Power transmission line tree obstacle risk monitoring method and device, terminal equipment and storage medium

Abstract

The invention discloses a power transmission line tree obstacle risk monitoring method and device, terminal equipment and a storage medium, and relates to the field of power transmission safety monitoring, and the method comprises the steps: obtaining environment data, historical interference data and a plurality of polarization interference images of a to-be-detected power transmission corridor; screening the target interference pair set, extracting polarization entropy and first interference coherence, and constructing first polarization coherence entropy; unwrapping the winding phase of the target interference pair, and constructing a phase consistency constraint based on a real phase; setting a plurality of inversion tree height values under constraints, and calculating corresponding second interference coherence and second polarization coherence entropy; constructing a target function according to the difference of interference coherence and the difference of polarization coherence entropy; and taking the inversion tree height value corresponding to the minimum value of the target function as a target tree height value, and determining the tree barrier risk according to the corresponding second polarization coherence entropy. According to the invention, the method can solve a problem of incoherence caused by the dynamic change of vegetation and a multi-layer structure in the traditional tree height inversion, and achieves the high-precision tree height inversion.

Description

Technical Field

[0001] This invention relates to the field of power transmission safety monitoring, and in particular to a method, device, terminal equipment, and storage medium for monitoring tree obstruction risks along power transmission lines. Background Technology

[0002] With the intensification of global climate change and the increasing frequency and intensity of extreme weather events, power transmission corridors face increasingly severe environmental challenges. Among these, vegetation encroachment on transmission corridors has become a major threat to the safe operation of the power grid. In recent years, transmission line faults caused by excessive vegetation growth, fallen trees pressing on power lines, and insufficient safe distances between trees and conductors have occurred frequently. This not only leads to large-scale power outages, causing great inconvenience to residents and businesses, but also may trigger secondary disasters such as forest fires, resulting in severe economic losses and social impacts. Therefore, monitoring tree obstruction in transmission corridor areas is of paramount importance in vegetation management.

[0003] In tree barrier monitoring along power transmission corridors, tree height inversion is a crucial step in assessing vegetation growth risk. Traditional tree height inversion methods, such as those based on the conventional random vegetation layer (RVOG) model, suffer from significantly reduced radar signal coherence due to the multi-layered scattering structure and dynamic changes of vegetation—a problem known as decoherence. When coherence is insufficient, the phase information relied upon by traditional models becomes subject to substantial noise or distortion, leading to significant discrepancies between the inverted tree height results and actual values. Summary of the Invention

[0004] This invention provides a method, device, terminal equipment, and storage medium for monitoring tree obstruction risks along power transmission lines. It can solve the problem of incoherence caused by dynamic changes in vegetation and multi-layered structures in traditional tree height inversion, and achieve high-precision tree height inversion.

[0005] An embodiment of the present invention provides a method for monitoring tree obstruction risk on power transmission lines, comprising: Acquire environmental data, historical interferometric data, and several polarization interferometric images of the power transmission corridor under test; Based on environmental data, historical interferometric data, and several polarimetric interferometric images, the target interferometric pair set is determined; Extract the polarization entropy and first interference coherence of each target interference pair in the target interference pair set; Based on the polarization entropy and the first interference coherence, construct the first polarization coherence entropy; The entangled phase generated by each target interferometer pair in the target interferometer pair set is untangled to obtain the true phase; Based on the true phase, a phase consistency constraint is constructed, and under the constraint of the phase consistency constraint, several different inversion tree height values ​​are set. For each inversion tree height value, the inversion tree height value is input into a preset vegetation interference coherence model to determine the second interference coherence, and the inversion tree height value is input into a preset polarization coherence entropy model to determine the second polarization coherence entropy. Based on the difference between the first interference coherence and the second interference coherence, and the difference between the first polarization coherence entropy and the second polarization coherence entropy, an objective function is constructed. The height of the inversion tree corresponding to the minimum value of the objective function is taken as the height of the objective tree. The tree obstacle risk level is determined based on the second polarization coherence entropy corresponding to the target tree height, and an early warning signal is issued when the tree obstacle risk level exceeds the preset risk level.

[0006] Furthermore, after acquiring environmental data, historical interferometric data, and several polarization interferometric images of the transmission corridor under test, the process also includes: Select one polarimetric interferometric image from all polarimetric interferometric images as the master image; Based on the main image, all polarimetric interferometric images are registered to obtain the registered polarimetric interferometric images.

[0007] Furthermore, based on environmental data, historical interferometric data, and several polarimetric interferometric images, the target interferometric pair set is determined, including: Construct the current state vector based on environmental data and historical interference data; For each interference pair, the corresponding spatial gradient is calculated based on the interference phase formed by the interference pair, and the spatial gradient is discretized to obtain the corresponding gradient energy; the interference pair is composed of any two different polarization interference images. Interference pairs corresponding to gradient energies less than a preset stability threshold are selected as candidate interference pairs. Construct the action space based on the temporal baselines of all candidate interference pairs; In the action space, based on the current state vector, the time baseline of the next interferometric pair is selected step by step according to the current policy Q, until the total time span of the interferometric pair corresponding to the selected time baseline covers the time period of all polarization interferometric images, thus obtaining the target interferometric pair set.

[0008] Furthermore, the polarization entropy and first interference coherence of each target interference pair in the target interference pair set are extracted, including: For each target interference pair in the target interference pair set, a coherence matrix is ​​constructed based on the complex scattering components of horizontal transmission-horizontal reception, horizontal transmission-vertical reception, and vertical transmission-vertical reception in the target interference pair. Hermitian eigenvalues ​​are obtained by performing Hermitian eigenvalue decomposition on the coherence matrix; Each scattering eigenvalue is normalized to obtain the normalized scattering eigenvalue; The polarization entropy is determined based on all normalized scattering eigenvalues. The first interference coherence is determined based on the scattering vector of each target interference pair in the target interference pair set.

[0009] Furthermore, the entangled phase generated by each target interferometer pair in the target interferometer pair set is unwrapped to obtain the true phase, including: Generate the corresponding entanglement phase for each target interference pair in the target interference pair set; A coupling term is constructed based on the intertwined phase of adjacent pixels; Obtain the terrain prior knowledge DEM elevation information, and construct the bias term based on the winding phase and the terrain prior knowledge DEM elevation information; Construct the binary compensation variables to be solved; Construct the Hamiltonian energy function of the Ising model based on the coupling term, bias term, and binary compensation variable; Quantum annealing is performed on the Hamiltonian energy function of the Ising model to obtain the target compensation variable; The untangling phase is determined based on the entanglement phase and the target compensation variable; Anisotropic diffusion filtering is applied to the unwrapped phase to obtain the true phase.

