A power transmission line abnormality monitoring system and method based on shock absorber vibration monitoring

Abstract

The application provides a kind of transmission line abnormal monitoring system and method based on shock absorber vibration monitoring, belongs to the technical field of power system monitoring.The system includes vibration sensor installed on each shock absorber, data collection and energy extraction unit set on tower, and intelligent monitoring terminal.The method includes: obtaining the vibration waveform data of each shock absorber and dividing clusters;for each shock absorber, compare its waveform features with the features of adjacent shock absorbers, filter abnormal shock absorbers and determine their fault type and location;for each shock absorber cluster, eliminate the above abnormal points to form a purified cluster, and locate the abnormal line section by comparing the feature differences between clusters;finally, determine whether the vibration of the abnormal line is dominated by wind-induced vibration, if not, further identify the specific risk type.The application realizes accurate positioning and classification of shock absorber individual faults, as well as intelligent identification and early warning of transmission line section systematic risk.

Description

Technical Field

[0001] This invention relates to the field of power system monitoring technology, and specifically to a transmission line anomaly monitoring system and method based on vibration monitoring of anti-vibration hammers. Background Technology

[0002] High-voltage transmission lines are exposed to complex outdoor environments for extended periods, making them particularly susceptible to wind-induced vibrations. To suppress conductor vibrations and prevent strand breakage or fracture due to long-term fatigue, vibration dampers are widely installed on transmission lines. The mechanical condition of the vibration dampers themselves (e.g., whether they are loose, corroded, or damaged) and their effectiveness in suppressing conductor vibrations directly affect the mechanical safety and operational lifespan of the transmission lines. Therefore, effective condition monitoring of the vibration dampers and the line sections they protect is a crucial aspect of intelligent operation and maintenance and disaster prevention early warning in power systems.

[0003] Currently, monitoring in this field mainly relies on regular manual inspections and simple single-point threshold alarms. These methods have significant drawbacks: First, manual inspections are inefficient and costly, and cannot achieve real-time or near-real-time monitoring, making it difficult to detect sudden faults or gradual deterioration in a timely manner. Second, existing technologies lack precise and automated diagnostic capabilities for mechanical faults in the vibration dampers themselves, and also fail to effectively distinguish the root causes of abnormal line vibrations. For example, they cannot reliably distinguish between normal vibrations caused by generalized wind forces and abnormal vibrations caused by potential risks such as line metal fatigue and structural resonance, leading to generalized early warning information and failing to provide clear guidance for precise maintenance and risk prevention. Furthermore, the lack of scientific basis for material selection and proportioning design for specific road conditions makes it difficult to conduct standardized simulations and comparative tests. Summary of the Invention

[0004] This invention provides a transmission line anomaly monitoring system and method based on vibration monitoring of anti-vibration hammers, which solves the problems of low efficiency due to reliance on manual inspection, difficulty in real-time and accurate location and diagnosis of mechanical faults of the anti-vibration hammers themselves, and inability to effectively distinguish the causes of line anomalies in the existing technology.

[0005] To achieve the above objectives, one embodiment of the present invention provides a transmission line anomaly monitoring system based on vibration monitoring of anti-vibration hammers, comprising: a vibration sensor installed on each anti-vibration hammer of the transmission line for collecting vibration waveform data of the corresponding anti-vibration hammer, each vibration sensor having a pre-set unique location identifier; a data collection and energy harvesting unit disposed on each transmission tower supporting the anti-vibration hammer, the data collection and energy harvesting unit including a wireless communication unit and a current transformer, the wireless communication unit being used to wirelessly communicate with the vibration sensor to receive the vibration waveform data and location identifier transmitted by it; the current transformer being used to sense and obtain electrical energy from the transmission line and to power itself and the vibration sensor; and an intelligent monitoring terminal communicatively connected to the data collection and energy harvesting unit for receiving vibration waveform data, analyzing and identifying the operating status of the anti-vibration hammer and the line based on the vibration waveform data, and outputting the transmission line monitoring results.

[0006] On the other hand, a method for monitoring transmission line anomalies based on vibration damper monitoring is also provided, applied to the aforementioned transmission line anomaly monitoring system, including: Step S1: Acquire the vibration waveform data of each vibration damper and extract the waveform features corresponding to each vibration damper; divide the vibration dampers into multiple vibration damper clusters according to a preset spatial adjacency relationship; Step S2: For each vibration damper, compare and analyze its waveform features with the waveform features of other vibration dampers within its adjacent preset range to screen out abnormal vibration dampers; Step S3: Based on the waveform data of the abnormal vibration dampers, determine the anomaly type of the abnormal vibration damper through an anomaly feature classification strategy, and combine it with the location identifier of the abnormal vibration damper to form a first monitoring result; Step S4: For each vibration damper... The first monitoring result is used to eliminate abnormal vibration dampers and form a clean vibration damper cluster. Based on the vibration waveform data of all vibration damper clusters, a difference analysis strategy is used to compare between clusters to identify abnormal vibration damper clusters, and the line range corresponding to the abnormal vibration damper clusters is taken as the abnormal line. Step S5: For the abnormal line, based on the vibration waveform data of its clean vibration damper cluster, it is determined whether it is wind-induced vibration. If it is determined to be non-wind-induced vibration, the risk type of the abnormal line is identified based on the waveform characteristics of the clean vibration damper cluster through a risk pattern recognition strategy, and a second monitoring result is formed by combining the section location information of the abnormal line. The first monitoring result and the second monitoring result are fused to form the transmission line detection result.

[0007] Optionally, the abnormal feature classification strategy includes: extracting a multi-dimensional feature vector containing time-domain features, frequency-domain features, and time-frequency-domain features from the waveform data of the abnormal vibration damper; comparing each feature value in the multi-dimensional feature vector with a preset multi-dimensional threshold condition set, wherein the multi-dimensional threshold condition set includes logical judgment conditions corresponding to different abnormal types, and each logical judgment condition includes multiple feature dimension thresholds; and determining the abnormal type that satisfies all of its logical judgment conditions as the abnormal type of the abnormal vibration damper.

[0008] Optionally, the formation of the purification anti-vibration hammer cluster includes: locating the anti-vibration hammer cluster to which each abnormal anti-vibration hammer belongs based on the location identifier of each abnormal anti-vibration hammer in the first monitoring result; removing the vibration waveform data of the abnormal anti-vibration hammer from the dataset of the anti-vibration hammer cluster to which it belongs; and the remaining anti-vibration hammers and their vibration waveform data after removal constitute the purification anti-vibration hammer cluster.

[0009] Optionally, the difference analysis strategy includes: for each vibration damper cluster, extracting the core feature parameters of the vibration waveform of the vibration damper based on the dataset of its corresponding purification vibration damper cluster; performing statistical aggregation operations on the core feature parameters to calculate feature statistical values ​​to characterize the overall vibration state of the cluster; comparing the feature statistical values ​​of the current vibration damper cluster with the feature statistical values ​​of other vibration damper clusters within its adjacent preset range to calculate their difference degree; if the difference degree exceeds a preset cluster difference threshold, then the current vibration damper cluster is determined to be an abnormal vibration damper cluster.

