identify abnormal areas that deviate from the threshold. The innovative introduction of strain direction synergy analysis effectively distinguishes local anomalies from global effects, and can quantitatively evaluate the impact of potential risks on the overall performance of the structure. The present invention realizes a complete closed loop from data acquisition, model construction to risk assessment: the front end obtains high-fidelity data through a multimodal sensor network, the middle stage uses a spatiotemporal coupling algorithm for intelligent analysis, and the back end establishes a diagnostic model based on physical mechanisms, breaking through the limitations of traditional monitoring technologies in terms of spatial coverage, temporal continuity, and anomaly recognition capabilities, providing a full life cycle health monitoring solution for transmission towers, significantly reducing operation and maintenance costs while improving the structural safety level.

[0150] The above disclosure is merely a preferred embodiment of the present invention and is not intended to limit the scope of the present invention. Therefore, equivalent changes made according to the claims of the present invention are still within the scope of the present invention.

Claims

1. A method for monitoring deformation of a transmission tower, characterized in that: include: Step S1, using multi-view digital image acquisition devices deployed at different preset structural parts of the transmission tower, synchronously acquiring multi-angle captured images of random speckle patterns on the surface of the transmission tower; Step S2, constructing an image sequence based on the synchronously acquired multi-angle images, reconstructing a three-dimensional displacement field of the transmission tower surface using a stereo vision reconstruction algorithm, and deriving a strain tensor field based on the three-dimensional displacement field; Step S3, performing time series processing on the strain tensor field at multiple time nodes through a continuous time series strain evolution analysis model to extract the strain change trend of each preset structural part; Step S4, performing spatial consistency mapping and weighted interpolation processing according to the strain variation trends of the plurality of preset structural parts to calculate the overall structural strain of the transmission tower; Step S5: combining the historical strain data with the current overall structural strain, and comparing the spatiotemporal evolution pattern of the strain tensor field to identify abnormal deformation areas of the transmission tower.

2. The method according to claim 1, characterized in that The step S2 specifically includes: The images collected at different angles at the same time are combined into an image sequence according to the spatial configuration relationship. Each set of image sequences corresponds to the surface state of the transmission tower at a time point. Applying a stereo vision reconstruction algorithm to perform dense stereo matching on the image pairs with parallax in the image sequence, converting the matched disparity map into a three-dimensional point cloud through triangulation, and constructing a three-dimensional geometric model of the tower structure surface; Obtaining the spatial displacement vectors of each of the preset structural parts at different times by three-dimensional point cloud registration and surface correspondence tracking at consecutive time points; For any of the preset structural parts, based on the spatial displacement vectors at different times, a deformation field is constructed by spatial interpolation using least squares fitting for a local area in the three-dimensional point cloud, and a full-order strain tensor is derived from the displacement gradient to construct a local deformation gradient tensor of the neighborhood of the preset structural part in continuous space; A strain tensor field of the transmission tower is constructed based on the local deformation gradient tensors of all the preset structural parts.

3. The method according to claim 2, characterized in that For any of the preset structural parts, based on the spatial displacement vectors at different times, a deformation field is constructed by spatial interpolation using least squares fitting for the local area in the three-dimensional point cloud, and a full-order strain tensor is derived from the displacement gradient, thereby constructing a local deformation gradient tensor of the neighborhood of the preset structural part in the continuous space, wherein the full-order strain tensor is expressed as follows: Among them, ε ij is the i-th row and j-th column component of the full-order strain tensor, u i is the displacement component of the preset structural part in the i-th direction during the deformation process, u j is the displacement component of the preset structural part in the jth direction during the deformation process, x i is the i-th coordinate component of the preset structural part in the undeformed state, x j is the jth coordinate component of the preset structural part in the undeformed state, represents the first-order partial derivative of the i-th displacement component with respect to the j-th coordinate component, represents the first-order partial derivative of the j-th displacement component with respect to the i-th coordinate component, is the geometric nonlinear term.

4. The method according to claim 1, wherein The step S3 specifically includes: defining a set of continuous discrete time nodes to form a strain data sequence containing time tags, wherein each of the discrete time nodes corresponds to an image captured by a multi-view digital image acquisition device at any moment of the preset structural part; For each of the discrete time nodes, obtaining the strain tensor field of the preset structural part; Normalizing and mapping the plurality of strain tensor fields in a unified spatial coordinate system; For multiple strain tensor fields of normalized mapping, time series data are extracted, and the strain change rate is further extracted by fitting method.