[0010] Furthermore, phase consistency constraints include: ; in, Indicates the true phase. Indicates the height of the inversion tree. Indicates the height of the inversion tree The corresponding second interference coherence.

[0011] Furthermore, the objective function includes: ; in, Describe the objective function. Indicates the first interference coherence. Indicates the second interference coherence. Represents the first polarization coherence entropy. Represents the coherent entropy of the second polarization. This indicates the ratio of scattering contribution from the land surface to that from vegetation.

[0012] Based on the above method embodiments, the present invention provides corresponding device embodiments, including: a power transmission corridor data acquisition module, an interference pair screening module, a polarization interference parameter extraction module, a polarization coherence entropy construction module, a phase unwrapping module, a tree height inversion module, an objective function construction module, a target tree height determination module, and a tree obstacle risk warning module; The power transmission corridor data acquisition module is used to acquire environmental data, historical interferometric data, and several polarization interferometric images of the power transmission corridor under test. The interferometric pair screening module is used to determine the target set of interferometric pairs based on environmental data, historical interferometric data, and several polarimetric interferometric images. The polarization interferometry parameter extraction module is used to extract the polarization entropy and first interferometric coherence of each target interferometry pair in the target interferometry pair set; The polarization coherence entropy construction module is used to construct the first polarization coherence entropy based on the polarization entropy and the first interference coherence. The phase unwrapping module is used to unwrap the wrapped phase generated by each target interferometer pair in the target interferometer pair set to obtain the true phase. The tree height inversion module is used to construct phase consistency constraints based on the true phase, and set several different inversion tree height values ​​under the constraints of phase consistency constraints. For each inversion tree height value, the inversion tree height value is input into a preset vegetation interference coherence model to determine the second interference coherence, and the inversion tree height value is input into a preset polarization coherence entropy model to determine the second polarization coherence entropy. The objective function construction module is used to construct an objective function based on the difference between the first interference coherence and the second interference coherence, as well as the difference between the first polarization coherence entropy and the second polarization coherence entropy. The target tree height determination module is used to take the inversion tree height value corresponding to the minimum value of the objective function as the target tree height value. The tree obstacle risk early warning module is used to determine the tree obstacle risk level based on the second polarization coherence entropy corresponding to the target tree height, and to issue an early warning signal when the tree obstacle risk level exceeds the preset risk level.

[0013] Based on the above method embodiments, the present invention provides a corresponding terminal device embodiment, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps of the transmission line tree obstacle risk monitoring method as described in the present invention.

[0014] Based on the above method embodiments, the present invention provides a corresponding computer-readable storage medium embodiment, including: a stored computer program, which, when the computer program is running, controls the device where the computer-readable storage medium is located to execute the steps of the transmission line tree obstacle risk monitoring method as described in the present invention.

[0015] Compared with the prior art, the beneficial effects of this embodiment are as follows: This invention uses environmental data, historical interferometric data, and several polarimetric interferometric images of the transmission corridor under test to screen a set of target interferometric pairs. Then, it extracts the polarization entropy and first interferometric coherence of each target interferometric pair in the set. The polarization entropy quantifies the multilayer scattering structure of vegetation, and the interferometric coherence captures dynamically changing phase stability. The first polarization coherence entropy is constructed to reduce information loss caused by decoherence. The entangled phases generated by each target interferometric pair in the set are untangled to obtain the true phase. Based on the true phase, a phase consistency constraint is constructed to physically ensure the consistency between tree height and phase characteristics, avoiding interference from phase distortion caused by decoherence on tree height inversion. Under the constraint of phase consistency, several different inverted tree height values ​​are set. For each inverted tree height value, the value is input into a preset vegetation interferometric coherence model to determine the second interferometric coherence, and input into a preset polarization coherence entropy model to determine the second polarization coherence entropy. Based on the differences between the first and second interferometric coherence, and the differences between the first and second polarization coherence entropies, an objective function is constructed. This ensures both the phase physical rationality of the tree height inversion and adapts to the structural dynamic characteristics of multi-layer scattering in vegetation. The inverted tree height value corresponding to the minimum value of the objective function is taken as the target tree height value. Finally, based on the second polarization coherence entropy corresponding to the target tree height value, the tree barrier risk level is determined, and an early warning signal is issued if the tree barrier risk level exceeds the preset risk level.

[0016] In summary, this invention constructs a multi-dimensional polarization coherence entropy and introduces a phase consistency constraint on the true phase, enabling tree height inversion to reduce the loss of decoherent information from both structural and dynamic dimensions, while ensuring the consistency between the phase and tree height from a physical perspective, under the conditions of multi-layered scattering structures and dynamic changes in vegetation. This solves the decoherence problem caused by dynamic changes and multi-layered structures in traditional tree height inversion and achieves high-precision tree height inversion. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a method for monitoring tree obstruction risks in power transmission lines according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating the training process of an adaptive baseline selection network model according to an embodiment of the present invention. Figure 3 This is a schematic diagram of the structure of a power transmission line tree obstacle risk monitoring device provided in an embodiment of the present invention. Detailed Implementation

[0018] 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 skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] In the description of this invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated.

[0020] like Figure 1 As shown, in order to solve the problem of incoherence caused by dynamic changes in vegetation and multi-layered structures in traditional tree height inversion, an embodiment of the present invention provides a method for monitoring tree obstruction risks in transmission lines, which includes at least the following steps: Step S1: Acquire environmental data, historical interferometric data, and several polarization interferometric images of the transmission corridor under test; For step S1, several polarization interferometric images are acquired for the transmission corridor under test during the observation period. , For the first A polarization interferometric image, The total number of polarization interferograms. The time for each polarization interferogram. (Months) are used for seasonal coding, with spring being... Summer is Autumn is Winter is ; The above environmental data includes: monthly cumulative rainfall during the observation period. Average monthly temperature Standardization , ; The aforementioned historical interferometric data includes: the set of baselines selected in the past K times. Historical regional average coherence , ;in, Indicates the first The time interval corresponding to the second selected baseline is the time difference between the two interferometric images associated with that baseline.