[0010] Optionally, the determination of wind-induced vibration includes: aligning the vibration waveform data of all vibration dampers in the purification vibration damper cluster corresponding to the abnormal line on the time axis to reduce data acquisition delay deviation; calculating the time-domain synchronization coefficient and frequency domain dominant frequency band overlap of the vibration waveform of each vibration damper in the purification vibration damper cluster based on the aligned vibration waveform data; and comparing the time-domain synchronization coefficient and frequency domain dominant frequency band overlap with the wind-induced vibration determination threshold to determine whether it is wind-induced vibration.

[0011] Optionally, the calculation of the time-domain synchronization coefficient and the frequency domain dominant frequency band overlap includes: calculating the cross-correlation coefficient between each pair of vibration waveforms of each vibration damper in the purification vibration damper cluster, and taking the average of all cross-correlation coefficients as the time-domain synchronization coefficient; performing frequency domain transformation on the vibration waveform data of each vibration damper, extracting the dominant frequency band with the highest vibration energy proportion of each vibration damper, and calculating the ratio of the intersection interval length of all dominant frequency bands in the purification vibration damper cluster to the average length of each dominant frequency band as the frequency domain dominant frequency band overlap.

[0012] Optionally, the risk pattern recognition strategy includes: extracting line-level feature parameters from the vibration waveform data of each purification vibration damper cluster corresponding to the abnormal line, wherein the line-level feature parameters include the dominant frequency of the vibration mode, the cross-cluster vibration phase consistency coefficient, and the vibration attenuation rate; comparing each line-level feature parameter with a preset normal state feature baseline, calculating the deviation degree and deviation duration of each parameter, and forming a multi-dimensional deviation dataset; based on the multi-dimensional deviation dataset, analyzing the deviation patterns of each line-level feature parameter and the deviation degree combination relationship between different line-level feature parameters; and identifying the metal fatigue risk type corresponding to the abnormal line based on the deviation patterns and deviation degree combination relationships of each line-level feature parameter, combined with a preset metal fatigue risk judgment rule.

[0013] Optionally, the calculation process of the deviation pattern and the deviation degree combination relationship includes: for each line-level feature parameter in the multi-dimensional deviation dataset, extracting the sequence data of its deviation degree changing over time, and obtaining the changing trend parameter of the line-level feature parameter through trend fitting, as the deviation pattern; merging and calculating all deviation degrees to obtain a comprehensive combined deviation value, and numerically sorting the deviation degrees of each line-level feature parameter to determine the dominant and auxiliary parameters of the deviation degree; and using the comprehensive combined deviation value and the parameter primary and secondary sorting results as the deviation degree combination relationship.

[0014] Optionally, the comparison analysis includes: calculating the feature difference degree between the waveform features of the current vibration damper and the waveform features of other vibration dampers within an adjacent preset range, comparing the feature difference degree with a preset single-point anomaly threshold; when the feature difference degree exceeds the single-point anomaly threshold, determining that the vibration damper is an abnormal vibration damper.

[0015] The present invention provides a transmission line anomaly monitoring system and method based on vibration monitoring of anti-vibration hammers. Through a monitoring system consisting of a sensor network deployed on the anti-vibration hammers and self-powered units on the tower side, it realizes real-time and automated data acquisition and transmission of the vibration status of transmission lines. Based on a progressive diagnostic method from individual to cluster and from component to system, it can simultaneously and accurately locate and classify mechanical faults of individual anti-vibration hammers, as well as intelligently identify the systemic risk types of line sections. This upgrades the operation and maintenance mode from inefficient manual periodic inspections to efficient intelligent online diagnosis and accurate early warning, significantly improving the safety and intelligent operation and maintenance level of transmission lines. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0017] Figure 1 This is a flowchart of the transmission line anomaly monitoring method provided in the embodiments of the present invention; Figure 2 This is a flowchart of single-point anomaly screening and classification provided in an embodiment of the present invention; Figure 3 This is a flowchart of the purification cluster and abnormal line determination provided in the embodiments of the present invention; Figure 4 This is a flowchart of risk pattern recognition provided in an embodiment of the present invention. Detailed Implementation

[0018] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0019] It should be noted that the acquisition, transmission, storage, use, and processing of data in the technical solution of this application all comply with the relevant provisions of national laws and regulations. In the embodiments of this application, certain existing industry solutions such as software, components, and models may be mentioned. These should be considered exemplary, intended only to illustrate the feasibility of implementing the technical solution of this application, and do not imply that the applicant has already used or necessarily used such solutions.

[0020] The following is combined Figures 1-4 This invention is described in detail.

[0021] This invention provides a transmission line anomaly monitoring system based on vibration monitoring of anti-vibration hammers, comprising: vibration sensors installed on each anti-vibration hammer of the transmission line to collect vibration waveform data of the corresponding anti-vibration hammer, each vibration sensor having a pre-set unique location identifier; a data collection and energy harvesting unit disposed on each transmission tower supporting the anti-vibration hammer, the data collection and energy harvesting unit including a wireless communication unit and a current transformer, the wireless communication unit being used to wirelessly communicate with the vibration sensor to receive the vibration waveform data and location identifier transmitted by it; the current transformer being used to sense and obtain electrical energy from the transmission line and to power itself and the vibration sensor; and an intelligent monitoring terminal communicatively connected to the data collection and energy harvesting unit, used to receive vibration waveform data, analyze and identify the operating status of the anti-vibration hammer and the line based on the vibration waveform data, and output the transmission line monitoring results.

[0022] The vibration damper is a core protective component installed on power transmission lines to suppress conductor vibration and protect the line structure. The vibration sensor is a sensing component that converts mechanical vibration signals into electrical signals. The vibration waveform data it collects is the fundamental data reflecting the real-time vibration state of the vibration damper, including dynamic changes in vibration. The location marker is a unique identifier for each vibration sensor, ensuring a one-to-one correspondence between the sensor and its installation point, guaranteeing accurate association of collected data with the corresponding vibration damper. The data collection and energy harvesting unit is an integrated hardware component combining data transmission and power acquisition functions. This unit is fixedly installed on the transmission tower and is suitable for outdoor applications without external wiring. The wireless communication unit is a communication component that uses wireless transmission for data interaction. This component establishes a stable communication link with the vibration sensor, reliably receiving vibration waveform data and location markers. The current transformer is an electrical component operating on the principle of electromagnetic induction. This component can obtain induced electrical energy from the current-carrying conductors of the transmission line in a non-contact manner, meeting the power requirements of itself and the vibration sensor. The intelligent monitoring terminal is a core control device capable of receiving, analyzing, processing, and outputting data. This device acquires data collected along the entire line via a communication link and uses data processing to identify and determine the operational status. The operational status includes two types of information: the working status of the vibration damper itself and the operational status of the transmission line itself. This is the core object of the system's monitoring and analysis. The transmission line monitoring results are the final conclusions output by the system after analysis and calculation. These conclusions directly reflect whether there are any abnormalities in the operation of the line and the vibration damper.