5. The method according to claim 1, wherein The step S4 specifically includes: The strain variation trends of the plurality of preset structural parts are uniformly projected into the global coordinate system of the transmission tower through spatial consistency mapping, thereby forming strain data that is spatially distributed, continuous, and temporally sequential; At each monitoring time node, the strain change trends of each preset structural part are jointly analyzed, and a mapping relationship from local strain to global strain is constructed using a method based on weighted spatial interpolation to obtain the overall strain of the structure. The weight parameters are set based on the force flow proportion of each part in the structural force system, the node importance coefficient and the density of the structural energy dissipation position.

6. The method according to claim 1, characterized in that The step S5 specifically includes: Constructing a complete spatiotemporal strain tensor database of the transmission tower over its structural history, the spatiotemporal strain tensor database covering strain evolution data of each preset structural part under multiple operating cycles, environmental changes, and load response conditions; Pairing the strain evolution data with the overall structural strain at the corresponding time node for analysis, forming a normal operating mode of the structural response through spatial mapping and time series modeling, wherein the normal operating mode is expressed in a multidimensional feature space, including the mean of the principal strain, the rate of change, the direction of the maximum shear strain, the amplitude of the strain gradient mutation, and the coupling coefficient between the strain evolution data and the overall structural strain; Comparing each of the strain tensor fields at the current time node with the strain tensor corresponding to the normal operating mode point by point, and measuring the degree of deviation of the strain tensor field of the preset structural part from the historical mode; Initially screen the pre-set structural parts with deviations greater than or equal to the pre-set threshold and mark them as potential abnormal areas; In combination with the overall structural strain of the transmission tower and the strain evolution data, analyzing whether there is non-cooperative change between the overall structural deformation and the local strain of the potential abnormal area; If it is determined that there is a non-cooperative change between the overall structural deformation and the local strain of the potential abnormal area, the principal direction change of the strain tensor of the potential abnormal area is combined with the evolution of the tensor eigenvalues and eigenvectors to capture the sudden change of the local deformation mode and evaluate the influence of the potential abnormal area on the overall structural response; The key potential deformation risk points with global structural impact are screened according to the impact degree.

7. The method according to claim 1, characterized in that The step S1 specifically includes: Arranging random speckle patterns at different predetermined structural locations, and installing a plurality of digital image acquisition devices in a multi-viewing manner around the random speckle patterns; Image synchronization is achieved based on timing control or triggering mechanisms, and the initial image is obtained through edge computing nodes or wireless transmission modules in real time. The acquisition system is calibrated to eliminate lens distortion errors, and time series filtering and image enhancement algorithms are introduced to process the initial image to obtain the multi-angle acquisition image.

8. A transmission tower deformation monitoring device, characterized in that: Includes: An acquisition module is configured to synchronously acquire multi-angle images of random speckle patterns on the surface of the transmission tower using multi-angle digital image acquisition devices arranged at different preset structural locations of the transmission tower; A processing module is configured to construct an image sequence based on the synchronously acquired multi-angle images, reconstruct a three-dimensional displacement field on the surface of the transmission tower using a stereoscopic vision reconstruction algorithm, and derive a strain tensor field based on the three-dimensional displacement field; perform time series processing on the strain tensor field at multiple time nodes using a continuous time series strain evolution analysis model to extract strain change trends at each preset structural part; perform spatial consistency mapping and weighted interpolation processing based on the strain change trends of the multiple preset structural parts to calculate the overall structural strain of the transmission tower; The output module is used to combine historical strain data with the current overall structural strain and identify abnormal deformation areas of the transmission tower by comparing the spatiotemporal evolution patterns of the strain tensor field.

9. A transmission tower deformation monitoring device, characterized in that: include: one or more processors; Memory; One or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, and the one or more applications are configured to execute the transmission tower deformation monitoring method according to any one of claims 1 to 7.

10. A computer program product, characterized in that The method comprises computer instructions, wherein the computer instructions instruct a computer device to perform operations corresponding to the method according to any one of claims 1 to 7.

Images