[0021] In a preferred embodiment, after acquiring environmental data, historical interferometric data, and several polarization interferometric images of the transmission corridor under test, the method further includes: Select one polarimetric interferometric image from all polarimetric interferometric images as the master image; Based on the main image, all polarimetric interferometric images are registered to obtain the registered polarimetric interferometric images.

[0022] In one embodiment of the present invention, a polarization interferometric image is selected as the master image from all polarization interferometric images. The remaining polarization interferometry images are from the image Subpixel-level fine registration is performed between polarimetric interferometric images, and a phase gradient adaptive algorithm is used to eliminate orbital errors, so that the displacement field is obtained after registration. , ,satisfy: ; in, Indicates the main image in pixel coordinates The value at that location, and This represents the displacement field, also called offset, and is used to characterize the pixel offset of the image relative to the main image in the horizontal and vertical directions. Represents the pixel coordinates of the image after offset. The value at that location.

[0023] The final registered polarimetric interferometric image is obtained with a registration error of <0.1 pixels, ensuring that the master and slave images are spatially consistent, thereby accurately extracting phase information.

[0024] Step S2: Determine the target interferometric pair set based on environmental data, historical interferometric data, and several polarimetric interferometric images; In a preferred embodiment, a set of target interferometric pairs is determined based on environmental data, historical interferometric data, and several polarization interferometric images, including: Construct the current state vector based on environmental data and historical interference data; For each interference pair, the corresponding spatial gradient is calculated based on the interference phase formed by the interference pair, and the spatial gradient is discretized to obtain the corresponding gradient energy; the interference pair is composed of any two different polarization interference images. Interference pairs corresponding to gradient energies less than a preset stability threshold are selected as candidate interference pairs. Construct the action space based on the temporal baselines of all candidate interference pairs; In the action space, based on the current state vector, the time baseline of the next interferometric pair is selected step by step according to the current policy Q, until the total time span of the interferometric pair corresponding to the selected time baseline covers the time period of all polarization interferometric images, thus obtaining the target interferometric pair set.

[0025] For step S2, based on the environmental data and historical interference data obtained in step S1, the current state vector is constructed. , An interference pair is formed by any two different polarization interferograms, and each interference pair corresponds to an interference phase diagram. For each interferometric phase pattern, calculate its spatial gradient. , which is the rate of change of phase in the horizontal and vertical directions, is given by the following formula: ; in, The spatial gradient of the interferometric phase map is also the phase gradient. and Indicates phase Spatial coordinates and The partial derivative of the phase. The gradient reflects the local change of the phase in space; a large gradient value indicates a drastic phase change, while a small gradient value indicates a smooth phase.

[0026] Since the phase map is a discrete pixel grid, it is necessary to approximate the continuous derivative through difference. This invention uses the Sobel operator to discretize and approximate the gradient, estimating the gradient using the phase difference between adjacent pixels, thereby transforming the continuous gradient into a computable operation on the discrete pixel grid. The specific formula is as follows: ; Define gradient energy as Used to measure global phase smoothness: ; Gradient energy By combining phase change information in both the horizontal and vertical directions, the smaller the gradient energy, the smoother the overall phase map, and the higher the quality of the interferometer pair.

[0027] Preset stability threshold Determined by empirical values ​​of vegetation scattering stability, only gradient energy is retained. Interference pairs are selected as candidate interference pairs, thereby excluding low-quality interference pairs with drastic phase changes, reducing computational complexity while ensuring the accuracy of the results.

[0028] Next, the temporal baselines of all candidate interferometer pairs are mapped to a discrete set of actions, such as polarimetric interferometry images. , , , , , ..., each action corresponds to a time baseline selection strategy, forming an action space; It should be noted that the total coverage area after combining the time spans of all selected interference pairs must match the total time period of the polarization interferometric image and must not exceed the effective coverage range of the data.

[0029] The current state vector and action space are input into the adaptive baseline selection network model, so that the adaptive baseline selection network model can, in the action space, adjust the current state vector... The time baseline for selecting the next intervention pair is gradually chosen based on the current strategy Q. In the selection process, an ε-greedy strategy is adopted. In this embodiment, when ε=0.1, an action is randomly selected with a 10% probability, and the action corresponding to the highest value of the current strategy Q is selected with a 90% probability. Update the next state vector Simultaneously, selected actions are removed from the action space until the total time span of the interferometric pairs corresponding to the selected time baseline can completely cover the time periods of all polarimetric interferometric images. For example, after three selection steps, the first step outputs... ,cover to The second step output The coverage area has been expanded to to Third step output Further coverage to At this point, complete coverage to The selection process ends at all corresponding time periods, ultimately yielding the target interference pair set.

[0030] It should be noted that the current policy Q in the adaptive baseline selection network model is obtained through training. It can evaluate the expected benefits corresponding to different time baselines under the current state vector, thereby assisting in the selection of intervention pairs that can effectively advance the coverage progress of the time span.

[0031] Specifically, the current policy Q in the adaptive baseline selection network model is determined in the following way: like Figure 2 As shown, firstly, temporal polarimetric SAR data covering multiple growth cycles is acquired to generate candidate interferometric pair samples, as well as LiDAR measured tree height variation values. This serves as a reference benchmark for the accuracy of tree height inversion.

[0032] The temporal polarimetric SAR data and the annotation information of LiDAR measured tree heights are divided into training datasets and test datasets.

[0033] All state-action pairs The value is initialized to 0; each round starts from a random initial state. Begin by repeatedly executing the following steps until the maximum number of training rounds is reached, at which point the iteration ends, taking 1000 values: Based on the current state An ε-greedy strategy is used to select actions. Execute actions , obtain a new state By combining the tree height retrieved from the current interferometric pair with the LiDAR-measured tree height, the reward for this action is calculated using a reward function. : ; in, Represents the reward function, This represents the tree height value retrieved from SAR data, which is a model prediction. This represents the measured tree height variation value using LiDAR. Indicates the penalty coefficient. Indicates the time baseline of the next interferometric pair, time baseline The longer the time frame, the higher the penalty. The reward function of this invention allows the policy to find an optimal balance between high-precision inversion and a reasonable time baseline, ensuring the effectiveness of the trained policy. It can effectively advance the time coverage schedule while ensuring the quality of tree height inversion.