[0023] Specifically, during system operation, vibration sensors pre-installed on each vibration damper continuously collect vibration waveform data of the corresponding damper. Each vibration sensor distinguishes the collected information by a preset unique position identifier. The data collection and energy harvesting unit installed on the transmission tower establishes wireless communication with the vibration sensor through its internal wireless communication unit, receiving the vibration waveform data and position identifier transmitted by the vibration sensor. The current transformer inside this unit senses and obtains electrical energy from the transmission line, providing the power required for the data collection and energy harvesting unit itself and the vibration sensor. The data collection and energy harvesting unit transmits the received relevant data to the intelligent monitoring terminal. The intelligent monitoring terminal analyzes and processes the acquired vibration waveform data, identifies the operating status of the vibration damper and the transmission line, and finally outputs the corresponding transmission line monitoring results.

[0024] The power transmission line anomaly monitoring system provided in this embodiment of the invention achieves accurate vibration data acquisition and point binding through vibration sensors with dedicated location identifiers. It completes wireless data transmission and outdoor self-powered operation by relying on a data collection and energy harvesting unit that integrates wireless communication and line induction energy harvesting. In addition, it achieves data processing and status identification by intelligent monitoring terminal. It can operate stably in field power transmission scenarios without external power supply, accurately collect and transmit data, and efficiently complete the monitoring of the operating status of the line and vibration damper and output the results.

[0025] like Figure 1 As shown, this embodiment of the invention also provides a method for monitoring transmission line anomalies based on vibration monitoring of anti-vibration hammers, applied to the aforementioned transmission line anomaly monitoring system, including: Step S1: Acquire vibration waveform data of each anti-vibration hammer and extract waveform features corresponding to each anti-vibration hammer; divide the anti-vibration hammers into multiple anti-vibration hammer clusters according to a preset spatial adjacency relationship; Step S2: For each anti-vibration hammer, compare and analyze its waveform features with the waveform features of other anti-vibration hammers within its adjacent preset range to screen out abnormal anti-vibration hammers; Step S3: Based on the waveform data of abnormal anti-vibration hammers, determine the anomaly type of the abnormal anti-vibration hammer through an anomaly feature classification strategy, and form a first monitoring result by combining the location identifier of the abnormal anti-vibration hammer; Step S4: For each The vibration damper cluster is formed by removing abnormal vibration dampers based on the first monitoring result, resulting in a purified vibration damper cluster. Based on all vibration waveform data within the purified vibration damper cluster, an inter-cluster comparison is performed using a difference analysis strategy to identify abnormal vibration damper clusters, and the line range corresponding to these abnormal vibration damper clusters is designated as the abnormal line. Step S5: For the abnormal line, based on the vibration waveform data of its purified vibration damper cluster, it is determined whether the vibration is wind-induced. If it is determined to be non-wind-induced, the risk type of the abnormal line is identified using a risk pattern recognition strategy based on the waveform characteristics of the purified vibration damper cluster, and a second monitoring result is formed by combining the section location information of the abnormal line. The first monitoring result and the second monitoring result are then fused to form the transmission line detection result.

[0026] Among them, waveform characteristics refer to the set of parameters that characterize the vibration properties of vibration dampers. Spatial adjacency refers to the proximity association formed based on the distribution of transmission line towers, span division, and the actual installation location of vibration dampers. Vibration damper cluster refers to a group of multiple vibration dampers grouped together according to spatial adjacency. Preset adjacent range refers to the range of surrounding equipment defined for feature comparison for a single vibration damper. First monitoring result refers to the monitoring conclusion that integrates the location and anomaly category of abnormal vibration dampers. Cleaned vibration damper cluster refers to the cluster consisting of normal equipment and valid data after removing abnormal vibration damper data. Inter-group comparison refers to the comparison of characteristic parameters between different vibration damper clusters. Abnormal vibration damper cluster refers to a cluster of vibration dampers whose overall vibration characteristics deviate from the standard state. Abnormal line refers to the transmission line range corresponding to the abnormal vibration damper cluster. Wind-induced vibration refers to non-fault-related natural vibration caused by natural wind. Risk type refers to the form of structural anomaly or performance degradation of the transmission line itself. Section location information refers to the location information used to clarify the specific distribution range of abnormal lines. The second monitoring result refers to the monitoring conclusion that integrates information on line risk types and section locations. The transmission line inspection result refers to the final monitoring conclusion covering both vibration dampers and the line itself. Wind-induced vibration refers to vibration caused by wind.

[0027] Specifically, waveform features include time-domain and frequency-domain features. Time-domain features include vibration amplitude and duration, while frequency-domain features include dominant vibration frequency and waveform distortion rate. By performing time-domain statistical calculations and frequency-domain transformations on the acquired vibration waveform data, the corresponding waveform features can be extracted. When dividing vibration damper clusters based on preset spatial adjacency relationships, all vibration dampers deployed between three consecutive transmission towers can be grouped into one cluster. The preset adjacency range is preferably the area covered by three adjacent vibration dampers before and after the current damper, ensuring the accuracy and comprehensiveness of feature comparison.

[0028] Specifically, the segment location information is determined through multi-layer location data association. The location identifier of the vibration sensor is pre-bound to the corresponding transmission tower number and line span number. Each vibration damper cluster is pre-associated with a fixed starting transmission tower number, ending transmission tower number, and corresponding line span range. After identifying an abnormal vibration damper cluster, the pre-bound tower number and span range of that cluster can be directly retrieved to accurately determine the segment location of the abnormal line. For example, if an abnormal vibration damper cluster is pre-bound to starting transmission tower number 22 and ending transmission tower number 25, corresponding to line spans 22 to 23, 23 to 24, and 24 to 25, then the segment location of the abnormal line can be directly determined to be the transmission line segment between transmission tower number 22 and transmission tower number 25.

[0029] The transmission line anomaly monitoring method provided in this invention can accurately screen and classify single-point anomalies of anti-vibration hammers through hierarchical monitoring and analysis, locate abnormal lines by relying on data purification to eliminate interference, effectively distinguish between wind-induced vibration and line body risks, and finally integrate to form a comprehensive monitoring result, which greatly improves the accuracy and reliability of transmission line anomaly judgment.

[0030] like Figure 2 As shown, preferably, the comparison analysis includes: calculating the feature difference degree between the waveform features of the current vibration damper and the waveform features of other vibration dampers within an adjacent preset range, comparing the feature difference degree with a preset single-point anomaly threshold; when the feature difference degree exceeds the single-point anomaly threshold, the vibration damper is determined to be an abnormal vibration damper.

[0031] Further preferably, the abnormal feature classification strategy includes: extracting a multi-dimensional feature vector containing time-domain features, frequency-domain features, and time-frequency-domain features from the waveform data of the abnormal vibration damper; comparing each feature value in the multi-dimensional feature vector with a preset multi-dimensional threshold condition set, wherein the multi-dimensional threshold condition set includes logical judgment conditions corresponding to different abnormal types, and each logical judgment condition includes multiple feature dimension thresholds; and determining the abnormal type that satisfies all of its logical judgment conditions as the abnormal type of the abnormal vibration damper.