[0034] Optimize according to Q-learning update rules value: ; in, Indicates the current state Next, select an action. The expected total reward that can be obtained later. This represents the learning rate, set to 0.3. This represents the discount factor, which is 0.75. Indicates the execution of an action Then enter a new state Optional actions.

[0035] In a new state If the current state for the next time step has not yet reached 1000 iterations, then hyperparameter tuning is performed, and the learning rate is optimized using gradient descent. Discount Factor and penalty coefficient The specific adjustment formula is as follows: ; Repeat the above steps until 1000 training rounds are completed, at which point the trained policy is finally obtained. .

[0036] Existing temporal polarimetric interferometry methods, such as SBAS-InSAR, require a fixed spatial baseline network, which makes it difficult to adapt to the dynamic deformation characteristics of vegetation scenes. Moreover, the computational complexity increases by O(N²) with the number of phases. This invention solves the signal aliasing problem caused by the fixed baseline in vegetation areas in traditional SBAS methods by combining reinforcement learning and phase gradient constraints to adaptively select the optimal combination of interferometric pairs. This optimizes the temporal baseline and spatial coherence, and improves the inversion accuracy of tree height in vegetation areas.

[0037] Step S3: Extract the polarization entropy and first interference coherence of each target interference pair in the target interference pair set; In a preferred embodiment, extracting the polarization entropy and first interference coherence of each target interference pair in the target interference pair set includes: For each target interference pair in the target interference pair set, a coherence matrix is ​​constructed based on the complex scattering components of horizontal transmission-horizontal reception, horizontal transmission-vertical reception, and vertical transmission-vertical reception in the target interference pair. Hermitian eigenvalues ​​are obtained by performing Hermitian eigenvalue decomposition on the coherence matrix; Each scattering eigenvalue is normalized to obtain the normalized scattering eigenvalue; The polarization entropy is determined based on all normalized scattering eigenvalues. The first interference coherence is determined based on the scattering vector of each target interference pair in the target interference pair set.

[0038] For step S3, since the electromagnetic waves from the coherent target undergo a linear change upon scattering, this change can be represented by the Sinclair scattering matrix. Represented by a matrix. Based on the reciprocity theorem of a single-base radar observation system using the same antenna for both transmission and reception, and the antenna reciprocity principle of a single-base system, The cross-polarization terms in the matrix are equivalent, that is... Therefore, for each target interferometric pair in the target interferometric pair set, the complex scattering components of horizontal emission-horizontal reception, horizontal emission-vertical reception, and vertical emission-vertical reception of the main image and auxiliary image are extracted respectively to construct the polarization scattering matrix. : ; in, , and These represent the complex scattering components of horizontal transmission-horizontal reception, horizontal transmission-vertical reception, and vertical transmission-vertical reception, respectively, and include amplitude and phase information.

[0039] Subsequently, by analyzing the polarization scattering matrix Its conjugate transpose Perform the outer product calculation to obtain the coherence matrix. The specific formula is as follows: ; in, , and They represent complex conjugate, complex conjugate and .

[0040] For ease of description, the coherence matrix Simplified representation: ; To suppress the influence of speckle noise, this invention performs spatial averaging on adjacent pixel values ​​to obtain the coherence matrix. : ; in, Indicates spatial average. This represents the number of pixels in the spatial neighborhood multi-view window that participated in the averaging calculation. This indicates the sequence number of the neighboring pixels that participate in the calculation of the spatial average. This represents the combination of polarization scattering channels at the neighborhood pixels that participate in the calculation of spatial averaging.

[0041] For the above coherence matrix Polarization entropy is extracted using an eigenvalue decomposition-based approach. and average scattering angle Specifically: First, for the coherence matrix Performing Hermiian eigenvalue decomposition yields three non-negative eigenvalues, as shown in the following formula: ; in, Eigenvalues ​​are used to characterize the intensity of different scattering mechanisms. ; This represents the normalized eigenvector, used to describe the polarization state of the scattering mechanism.

[0042] Secondly, the eigenvalues ​​are normalized into a probability distribution: ; in, Indicates the first The proportion of each scattering mechanism in the total scattering, if For a deterministic scattering target such as a flat surface, if , for forest body scattering and other scattering height random targets.

[0043] Then, using scattering entropy The randomness of the scattering process is quantified by the following formula: ; in, Representing scattering entropy, it is used to quantify the randomness of the scattering process, and also to characterize the randomness of the scattering process under the polarization dimension; therefore, it is also polarization entropy. For a single scattering mechanism, the eigenvalues ​​exhibit the following characteristics: , For completely random scattering, the eigenvalues ​​are as follows: The higher the scattering entropy value, the more complex the characterization of the scattering target; In this embodiment, polarization entropy When vegetation scattering dominates, it exhibits a high entropy value. Low entropy values ​​indicate that regular targets such as exposed ground surfaces or conductors exhibit this characteristic. .

[0044] At the same time, using the average scattering angle The type of dominant scattering mechanism of the target scatterer is represented by the following formula: ; in, Indicates the first The scattering angle corresponding to each scattering mechanism.

[0045] Finally, based on the interference coherence between the two images To quantify the phase stability between two SAR images, let the scattering matrix of the main image be... supplementary images are For any polarization combination such as , , First interference coherence Defined as: ; in, The representation space is mostly based on average. Indicates complex amplitude. Indicates the main image in polarization combination The complex scattering component below, Indicates auxiliary image in polarization combination The complex scattering component below.

[0046] In this embodiment, the first interference coherence The vegetation zone along the power transmission corridor exhibits low coherence due to volume scattering. Stable surfaces such as rocks have higher coherence. .

[0047] Step S4: Construct the first polarization coherence entropy based on the polarization entropy and the first interference coherence; For step S4, the polarization entropy obtained in step S3 is used as a basis. and first interference coherence Constructing the first polarization coherent entropy : ; Among them, the first item This is a physical description of structural randomness, representing the coupling relationship between the randomness of polarization entropy and phase consistency between images. When vegetation is lush, the polarization entropy... Approaching 1, first interferometric coherence The polarization entropy approaches 0, with the first term approaching 1, indicating high randomness; when trees fall or are removed, the polarization entropy... Reduce, first interference coherence The increase in the first item and the decrease in the second item indicate that the risk of vegetation collapse in the vegetation-covered areas of the power transmission corridor has been mitigated. Second item Represents the dynamic characteristics of time series changes, where the time derivative of coherence is... Used to capture mutation events, such as vegetation lodging in the study area leading to... Sudden rise, This represents the adaptive weighting coefficient, used to control the dynamically changing contribution, by maximizing... Changes in tree height as measured on the ground The correlation coefficient is determined using the following formula: ; In order to accurately quantify Changes in tree height as measured on the ground The linear correlation between them is optimized using the Pearson correlation coefficient to find the optimal adaptive weighting coefficient. The specific formula can be transformed into the following form: ; in, The Pearson correlation coefficient is a commonly used standardized indicator used to measure the degree of linear correlation between two variables; the larger the absolute value, the stronger the correlation.

[0048] Optimal adaptive weight coefficient The specific steps for solving this problem are as follows: First, data preparation work is carried out, including acquiring time-series observation data, including SAR time-series observation parameters. , and the corresponding measured tree height ; Then, use the measured tree height at the next moment. Subtract the measured tree height at the current moment Calculate the rate of change of tree height ,Right now: ; Construct a gradient descent algorithm for grid search: In Iterate through the candidate values ​​within the interval, for each calculate and with Calculate the Pearson correlation coefficient. Ultimately, the choice made Maximum adaptive weight coefficient As the optimal adaptive weight coefficient Used for the first polarization coherent entropy The adaptive weight coefficient of the second term .

[0049] Furthermore, due to the original SAR temporal interferometry coherence The data will fluctuate randomly due to observation noise, environmental interference, etc., if the raw data is used directly... Calculate the time derivative of coherence Noise can be amplified, making it easy to misinterpret noise fluctuations as sudden events such as vegetation lodging, leading to... The accuracy has decreased.

[0050] Therefore, this invention employs a filtering method based on local polynomial least squares fitting, using a Savitzky-Golay (SG) filter to smooth and denoise the temporal coherence data of the vegetation-covered area of ​​the power transmission corridor. The specific process of derivative calculation is as follows: right Performing a polynomial fit, we have: ; in, This indicates the smoothed interference coherence. Indicates the half-width of the window and the width of the window. , The coefficients represent the polynomial fitting coefficients, which are determined by the least squares polynomial fitting.

[0051] The time derivative is approximately: ; Among them, in the above formula The time sampling interval is represented by the derivative calculated from the difference between the smoothed values ​​at different times to suppress high-frequency noise.

[0052] This invention constructs polarization coherent entropy As a characterization of polarization entropy and interference coherence The dynamically changing joint index integrates the randomness of polarization entropy, the stability of coherence, and information on temporal mutations.

[0053] Step S5: Unwrap the entangled phase generated by each target interferometer pair in the target interferometer pair set to obtain the true phase; In a preferred embodiment, the entangled phase generated by each target interferometer pair in the target interferometer pair set is unwrapped to obtain the true phase, including: Generate the corresponding entanglement phase for each target interference pair in the target interference pair set; A coupling term is constructed based on the intertwined phase of adjacent pixels; Obtain the terrain prior knowledge DEM elevation information, and construct the bias term based on the winding phase and the terrain prior knowledge DEM elevation information; Construct the binary compensation variables to be solved; Construct the Hamiltonian energy function of the Ising model based on the coupling term, bias term, and binary compensation variable; Quantum annealing is performed on the Hamiltonian energy function of the Ising model to obtain the target compensation variable; The untangling phase is determined based on the entanglement phase and the target compensation variable; Anisotropic diffusion filtering is applied to the unwrapped phase to obtain the true phase.

[0054] For step S5, based on each target interferometer pair in the target interferometer pair set obtained in step S2, a wrapped phase is generated by InSAR interferometry. Its true phase for: ; in, Let be the integer number of transitions to be solved.

[0055] Traditional phase unwrapping is achieved by minimizing the phase transition difference between adjacent pixels, i.e.: ; in, Indicates adjacent pixel pairs, It represents the set of all adjacent pixel pairs. Indicates the true phase in adjacent pixels gradient between, Indicates the entangled phase in adjacent pixels The gradient between them.

[0056] However, traditional methods, based on minimizing gradient differences and optimizing pixel by pixel, are prone to accumulating errors in noisy or complex terrains such as vegetated areas, leading to global inconsistencies. Therefore, this invention uses the Ising model for variable mapping, specifically: First, define the binary compensation variable. , representing pixels Is it necessary? Compensation, when Time compensation, when No compensation will be provided. Secondly, establish coupling terms. This will force adjacent pixels to... The process of phase transition consistency is represented by the following formula: ; in, Represents pixels The entanglement phase, Represents pixels adjacent pixels The winding phase. If the winding phase difference between adjacent pixels is small, then Approaching 1, strong coupling requirement If the phase difference is large, When the value is close to -1 or small, the coupling weakens.

[0057] Introducing prior topographic knowledge and DEM elevation information To constrain the phase jump and the bias term related to terrain The solution is given by the following formula: ; in, Represents pixels DEM elevation values, Indicates the reference elevation.

[0058] Finally, phase unwrapping is transformed into minimizing the Hamiltonian energy function of the Ising model. The specific formula is as follows: ; To solve for the Hamiltonian energy function in the Ising model, a quantum annealing model is used, by adjusting the transverse magnetic field... Controlled quantum fluctuation processes are caused by coupling terms With bias term The controlled classical potential field allows the system to gradually evolve from a quantum superposition state to the classical ground state. The process of mapping the Ising model to the Pegasus topology of the D-Wave quantum processing unit requires chain coupling to address hardware connectivity constraints. The specific annealing scheduling process mainly consists of the following three steps: Initial time Time: in the horizontal field The system is in a uniform superposition state; Annealing process Gradually decrease Enhance and The impact; Final moments The system collapses to the classical minimum energy state.

[0059] At this point, the Hamiltonian energy function of the Ising model The expression is transformed into: ; in, and Represents the chain coupling coefficient. Represents pixels The binary compensation variable is mapped to a quantum bit. Directional spin state, Represents pixels The binary compensation variable is mapped to a quantum bit. Directional spin state, Represents pixels The binary compensation variable is mapped to a quantum bit. Directional spin state. It should be noted that pixels... and pixels It is the index of adjacent pixel pairs, not an independent variable.