[0032] Among them, multidimensional feature vectors refer to feature sets formed by combining various types of vibration features. Time-domain features refer to parameters representing vibration changes extracted from the time dimension of the vibration waveform. Frequency-domain features refer to parameters representing the vibration frequency distribution extracted after converting the vibration waveform to the frequency dimension. Time-frequency domain features refer to parameters that reflect the dynamic change law of vibration extracted by combining the time and frequency dimensions. Multidimensional threshold condition set refers to the parameter set storing the judgment criteria corresponding to various anomaly types. Logical judgment conditions refer to all feature constraints that must be satisfied to distinguish a certain anomaly type. Feature dimension threshold refers to the judgment critical value set for a single feature dimension. Anomaly type refers to the specific classification and naming of the mechanical faults or performance degradation of the vibration damper itself. Anomaly types include mechanical loosening, rust jamming, structural cracks or damage. Feature difference degree refers to the numerical index used to quantify the degree of difference between the waveform features of a single vibration damper and adjacent vibration dampers. Single-point anomaly threshold refers to the critical value used to determine whether a single vibration damper has an anomaly.

[0033] Specifically, when conducting comparative analysis on a single vibration damper, the waveform characteristics of the current vibration damper and other vibration dampers within a preset adjacent range are first extracted. These waveform characteristics include time-domain and frequency-domain features. The feature difference degree is obtained by calculating the absolute value, variance, or Euclidean distance between these parameters. This feature difference degree is then numerically compared with a preset single-point anomaly threshold. If the feature difference degree is greater than the single-point anomaly threshold, the vibration damper is determined to be an abnormal vibration damper. For example, the effective vibration value is selected as the time-domain feature parameter for comparison. The system reads the digital acceleration sampling sequence of the current vibration damper #5 within the most recent analysis time window. By calculating the square root of the square of all sampling points in the sequence, taking the average, and then the square root, its effective vibration value is obtained as 0.8g. The preset adjacent range is set to two vibration dampers before and after it, namely #3, #4, #6, and #7. The calculated effective vibration values ​​of these four adjacent vibration dampers are 0.2g, 0.25g, 0.18g, and 0.22g, respectively. The statistical measure—the average value—of the effective values ​​of this group of adjacent hammers is calculated, yielding 0.2125g. The characteristic difference between the characteristic value of target hammer #5 and the average value of adjacent hammers is calculated using the absolute value of the difference method, i.e., |0.8g - 0.2125g| = 0.5875g. The system's preset single-point anomaly threshold is 0.4g. Since the calculated characteristic difference of 0.5875g is greater than the threshold of 0.4g, the system determines that vibration damper #5 is an abnormal vibration damper, completing the initial screening.

[0034] Specifically, the multidimensional feature vector is composed of time-domain features, frequency-domain features, and time-frequency-domain features. The time-domain features include parameters such as vibration effective value, peak value, and waveform factor to quantify vibration intensity and impact. The frequency-domain features include parameters such as dominant frequency, center of gravity frequency, and energy proportion of a specific frequency band to describe the frequency distribution characteristics of vibration energy. The time-frequency-domain features include parameters such as wavelet energy entropy and wavelet coefficient energy at a specific scale to characterize the local time-frequency characteristics of non-stationary vibration. By performing multidimensional analysis on the vibration waveform data of the abnormal vibration damper, the corresponding multidimensional feature vector can be obtained.

[0035] Specifically, each anomaly type in the multidimensional threshold condition set corresponds to an independent logical judgment condition. Each logical judgment condition contains threshold constraints for multiple feature dimensions. Each feature value in the multidimensional feature vector is compared with the feature dimension threshold in the corresponding logical judgment condition. When all feature values ​​meet the corresponding threshold requirements, the corresponding anomaly type can be matched and determined.

[0036] For example, the construction and execution of logical judgment conditions: A logical judgment condition is defined for each anomaly type to be identified. Each condition consists of multiple feature dimension thresholds connected by a logical AND operator. For instance, the judgment condition for mechanical loosening can be set as: the effective vibration value is greater than threshold A and the dominant frequency fluctuation range is less than threshold B. The judgment condition for rust and jamming can be set as: the total vibration energy is less than threshold C and the energy proportion above 50Hz is greater than threshold D. The judgment condition for structural cracks or damage can be set as: the peak value of the impact component in the time-domain waveform is greater than threshold E and the wavelet energy entropy is lower than threshold F. During classification, the system substitutes the multi-dimensional feature vector extracted from the abnormal vibration damper into each logical judgment condition for calculation. If all inequalities in a certain condition are satisfied, the vibration damper is determined to belong to the anomaly type corresponding to that condition.

[0037] The technical content described in the preferred embodiments of this invention, through a difference calculation method based on proximity comparison, can efficiently and automatically screen potential abnormal individuals from a large number of vibration dampers, achieving rapid preliminary location of fault points. Furthermore, through a classification strategy based on multi-dimensional features and logical threshold judgment, the screened abnormal vibration dampers can be analyzed in depth to accurately determine their specific mechanical fault types. The combination of these two methods effectively overcomes the shortcomings of traditional monitoring methods, such as high false alarm rates and difficulty in identifying fault root causes, significantly improving the targeting and efficiency of transmission line condition-based maintenance, and accurately determining fault types and causes, directly supporting the decision-making upgrade from abnormal alarms to precise maintenance.

[0038] like Figure 3 As shown, preferably, the formation of the purification anti-vibration hammer cluster includes: locating the anti-vibration hammer cluster to which each abnormal anti-vibration hammer belongs based on the location identifier of each abnormal anti-vibration hammer in the first monitoring result; removing the vibration waveform data of the abnormal anti-vibration hammer from the dataset of the anti-vibration hammer cluster to which it belongs; and the remaining anti-vibration hammer and its vibration waveform data after removal constitute the purification anti-vibration hammer cluster.

[0039] Further preferably, the difference analysis strategy includes: for each vibration damper cluster, extracting the core feature parameters of the vibration waveform of each vibration damper based on the dataset of its corresponding purification vibration damper cluster; performing statistical aggregation operations on the core feature parameters to calculate feature statistical values ​​to characterize the overall vibration state of the cluster; comparing the feature statistical values ​​of the current vibration damper cluster with the feature statistical values ​​of other vibration damper clusters within its adjacent preset range to calculate their difference degree; if the difference degree exceeds a preset cluster difference threshold, then the current vibration damper cluster is determined to be an abnormal vibration damper cluster.

[0040] The first monitoring result refers to a structured dataset generated during the individual anomaly diagnosis phase, containing unique location identifiers of vibration dampers identified as abnormal and their corresponding anomaly types. A location identifier is coded information used to uniquely determine the physical installation location of a vibration damper within the system. A dataset refers to the collection of vibration waveform data and associated information for all vibration dampers belonging to the same vibration damper cluster. An abnormal vibration damper refers to a vibration damper that has been screened out in the individual comparison analysis and whose specific anomaly type has been determined through feature classification. A purified vibration damper cluster refers to a new set of data consisting of the remaining, considered normal vibration damper data, after removing all data entries corresponding to the abnormal vibration dampers from the original vibration damper cluster dataset. A core feature parameter refers to a key indicator extracted from the vibration waveform of a single vibration damper to quantify its vibration state. In a preferred embodiment of the invention, the core feature parameter is the vibration energy value, a time-domain parameter characterizing the overall vibration intensity level over a period of time. A cluster difference threshold refers to a pre-set critical value used to determine whether the difference between clusters is significant enough to identify one as abnormal.

[0041] Specifically, the core characteristic parameter is the vibration energy value. The calculation formula is: in, Let be the acceleration value of the i-th sampling point, and n be the total number of sampling points within this time window.

[0042] Specifically, the formula for statistical aggregation operations is:

[0043] in, The characteristic statistical values ​​representing the cluster of anti-vibration hammers for purification are as follows: To reduce the number of vibration dampers within the vibration damper cluster, Let be the vibration energy value of the j-th vibration damper.

[0044] Specifically, the formula for calculating the degree of difference is:

[0045] in, This represents the absolute value of the difference between two characteristic statistics. Indicates the larger of the two values. Dissimilarity It is a dimensionless number between 0 and 1, and the larger the value, the more significant the difference in vibrational energy levels between the two clusters.

[0046] For example, suppose a transmission line section contains adjacent vibration damper clusters A and B. Initially, cluster A contains 5 vibration dampers, numbered A1 to A5, with calculated vibration energy values ​​of [0.2, 0.8, 0.18, 0.22, 0.19]. Cluster B contains 4 vibration dampers, numbered B1 to B4, with vibration energy values ​​of [0.21, 0.19, 0.23, 0.20]. The first monitoring result shows that damper A2 is identified as having abnormal mechanical loosening. Based on its location identifier, the system removes A2's data from the original dataset of cluster A. After removal, the cleaned dataset of cluster A contains the energy values ​​of dampers A1, A3, A4, and A5: [0.2, 0.18, 0.22, 0.19]. Subsequently, the characteristic statistics of cluster A are calculated. =(0.2+0.18+0.22+0.19) / 4=0.1975g. The characteristic statistics of cluster B are: =(0.21+0.19+0.23+0.20) / 4=0.2075g. Then, the difference between clusters A and B is calculated using the difference calculation formula: = |0.1975 0.2075 | / 0.2075 ≈ 0.0482. Assuming the cluster difference threshold is 0.10, since the calculated difference degree D ≈ 0.0482 < 0.10, it does not exceed the threshold. Therefore, cluster A is determined not to be an abnormal vibration damper cluster. This indicates that after removing the known faulty damper A2, the overall vibration energy level of the line section under the jurisdiction of cluster A is not significantly different from that of the adjacent cluster B, and this section currently does not show any systemic abnormal risk.

[0047] In a preferred embodiment of the present invention, a data purification step effectively eliminates data interference from known individual faulty vibration dampers, ensuring the reliability of the data source used for subsequent analysis. On this basis, the difference analysis strategy, by quantitatively comparing the overall vibration characteristics of adjacent purified vibration damper clusters, can sensitively identify significant anomalies in the overall state of a certain line section. This achieves an upgrade from locating scattered component fault points to assessing continuous line risk sections, providing a systematic diagnostic basis from point to surface for the precise operation and maintenance of transmission lines.

[0048] Preferably, the determination of wind-induced vibration includes: aligning the vibration waveform data of all vibration dampers in the purification vibration damper cluster corresponding to the abnormal line on the time axis to reduce data acquisition delay deviation; calculating the time-domain synchronization coefficient and frequency domain dominant frequency band overlap of the vibration waveform of each vibration damper in the purification vibration damper cluster based on the aligned vibration waveform data; and comparing the time-domain synchronization coefficient and frequency domain dominant frequency band overlap with the wind-induced vibration determination threshold to determine whether it is wind-induced vibration.

[0049] Further preferably, the calculation of the time-domain synchronization coefficient and the frequency-domain dominant frequency band overlap includes: calculating the cross-correlation coefficient between each pair of vibration waveforms of each vibration damper in the purification vibration damper cluster, and taking the average of all cross-correlation coefficients as the time-domain synchronization coefficient; performing frequency domain transformation on the vibration waveform data of each vibration damper, extracting the dominant frequency band with the highest vibration energy proportion of each vibration damper, and calculating the ratio of the intersection interval length of all dominant frequency bands in the purification vibration damper cluster to the average length of each dominant frequency band as the frequency-domain dominant frequency band overlap.

[0050] Time axis alignment refers to a signal preprocessing step designed to eliminate minor deviations in the starting point of vibration waveform data within the same purification vibration damper cluster, caused by independent clocks of individual vibration sensors or different data upload timings, ensuring all data are aligned under a unified time reference. Data acquisition delay deviation refers to the time asynchrony error caused by the aforementioned reasons. The time-domain synchronization coefficient is a quantitative parameter characterizing the consistency of the temporal variation of vibration waveforms of each vibration damper within the purification vibration damper cluster; a higher value indicates stronger synchronization. The frequency domain dominant band overlap is a quantitative parameter characterizing the similarity of the frequency ranges where vibration energy of each vibration damper within the cluster is mainly concentrated; a higher value indicates higher consistency in frequency distribution. The wind-induced vibration judgment threshold refers to a set of pre-set numerical critical points used to determine whether the calculated time-domain synchronization coefficient and the frequency domain dominant band overlap reach the typical characteristic level of wind-induced vibration. The cross-correlation coefficient is the basis for calculating the time-domain synchronization coefficient, a statistical index used to measure the similarity of two vibration waveform signals under different time delays; its maximum value reflects the similarity when the two waveforms are optimally matched. The mean here specifically refers to the arithmetic mean, which is the average level obtained by summing a set of cross-correlation coefficients and dividing by the number of data points in the set. Frequency domain transformation refers to the mathematical process of transforming the vibration waveform from the time domain to the frequency domain; in a preferred embodiment of this invention, this is achieved through a Fast Fourier Transform. The dominant frequency band refers to the continuous frequency range with the highest proportion of vibration energy determined after analyzing the vibration spectrum of a single vibration damper. The intersection interval length refers to the frequency width of the common portion where multiple dominant frequency bands overlap. The average length refers to the arithmetic mean of the frequency widths of these dominant frequency bands.

[0051] Specifically, the formula for calculating the time-domain synchronization coefficient is as follows:

[0052] in, This represents the time-domain synchronization coefficient. The closer the value is to 1, the more synchronized the vibrations of each point within the cluster are in time. This indicates the total number of vibration dampers in the purification vibration damper cluster. This indicates all different shock absorber pairs ( Perform a traversal and summation. Indicates the calculation of the first and the The cross-correlation coefficients between the waveforms of the vibration damper are calculated, and the maximum value is taken over all time delays τ.