[0060] By adjusting the chain coupling coefficient , After completing the quantum tunneling to the ground state, the above process is sampled multiple times and annealed multiple times, such as 1000 times, and the optimal solution is statistically analyzed. The target compensation variable is obtained. Represents pixels Whether compensation for the 2π phase transition is needed. A value of 1 indicates compensation, and a value of 0 indicates no compensation.

[0061] Assuming each jump is to 2π, the actual value needs to be determined according to... The accumulated value is adjusted based on the target compensation variable obtained above. The specific formula for calculating the unwrapping phase is as follows: ; in, Represents pixels The untangling phase, Represents pixels The winding phase, in the above formula Represents pixels The target compensation variable, when When, in the formula The phase compensation amount is π; when When, in the formula No phase compensation.

[0062] After compensation, the untangling phase is achieved. To restore continuity and resolve phase ambiguity issues.

[0063] To further smooth out random fluctuations in the phase, this invention uses anisotropic diffusion filtering to eliminate residual noise and obtain the true phase. The specific formula is as follows: ; in, Indicates the diffusion coefficient. This represents the spatial gradient of the unwrapped phase.

[0064] Because phase unwrapping in densely vegetated areas is affected by multipath scattering, traditional least squares methods or network flow algorithms have large errors in low-coherence regions, resulting in poor accuracy of phase unwrapping results. This invention introduces quantum annealing unwrapping, encoding the phase gradient variable as the Ising model Hamiltonian, and using a quantum annealing machine to globally search for the optimal phase unwrapping path. This significantly reduces the number of residual points in high-noise regions with high coherence, thereby improving the efficiency and accuracy of phase unwrapping.

[0065] Step S6: Based on the true phase, construct phase consistency constraints, and under the constraints of phase consistency constraints, set several different inversion tree height values. For each inversion tree height value, input the inversion tree height value into the preset vegetation interference coherence model to determine the second interference coherence, and input the inversion tree height value into the preset polarization coherence entropy model to determine the second polarization coherence entropy. For step S6, in interferometry, the tree height value affects interferometric coherence. This invention utilizes the true phase obtained in step S5 to construct a phase consistency constraint. Preferably, the phase consistency constraint includes: ; in, Indicates the true phase. Indicates the height of the inversion tree. Indicates the height of the inversion tree The corresponding second interference coherence.

[0066] This constraint forces the true phase. It needs to be consistent with the coherence phase of the interferometric system at a specific tree height, so as to eliminate local deviations in the unwrapped phase and improve the physical rationality of the phase solution.

[0067] Phase consistency constraints Under the constraints, a series of inversion tree height values ​​are set. , and the height value of each inversion tree Input the data into the vegetation interference coherence model and calculate the corresponding second interference coherence. .

[0068] In this invention, the vegetation interference coherence model is the traditional Random Volume Over Ground (RVOG) model, assuming the vegetation layer height is... The contributions of surface scattering and vegetation canopy scattering are respectively... and Vegetation interference coherence is the sum of surface coherence and canopy coherence, and the specific formula is as follows: ; in, The weighting coefficients represent the surface scattering components. Weighting coefficients representing the scattering components of the vegetation canopy, and the bulk interference coherence of the vegetation canopy. Determined by the vertical structure of vegetation. Indicates the volume attenuation coefficient. , Indicates vertical complex beam. , Indicates the angle of incidence. This represents the viewing angle difference corresponding to the interferometric baseline. Indicates the radar wavelength.

[0069] Surface scattering Primarily dominated by surface scattering, its interference coherence is calculated as follows: ; in, Represents the surface phase, determined by the height of the surface relative to a reference surface. and vertical beam The magnitude is usually close to 1, and the surface phase calculation formula is: ; vegetation canopy scattering The interference coherence is primarily dominated by the coherence of randomly distributed scatterers, and is represented by an exponentially decaying volume scattering model, determined by the attenuation coefficient. and vegetation height Control, coherence with vertical complex beam The oscillation decay is expressed as follows: ; At the same time, the height value of each inversion tree Input the input to the polarization coherence entropy model and calculate the corresponding second polarization coherence entropy. The specific expression is: ; Among them, the structural randomness term Defined as: ; ; In the formula, In order to be at a high altitude The eigenvalues ​​of the polarization interference coherence matrix constructed by the RVOG model are shown below. This represents the second interference coherence amplitude at the corresponding height. As a weighting factor, and related to the first polarization coherence entropy Same weighting coefficient; This refers to the temporal variation of coherence related to vegetation height.

[0070] In this invention, the structural randomness, coherence correlation characteristics and dynamic time-varying features of polarization interference coherence at different heights of vegetation are comprehensively quantified by the polarization coherence entropy model, thereby reflecting the distribution of scatterers at different heights of vegetation, the complexity of scattering mechanisms and the temporal changes of coherence.

[0071] Step S7: Construct the objective function based on the difference between the first interference coherence and the second interference coherence, and the difference between the first polarization coherence entropy and the second polarization coherence entropy; For step S7, the present invention, based on the traditional RVOG model, and according to the first interference coherence... Second interference coherence The difference between them, and the first polarization coherence entropy Second polarization coherence entropy The differences between them are used to construct an objective function, thereby considering both the interference phase matching degree and the consistency of polarization scattering characteristics during the optimization process.

[0072] Preferably, the objective function includes: ; in, Describe the objective function. Indicates the first interference coherence. Indicates the second interference coherence. Represents the first polarization coherence entropy. Represents the coherent entropy of the second polarization. It represents the ratio of scattering contribution from the land surface and vegetation, reflecting the proportion of radar signal scattering from the land surface and vegetation layer.

[0073] While traditional RVOG models can invert vegetation height, their assumption of single-layer volume scattering leads to a scattering phase center localization error exceeding λ / 4 in densely vegetated areas, affecting the accuracy of temporal coherence modeling. This invention proposes multi-dimensional polarimetric coherence entropy (PCE) modeling, which quantifies the dynamic coupling effect of the vegetation canopy-land surface scattering mechanism by jointly decomposing the spatiotemporal polarimetric joint distribution of the interferometric coherence matrix, thus addressing the problem of insufficient characterization of multi-layer vegetation scattering by traditional methods.