[0053] Specifically, the formula for calculating the normalized cross-correlation coefficient between two waveforms is:

[0054] in, This represents the normalized cross-correlation coefficient under time delay τ, with a value range of [-1, 1]. , These represent the vibration acceleration time series after the two vibration dampers are aligned. , Representing sequences respectively and The mean. This represents the total number of sampling points within a fixed-duration analysis window, the length of which is reasonably set based on the dominant frequency of the vibration signal. τ represents the delay parameter, in [ The integer range [T / 2, T / 2] is taken to find the value that makes... The largest τ, the maximum value max( This indicates the similarity between two waveforms under optimal alignment.

[0055] Specifically, the formula for calculating the overlap of the dominant frequency band in the frequency domain is as follows:

[0056] in, This indicates the overlap of the dominant frequency bands in the frequency domain. The closer the value is to 1, the more consistent the main vibration frequency range of all vibration dampers, which is another characteristic of wind-induced vibration. The dominant frequency band of the k-th vibration damper is defined as the continuous frequency range with the highest energy in the spectrum of that vibration damper. , and These are the lower and upper limits of the frequency band, respectively. This represents the intersection of the dominant frequency bands of N vibration dampers, that is, the frequency range shared by all vibration dampers. Indicates frequency band Length, In determining wind-induced vibration, the calculated values ​​need to be considered. and Each with a preset threshold T t and T f Compare. If both conditions are met... >T t and >T f If so, the vibration is determined to be dominated by wind-induced vibration.

[0057] For example, suppose a vibration damper cluster corresponding to an abnormal power line contains three vibration dampers. The calculated normalized cross-correlation coefficients between any two pairs of vibration dampers after time-axis alignment are: 0.92 for damper 1 and damper 2, 0.88 for damper 1 and damper 3, and 0.90 for damper 2 and damper 3. Then, what is the time-domain synchronization coefficient of this cluster? =(0.92+0.88+0.90) / 3≈0.90. Meanwhile, spectral analysis revealed that the dominant vibration energy frequency bands of the three vibration dampers are [1.2Hz, 1.8Hz], [1.3Hz, 1.7Hz], and [1.25Hz, 1.75Hz], with their intersection at [1.3Hz, 1.7Hz], a length of 0.4Hz; the average length of each band is approximately 0.5Hz. Therefore, the overlap of the dominant frequency bands in the frequency domain is... =0.4 / 0.5=0.80. If the preset judgment threshold T... t =0.85, T f =0.70, because >T t and >T f If both conditions are met, the vibration of the abnormal line is determined to be dominated by wind-induced vibration.

[0058] The wind-induced vibration determination method described in the preferred embodiment of this invention calculates two quantitative indicators—the time-domain synchronization coefficient and the frequency-domain dominant frequency band overlap—of the vibration waveform of the purified anti-vibration hammer cluster, and compares them with preset thresholds. This enables intelligent and accurate differentiation of whether the vibration of abnormal lines is caused by general natural wind. This method effectively solves the problem of difficulty in identifying the root cause of vibration in traditional monitoring, avoids misjudging wind-induced vibration as line faults, thereby reducing unnecessary maintenance interventions. It allows maintenance resources to more accurately focus on real risks caused by non-wind factors, improving the targeting of status warnings and the economy of maintenance decisions.

[0059] like Figure 4 As shown, preferably, the risk pattern recognition strategy includes: extracting line-level feature parameters from the vibration waveform data of each purification vibration damper cluster corresponding to the abnormal line, wherein the line-level feature parameters include the dominant frequency of the vibration mode, the cross-cluster vibration phase consistency coefficient, and the vibration attenuation rate; comparing each line-level feature parameter with a preset normal state feature baseline, calculating the deviation degree and deviation duration of each parameter, and forming a multi-dimensional deviation dataset; based on the multi-dimensional deviation dataset, analyzing the deviation patterns of each line-level feature parameter, as well as the deviation degree combination relationship between different line-level feature parameters; and identifying the metal fatigue risk type corresponding to the abnormal line based on the deviation patterns and deviation degree combination relationships of each line-level feature parameter, combined with a preset metal fatigue risk judgment rule.

[0060] More preferably, the calculation process of the deviation pattern and the deviation degree combination relationship includes: for each line-level feature parameter in the multi-dimensional deviation dataset, extracting the sequence data of its deviation degree changing over time, and obtaining the changing trend parameter of the line-level feature parameter through trend fitting, as the deviation pattern; merging and calculating all deviation degrees to obtain a comprehensive combined deviation value, and numerically sorting the deviation degrees of each line-level feature parameter to determine the dominant and auxiliary parameters of the deviation degree; and using the comprehensive combined deviation value and the parameter primary and secondary sorting results as the deviation degree combination relationship.

[0061] Among them, line-level characteristic parameters refer to high-order characteristic parameters extracted from the vibration waveform data of one or more purification and vibration damper clusters corresponding to the abnormal line, used to characterize the overall structural dynamic state and vibration propagation characteristics of the line segment. Vibration mode dominant frequency refers to the dominant vibration frequency reflecting the inherent structural characteristics, identified through modal analysis or spectral clustering of the vibration response of the line segment. Cross-cluster vibration phase consistency coefficient is a quantitative parameter used to characterize the synchronicity and stability of the phase relationship of the wavefront or specific frequency components when vibration waves propagate between adjacent purification and vibration damper clusters; a higher value indicates more orderly vibration propagation. Normal state characteristic baseline refers to a reference model or threshold interval established through statistical analysis based on the line-level characteristic parameter data monitored under long-term healthy operation conditions, used to characterize its normal fluctuation range. Deviation degree refers to the numerical value of the quantified deviation obtained by comparing the real-time calculated line-level characteristic parameter value with the normal state characteristic baseline, such as relative deviation or standardized score. Deviation duration refers to the length of time that the deviation degree of the line-level characteristic parameter continuously exceeds a preset threshold. A multi-dimensional deviation dataset refers to a structured data set consisting of the deviation degree and duration of multiple line-level characteristic parameters over a continuous time period. Deviation patterns refer to the quantitative characteristics extracted from the deviation time series of a single line-level characteristic parameter, reflecting its changing patterns and trends. Deviation combination relationships refer to the numerical interaction, primary and secondary roles, and overall combined patterns of the deviation degrees of multiple different line-level characteristic parameters. Metal fatigue risk assessment rules refer to a set of logical judgment conditions based on the principles of materials mechanics, fracture mechanics, and historical fault data, used to infer whether conductors or fittings have accumulated metal fatigue risk based on the abnormal evolution patterns of line-level characteristic parameters. Trend parameters refer to parameters obtained by mathematically fitting the deviation time series, used to quantify its long-term change direction, such as the slope of the fitted straight line. Dominant deviation parameters refer to the line-level characteristic parameter with the largest deviation value within a specific time period. Auxiliary parameters refer to other line-level characteristic parameters with the next largest deviation values.