[0074] Step S8: Take the height of the inversion tree corresponding to the minimum value of the objective function as the height of the objective tree; For step S8, by minimizing the objective function, the tree height value that makes the model prediction coherence and polarization characteristics closest to the actual observation can be selected. This effectively integrates interferometry and polarization information, overcomes the limitations of a single data source, and finally obtains the tree height inversion target with enhanced physical constraints and multi-parameter collaborative optimization, which is used as the target tree height value.

[0075] Step S9: Determine the tree obstacle risk level based on the second polarization coherence entropy corresponding to the target tree height, and issue an early warning signal if the tree obstacle risk level exceeds the preset risk level.

[0076] For step S9, based on the target tree height obtained from the inversion in step S8, the corresponding second polarization coherence entropy is calculated using the polarization coherence entropy model. Ideally, ,in, These are the weighting coefficients. The interference coherence time change rate is used, but in actual calculations, it is constrained by normalization to... Interval.

[0077] It should be noted that, since the first polarization coherence entropy directly relies on the raw radar observation data, it is susceptible to interference from temporal decoherence (such as vegetation growth), spatial baseline changes, or thermal noise, resulting in significant short-term fluctuations and making it difficult to stably reflect long-term risk trends. In contrast, the inversion tree height value, through model constraints, has already undergone noise filtering and parameter optimization of the raw observation data during the calculation process. Therefore, the second polarization coherence entropy calculated based on the inversion tree height value possesses stronger noise resistance and data stability, making it more suitable as a risk criterion.

[0078] The tree barrier risk level corresponding to the pre-set interval values ​​of the second polarization coherence entropy is shown in Table 1. Simultaneously, a warning signal is pre-set to be issued when the risk level exceeds the medium risk level.

[0079] Table 1. Risk Level Threshold Comparison Table for Vegetated Areas Along Power Transmission Lines Calculate the corresponding second polarization coherence entropy value based on the target tree height, and then base the calculation on the value's position. The risk category of the interval range (refer to Table 1) is determined by the interval matching rules, and finally the risk level of the current tree barrier is determined by combining the correspondence between intervals and risk levels in Table 1.

[0080] If the risk level of tree obstruction exceeds the medium risk level, an early warning signal will be automatically triggered, prompting maintenance personnel to take timely measures, such as pruning trees and strengthening inspections, to prevent vegetation from falling and causing power outages or fires. If the risk level of tree obstruction does not exceed the medium risk level, it is determined to be low risk or medium risk, and there is no need to trigger an early warning immediately. For low-risk areas, vegetation status data should be recorded regularly and the basic monitoring frequency should be maintained. For medium-risk areas, the inspection frequency should be appropriately increased or short-term trend tracking should be initiated to continuously observe the dynamic changes in vegetation growth or structure.

[0081] In a preferred embodiment, a dynamic risk index is defined. This index is used to comprehensively reflect recent and long-term changes in tree barrier risk. When continuous polarization coherence entropy data is obtained, an integral formula is used to calculate the dynamic risk index, as follows: ; Among them, in the above formula This refers to the second polarization coherence entropy value calculated from the target tree height. Indicates the attenuation factor. It is used to control the weighting of historical data, giving more weight to recent data. This indicates the current time, and the integration interval covers the entire detection period.

[0082] For discrete time series The risk index is approximately: ; in, Indicates the total number of time points. Indicates a point in time The corresponding second polarization coherence entropy value, As an exponentially decaying term, it can effectively capture recent abrupt events, such as vegetation recovery after a fire, which is consistent with the time-sensitive characteristics of disaster risk.

[0083] like Figure 3 As shown, based on the above method embodiments, corresponding apparatus embodiments are provided; One embodiment of the present invention provides a tree obstacle risk monitoring device for transmission lines, comprising: a transmission corridor data acquisition module, an interference pair screening module, a polarization interference parameter extraction module, a polarization coherence entropy construction module, a phase unwrapping module, a tree height inversion module, an objective function construction module, a target tree height determination module, and a tree obstacle risk early warning module; The power transmission corridor data acquisition module is used to acquire environmental data, historical interferometric data, and several polarization interferometric images of the power transmission corridor under test. The interferometric pair screening module is used to determine the target set of interferometric pairs based on environmental data, historical interferometric data, and several polarimetric interferometric images. The polarization interferometry parameter extraction module is used to extract the polarization entropy and first interferometric coherence of each target interferometry pair in the target interferometry pair set; The polarization coherence entropy construction module is used to construct the first polarization coherence entropy based on the polarization entropy and the first interference coherence. The phase unwrapping module is used to unwrap the wrapped phase generated by each target interferometer pair in the target interferometer pair set to obtain the true phase. The tree height inversion module is used to construct phase consistency constraints based on the true phase, and set several different inversion tree height values ​​under the constraints of phase consistency constraints. For each inversion tree height value, the inversion tree height value is input into a preset vegetation interference coherence model to determine the second interference coherence, and the inversion tree height value is input into a preset polarization coherence entropy model to determine the second polarization coherence entropy. The objective function construction module is used to construct an objective function based on the difference between the first interference coherence and the second interference coherence, as well as the difference between the first polarization coherence entropy and the second polarization coherence entropy. The target tree height determination module is used to take the inversion tree height value corresponding to the minimum value of the objective function as the target tree height value. The tree obstacle risk early warning module is used to determine the tree obstacle risk level based on the second polarization coherence entropy corresponding to the target tree height, and to issue an early warning signal when the tree obstacle risk level exceeds the preset risk level.

[0084] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can implement the transmission line tree obstacle risk monitoring method provided by any of the above-described method embodiments of the present invention.

[0085] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can specifically be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0086] Based on the above embodiments of the transmission line tree obstacle risk monitoring method, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the transmission line tree obstacle risk monitoring method of any embodiment of the present invention.

[0087] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0088] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0089] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting various parts of the terminal device via various interfaces and lines.

[0090] Based on the above-described method embodiments, another embodiment is provided: another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the transmission line tree obstacle risk monitoring method described in any of the above-described method embodiments of the present invention.