[0062] Specifically, the vibration decay rate is a key physical parameter for quantifying how quickly vibration energy dissipates in space or time. In an embodiment of the present invention, considering the vibration propagation characteristics of transmission lines, the vibration decay rate can be calculated as follows: select two adjacent clusters of anti-vibration hammers within the abnormal line range, and calculate their average vibration energy respectively. and Let L be the distance between the center points of the two clusters along the line. The vibration attenuation rate α can be obtained by fitting an exponential attenuation model. The larger the value, the faster the vibration decays along the path.

[0063] Specifically, the risk pattern recognition strategy achieves in-depth diagnosis of non-wind-induced vibration anomalies through the following technical means. First, line-level feature parameters are extracted: vibration data of each purification damper cluster involved in the abnormal line are analyzed collaboratively. For example, the dominant frequency of the vibration mode can be obtained by clustering the spectra of all clusters and taking the frequency band with the most concentrated energy as the line dominant frequency; the cross-cluster vibration phase consistency coefficient can be obtained by calculating the statistical characteristics, such as standard deviation or circular square deviation, of the phase difference sequence of vibration signals from adjacent clusters at a specific frequency, such as the line dominant frequency; the vibration decay rate can be obtained by analyzing the distribution of vibration energy, such as the effective value, along the line direction and fitting an exponential decay curve to obtain its decay constant. Next, a multi-dimensional deviation dataset is established: the real-time extracted line-level feature parameter values ​​are compared with a preset normal state feature baseline, the deviation of each parameter is calculated, such as (current value - baseline mean) / baseline standard deviation, and the duration of the deviation is recorded. Thus, a dataset containing multiple feature dimensions and evolving over time is formed.

[0064] Specifically, deviation The calculation formula is:

[0065] in, This indicates the value of a certain line-level characteristic parameter obtained from the current calculation, such as the dominant frequency of the vibration mode. This represents the historical average value of the parameter within the normal state characteristic baseline. This indicates that the participant The historical standard deviation of the number in the normal state characteristic baseline.

[0066] Specifically, the formula for calculating the trend parameter is as follows:

[0067] in, The slope of the fitted line is the trend parameter. Indicates the first A specific point in time. This represents the total number of data points involved in the fitting process.

[0068] The formula for calculating the overall portfolio deviation is:

[0069] in, This is the combined deviation value. for .

[0070] Specifically, the metal fatigue risk assessment rule is a pre-defined, structured set of logical judgment conditions. Its input consists of the deviation patterns and combinations of deviation degrees of various line-level characteristic parameters, and its output is a specific risk level or type. The execution of the metal fatigue risk assessment rule is based on the quantitative results calculated in the preceding steps. The rule specifically quantifies the "deviation pattern" as the changing trend parameters of each line-level characteristic parameter. The deviation degree combination relationship is specifically quantified into a comprehensive combination deviation value. The ranking results of the dominant and auxiliary parameters of the deviation. Each judgment condition of the rule is a threshold comparison and logical combination of these quantified results. The complete rules are as follows: Rule 1: If the trend parameter k1 of the vibration mode dominant frequency changes < If α1 is true and the duration of this trend is T1 > Tth1, then it is marked as a metal fatigue risk concern. Rule 2: Based on Rule 1, if the trend parameter k2 of the vibration decay rate is < If α2 is exceeded and the deviation of the cross-cluster vibration phase consistency coefficient increases significantly, it is upgraded to a high-risk level for potential metal fatigue. Rule 3: Based on satisfying Rule 2, if the comprehensive combined deviation value... If the deviation is greater than β, and the dominant parameter of the deviation is the dominant frequency of the vibration mode or the vibration decay rate, then it is ultimately determined to be a high-probability metal fatigue damage. Among them, the thresholds α1, α2, Tth1, and β are determined based on the historical baseline data of the line, simulation analysis, and experimental calibration.

[0071] Suppose that after analyzing a certain abnormal line, the deviation sequence of the dominant vibration mode frequency is fitted, and the trend parameter k1 = 0.02 / day, negative growth, and has continued for 25 days (T1=25); the trend parameter k2 of the vibration decay rate is... 0.015 / day; Phase consistency coefficient deviation increased to 1.5; Current overall combination deviation value =2.8, and the dominant parameter is the dominant frequency of the vibration mode. The system substitutes the above results into the metal fatigue risk judgment rules: First, it matches rule 1: k1<0, and T1>20 days, and the status is upgraded to "concern"; then it matches rule 2: k2<0, and the phase consistency deviation>1.0, and the status is upgraded to "potentially high risk"; finally, it matches rule 3: With a value greater than 2.5 and the dominant parameters being correct, the system ultimately determined that the risk type of the line was high-probability metal fatigue damage.

[0072] The risk pattern recognition strategy described in the preferred embodiment of this invention extracts line-level characteristic parameters such as the dominant frequency and phase consistency of vibration modes, and performs multi-dimensional, time-series deviation analysis with historical normal baselines. This method can reveal the evolutionary laws of vibration modes related to deep mechanisms such as metal fatigue that traditional monitoring cannot capture. Its core lies in upgrading simple "abnormal alarms" to "risk source diagnosis," which can not only identify whether there are non-wind-induced anomalies in line sections, but also provide early warning of potential metal fatigue risks through quantitative analysis of the deviation trends and combinations of multiple characteristic parameters. This provides a crucial technical decision-making basis for shifting transmission lines from "post-event maintenance" to "predictive maintenance."

[0073] In summary, the transmission line anomaly monitoring system and method based on vibration monitoring of anti-vibration hammers provided by this invention achieves real-time, automated data acquisition and transmission of transmission line vibration status through an intelligent sensor network deployed on the anti-vibration hammers and a self-powered unit on the tower side. Based on a progressive intelligent diagnostic process from individual components to clusters and from parts to the system, it can simultaneously and accurately locate and classify mechanical faults in individual anti-vibration hammers, intelligently distinguish whether abnormal vibrations are caused by general wind forces, and further identify potential deep-seated risks in the line section after excluding the influence of wind. This invention upgrades the operation and maintenance mode of transmission lines from inefficient, blind manual periodic inspections and simple threshold alarms to efficient, accurate intelligent online diagnosis and risk warning, realizing a systematic and interpretable assessment of the mechanical safety status of the line from micro-components to macro-sections, significantly improving the predictive maintenance capability and operational reliability of the power grid.

[0074] The above description is merely a preferred embodiment of the technical solution of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A transmission line anomaly monitoring system based on vibration monitoring with a vibration damper, characterized in that, include: Vibration sensors are installed on each anti-vibration hammer of the transmission line to collect vibration waveform data of the corresponding anti-vibration hammer. Each vibration sensor is preset with a unique position mark. A data collection and energy harvesting unit is installed on each transmission tower supporting the vibration damper. The data collection and energy harvesting unit includes a wireless communication unit and a current transformer. The wireless communication unit is used to communicate wirelessly with the vibration sensor to receive the vibration waveform data and location identifier sent by it. The current transformer is used to obtain electrical energy from the transmission line and to power itself and the vibration sensor. The intelligent monitoring terminal is communicatively connected to the data collection and energy harvesting unit, and is used to receive vibration waveform data, analyze and identify the operating status of the vibration damper and the line based on the vibration waveform data, and output the monitoring results of the transmission line.