[0091] The modules / units integrated into the power transmission line tree obstacle risk monitoring device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0092] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for monitoring tree obstruction risk on power transmission lines, characterized in that, include: Acquire environmental data, historical interferometric data, and several polarization interferometric images of the power transmission corridor under test; Based on environmental data, historical interferometric data, and several polarimetric interferometric images, the target interferometric pair set is determined; Extract the polarization entropy and first interference coherence of each target interference pair in the target interference pair set; Based on the polarization entropy and the first interference coherence, construct the first polarization coherence entropy; The entangled phase generated by each target interferometer pair in the target interferometer pair set is untangled to obtain the true phase; Based on the true phase, a phase consistency constraint is constructed, and under the constraint of the phase consistency constraint, several different inversion tree height values ​​are set. For each inversion tree height value, the inversion tree height value is input into a preset vegetation interference coherence model to determine the second interference coherence, and the inversion tree height value is input into a preset polarization coherence entropy model to determine the second polarization coherence entropy. Based on the difference between the first interference coherence and the second interference coherence, and the difference between the first polarization coherence entropy and the second polarization coherence entropy, an objective function is constructed. The height of the inversion tree corresponding to the minimum value of the objective function is taken as the height of the objective tree. The tree obstacle risk level is determined based on the second polarization coherence entropy corresponding to the target tree height, and an early warning signal is issued when the tree obstacle risk level exceeds the preset risk level.

2. The method for monitoring tree obstruction risk on transmission lines according to claim 1, characterized in that, After acquiring environmental data, historical interferometric data, and several polarization interferometric images of the transmission corridor under test, the process also includes: Select one polarimetric interferometric image from all polarimetric interferometric images as the master image; Based on the main image, all polarimetric interferometric images are registered to obtain the registered polarimetric interferometric images.

3. The method for monitoring tree obstruction risk on transmission lines according to claim 1, characterized in that, Based on environmental data, historical interferometric data, and several polarimetric interferometric images, the target interferometric pair set is determined, including: Construct the current state vector based on environmental data and historical interference data; For each interference pair, the corresponding spatial gradient is calculated based on the interference phase formed by the interference pair, and the spatial gradient is discretized to obtain the corresponding gradient energy; the interference pair is composed of any two different polarization interference images. Interference pairs corresponding to gradient energies less than a preset stability threshold are selected as candidate interference pairs. Construct the action space based on the temporal baselines of all candidate interference pairs; In the action space, based on the current state vector, the time baseline of the next interferometric pair is selected step by step according to the current policy Q, until the total time span of the interferometric pair corresponding to the selected time baseline covers the time period of all polarization interferometric images, thus obtaining the target interferometric pair set.

4. The method for monitoring tree obstruction risk on transmission lines according to claim 1, characterized in that, Extract the polarization entropy and first interference coherence of each target interference pair in the target interference pair set, including: For each target interference pair in the target interference pair set, a coherence matrix is ​​constructed based on the complex scattering components of horizontal transmission-horizontal reception, horizontal transmission-vertical reception, and vertical transmission-vertical reception in the target interference pair. Hermitian eigenvalues ​​are obtained by performing Hermitian eigenvalue decomposition on the coherence matrix; Each scattering eigenvalue is normalized to obtain the normalized scattering eigenvalue; The polarization entropy is determined based on all normalized scattering eigenvalues. The first interference coherence is determined based on the scattering vector of each target interference pair in the target interference pair set.

5. The method for monitoring tree obstruction risk on transmission lines according to claim 1, characterized in that, The entangled phase generated by each target interferometer pair in the target interferometer pair set is unwrapped to obtain the true phase, including: Generate the corresponding entanglement phase for each target interference pair in the target interference pair set; A coupling term is constructed based on the intertwined phase of adjacent pixels; Obtain the terrain prior knowledge DEM elevation information, and construct the bias term based on the winding phase and the terrain prior knowledge DEM elevation information; Construct the binary compensation variables to be solved; Based on the coupling term, bias term, and binary compensation variable, construct the Hamiltonian energy function of the Ising model; Quantum annealing is performed on the Hamiltonian energy function of the Ising model to obtain the target compensation variable; The untangling phase is determined based on the entanglement phase and the target compensation variable; Anisotropic diffusion filtering is applied to the unwrapped phase to obtain the true phase.

6. The method for monitoring tree obstruction risk on transmission lines according to claim 1, characterized in that, The phase consistency constraint includes: ； in, Indicates the true phase. Indicates the height of the inversion tree. Indicates the height of the inversion tree The corresponding second interference coherence.

7. The method for monitoring tree obstruction risk on transmission lines according to claim 1, characterized in that, The objective function includes: ； in, Describe the objective function. Indicates the first interference coherence. Indicates the second interference coherence. Represents the first polarization coherence entropy. Represents the coherent entropy of the second polarization. This indicates the ratio of scattering contribution from the land surface to that from vegetation.

8. A power transmission line tree obstruction risk monitoring device, characterized in that, include: The system includes a power transmission corridor data acquisition module, an interference pair screening module, a polarization interference parameter extraction module, a polarization coherence entropy construction module, a phase unwrapping module, a tree height inversion module, an objective function construction module, a target tree height determination module, and a tree obstacle risk early warning module. The power transmission corridor data acquisition module is used to acquire environmental data, historical interferometric data and several polarization interferometric images of the power transmission corridor under test; The interference pair screening module is used to determine the target interference pair set based on environmental data, historical interference data, and several polarization interference images; The polarization interference parameter extraction module is used to extract the polarization entropy and the first interference coherence of each target interference pair in the target interference pair set; The polarization coherence entropy construction module is used to construct the first polarization coherence entropy based on the polarization entropy and the first interference coherence. The phase unwrapping module is used to unwrap the wrapped phase generated by each target interference pair in the target interference pair set to obtain the true phase. The tree height inversion module is used to construct phase consistency constraints based on the true phase, and set several different inversion tree height values ​​under the constraints of the phase consistency constraints. For each inversion tree height value, the inversion tree height value is input into a preset vegetation interference coherence model to determine the second interference coherence, and the inversion tree height value is input into a preset polarization coherence entropy model to determine the second polarization coherence entropy. The objective function construction module is used to construct an objective function based on the difference between the first interference coherence and the second interference coherence, and the difference between the first polarization coherence entropy and the second polarization coherence entropy. The target tree height determination module is used to take the inversion tree height value corresponding to the minimum value of the objective function as the target tree height value; The tree obstacle risk early warning module is used to determine the tree obstacle risk level based on the second polarization coherence entropy corresponding to the target tree height, and to issue an early warning signal when the tree obstacle risk level exceeds the preset risk level.

9. A terminal device, characterized in that, The method includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the transmission line tree obstacle risk monitoring method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform the transmission line tree obstacle risk monitoring method as described in any one of claims 1-7.

Images