2. A method for monitoring transmission line anomalies based on vibration monitoring with a vibration damper, applied to the transmission line anomaly monitoring system described in claim 1, characterized in that, include: Step S1: Obtain the vibration waveform data of each vibration damper and extract the waveform features corresponding to each vibration damper; Based on the preset spatial adjacency relationship, the vibration damper is divided into multiple vibration damper clusters; Step S2: For each vibration damper, compare and analyze its waveform characteristics with the waveform characteristics of other vibration dampers within its adjacent preset range to screen out abnormal vibration dampers. Step S3: Based on the waveform data of the abnormal vibration damper, determine the abnormal type of the abnormal vibration damper through the abnormal feature classification strategy, and form the first monitoring result by combining the location marker of the abnormal vibration damper. Step S4: For each vibration damper cluster, remove abnormal vibration dampers based on the first monitoring result to form a clean vibration damper cluster; based on all vibration waveform data in the clean vibration damper cluster, perform inter-cluster comparison through a difference analysis strategy to determine abnormal vibration damper clusters, and take the line range corresponding to the abnormal vibration damper cluster as the abnormal line. Step S5: For the abnormal line, determine whether it is wind-induced vibration based on the vibration waveform data of its purification anti-vibration hammer cluster; If the vibration is determined to be non-wind-induced, the risk type of the abnormal line is identified by the risk pattern recognition strategy based on the waveform characteristics of the purification vibration damper cluster, and a second monitoring result is formed by combining the section location information of the abnormal line. The results of the first and second monitoring are combined to form the transmission line inspection results.

3. The method for monitoring abnormalities in transmission lines according to claim 2, characterized in that, The abnormal feature classification strategy includes: From the waveform data of the abnormal vibration damper, extract a multi-dimensional feature vector containing time-domain features, frequency-domain features, and time-frequency-domain features; The feature values ​​in the multidimensional feature vector are compared with a preset multidimensional threshold condition set, which includes logical judgment conditions corresponding to different anomaly types, and each logical judgment condition includes multiple feature dimension thresholds. An anomaly type that satisfies all of its logical judgment conditions is identified as the anomaly type of the anomaly damper.

4. The method for monitoring abnormalities in transmission lines according to claim 2, characterized in that, The formation of the purification anti-vibration hammer cluster includes: locating the anti-vibration hammer cluster to which each abnormal anti-vibration hammer belongs based on the location identifier of each abnormal anti-vibration hammer in the first monitoring result; removing the vibration waveform data of the abnormal anti-vibration hammer from the dataset of the anti-vibration hammer cluster to which it belongs; and the remaining anti-vibration hammer and its vibration waveform data after removal constitute the purification anti-vibration hammer cluster.

5. The method for monitoring abnormalities in transmission lines according to claim 2, characterized in that, The difference analysis strategy includes: For each vibration damper cluster, the core feature parameters of the vibration waveform of the vibration damper are extracted based on the dataset of its corresponding purification vibration damper cluster. Statistical aggregation operations are performed on the core feature parameters to calculate feature statistics, which characterize the overall vibration state of the cluster. Compare the characteristic statistical values ​​of the current vibration damper cluster with the characteristic statistical values ​​of other vibration damper clusters within its adjacent preset range, and calculate their degree of difference. If the difference exceeds the preset cluster difference threshold, the current anti-vibration hammer cluster will be determined as an abnormal anti-vibration hammer cluster.

6. The method for monitoring abnormalities in transmission lines according to claim 2, characterized in that, The determination of wind-induced vibration includes: Time axis alignment is performed on the vibration waveform data of all vibration dampers in the purification vibration damper cluster corresponding to the abnormal line to reduce data acquisition delay and deviation. Based on the aligned vibration waveform data, calculate the time-domain synchronization coefficient and frequency domain dominance band overlap of the vibration waveforms of each vibration damper in the purification damper cluster. The time-domain synchronization coefficient and the frequency-domain dominant frequency band overlap are compared with the wind-induced vibration determination threshold to determine whether it is wind-induced vibration.

7. The method for monitoring abnormalities in transmission lines according to claim 6, characterized in that, The calculation of the overlap between the time-domain synchronization coefficient and the frequency-domain dominant frequency band includes: Calculate the cross-correlation coefficients between each pair of vibration waveforms of the vibration dampers in the purification damper cluster, and take the average of all cross-correlation coefficients as the time-domain synchronization coefficient. The vibration waveform data of each vibration damper is transformed in the frequency domain, and the dominant frequency band with the highest proportion of vibration energy of each vibration damper is extracted. The ratio of the intersection interval length of all dominant frequency bands in the purification vibration damper cluster to the average length of each dominant frequency band is calculated as the frequency domain dominant frequency band overlap.

8. The method for monitoring abnormalities in transmission lines according to claim 2, characterized in that, The risk pattern recognition strategy includes: Line-level characteristic parameters are extracted from the vibration waveform data of each purification anti-vibration hammer cluster corresponding to the abnormal line. The line-level characteristic parameters include the dominant frequency of the vibration mode, the cross-cluster vibration phase consistency coefficient, and the vibration attenuation rate. Each line-level feature parameter is compared with a preset normal state feature baseline, and the deviation degree and deviation duration of each parameter are calculated to form a multi-dimensional deviation dataset. Based on the multi-dimensional deviation dataset, the deviation patterns of each line-level feature parameter and the combination relationship of deviation degrees among different line-level feature parameters are analyzed. Based on the deviation patterns and combinations of deviation degrees of the characteristic parameters at each line level, and combined with the preset metal fatigue risk judgment rules, the metal fatigue risk type corresponding to the abnormal line is identified.

9. The method for monitoring abnormalities in transmission lines according to claim 8, characterized in that, The calculation process for the deviation pattern and the combination relationship of deviation degrees includes: For each line-level feature parameter in the multi-dimensional deviation dataset, extract the sequence data of its deviation degree changing over time, and obtain the changing trend parameter of the line-level feature parameter through trend fitting, as the deviation law; All deviations are integrated and calculated to obtain a comprehensive combined deviation value. The deviations of each line-level characteristic parameter are numerically sorted to determine the dominant and auxiliary parameters of the deviation. The comprehensive combined deviation value and the parameter ranking result are used as the deviation combination relationship.

10. The method for monitoring abnormalities in transmission lines according to claim 2, characterized in that, The comparison analysis includes: calculating the feature difference degree between the waveform features of the current vibration damper and the waveform features of other vibration dampers within an adjacent preset range, comparing the feature difference degree with a preset single-point anomaly threshold; when the feature difference degree exceeds the single-point anomaly threshold, the vibration damper is determined to be an abnormal vibration damper.

Images