Communication tower inclination attitude monitoring method and device based on inertial sensor

Abstract

The invention discloses a communication iron tower inclination attitude monitoring method and device based on an inertial sensor, and relates to the technical field of iron tower measurement, and the method comprises the steps: obtaining acceleration time sequence data; based on the acceleration time sequence data, determining a reference modal feature and instantaneous modal features at multiple moments, based on the instantaneous modal features and the reference modal feature, determining an equivalent icing quality index and a modal evolution cost matrix, and through an optimal allocation algorithm, performing calculation processing on the modal evolution cost matrix to obtain an optimal modal evolution path; on the basis of the optimal modal evolution path and the modal evolution cost matrix, a modal change concentration index is obtained through calculation; and based on the equivalent icing quality index and the modal change concentration index, carrying out risk judgment processing on the structural state of the communication iron tower so as to monitor the inclination attitude of the communication iron tower. According to the invention, the technical effect of timely early warning of dynamic instability of the iron tower is achieved.

Description

Technical Field

[0001] This application relates to the field of tower measurement technology, specifically to a method and device for monitoring the tilt attitude of communication towers based on inertial sensors. Background Technology

[0002] Communication towers, as the fundamental infrastructure for wireless communication networks, are often located in harsh outdoor environments. During cold, windy seasons, towers face severe icing threats. Icing not only increases the static load on the tower, but more dangerously, asymmetrical icing distribution alters the aerodynamic cross-sectional shape of tower components (such as angle steel and guy wires). According to aerodynamic theory, when wind flows through such irregular cross-sections, it can generate negative aerodynamic damping. Once this negative aerodynamic damping cancels out the structure's own mechanical damping, it triggers a low-frequency, high-amplitude self-excited vibration, known as galloping. Galloping can easily lead to structural fatigue damage and even collapse.

[0003] Existing monitoring methods based on inertial sensors typically trigger alarms by monitoring acceleration amplitude or tilt angle thresholds. However, this "phenomenon-based" approach has serious technical limitations: it cannot effectively distinguish between forced vibrations (fluttering) directly caused by strong wind pulsations and self-excited vibrations (galloping) caused by structural instability. While fluttering may have higher amplitudes, it usually does not exhibit a destructive divergence trend, whereas galloping is a precursor to structural collapse. Due to the lack of ability to identify the vibration generation mechanism, existing technologies face a dilemma: raising the alarm threshold may result in missing weak signals in the early stages of galloping; lowering the threshold will lead to frequent false alarms due to fluttering in strong winds, resulting in low maintenance efficiency.

[0004] Therefore, there is an urgent need for a monitoring method that can delve into the structural evolution mechanism and effectively distinguish between buffeting and galloping, so as to provide timely early warning of the dynamic instability of the tower. Summary of the Invention

[0005] To address the technical problem that related technologies typically rely solely on monitoring acceleration amplitude or tilt angle thresholds for alarms, lacking the ability to identify the vibration generation mechanism and thus failing to provide timely early warnings of dynamic instability of towers, this application provides a method and device for monitoring the tilt attitude of communication towers based on inertial sensors.

[0006] The specific technical solution adopted is as follows: Acquire acceleration time-series data collected by sensors on the communication tower within a preset time period; Based on acceleration time-series data, the reference modal characteristics and instantaneous modal characteristics at multiple time points are determined. The reference modal characteristics include the set of reference modal frequencies and the set of reference modal damping ratios. Based on instantaneous modal characteristics and baseline modal characteristics, the equivalent icing quality index and modal evolution cost matrix are determined. The optimal modal evolution path is obtained by calculating and processing the modal evolution cost matrix through the optimal allocation algorithm. Based on the optimal modal evolution path and the modal evolution cost matrix, the modal change concentration index is calculated. Based on the equivalent icing quality index and modal change concentration index, the structural status of communication towers is assessed for risk, in order to monitor the tilt attitude of the communication towers.

[0007] In one possible implementation of this application, determining reference modal characteristics and instantaneous modal characteristics at multiple time points based on acceleration time-series data includes: The acceleration time series data is processed by a preset recognition algorithm to obtain the modal frequency and modal damping ratio. The data set corresponding to the preset number of modal frequencies and modal damping ratios before sorting, the frequency variance vector and the damping ratio variance vector corresponding to the data set, are used as the reference modal features; The acceleration time series data within the current analysis period are processed for parameter identification to obtain instantaneous modal characteristics at multiple moments. The duration of the current analysis period is shorter than the duration of the preset time period.

[0008] In one possible implementation of this application, the equivalent icing quality index and the modal evolution cost matrix are determined based on instantaneous modal characteristics and baseline modal characteristics, including: Extract the frequency variance vector and damping ratio variance vector from the baseline modal features; Based on the frequency variance vector and the damping ratio variance vector, the frequency change weight and the damping change weight are calculated. Based on frequency change weights, damping change weights, instantaneous modal features, and baseline modal features, a modal evolution cost matrix is ​​constructed. The equivalent icing quality index is calculated based on the ratio between the instantaneous modal characteristics and the baseline modal characteristics.

[0009] In one possible implementation of this application, after determining the equivalent icing quality index and the modal evolution cost matrix based on instantaneous modal features and baseline modal features, the method further includes: When the equivalent icing mass index meets the preset conditions, the acceleration time series data in the current analysis period are processed by the preset identification algorithm to obtain the candidate reference mode characteristics of the current analysis period. By using a preset health metric function, the baseline health metric of the baseline modal features and the candidate health metric of the candidate baseline modal features are calculated. When the candidate health metric is greater than the baseline health metric, the candidate baseline modal feature replaces the baseline modal feature.

[0010] In one possible implementation of this application, a modal change concentration index is calculated based on the optimal modal evolution path and the modal evolution cost matrix, including: Calculate the change vector of each mode based on the optimal mode evolution path and the mode evolution cost matrix; When the change vector satisfies the preset boundary conditions, the first concentration index is calculated based on the change vector. When the change vector does not meet the preset boundary conditions, the second concentration index is calculated based on the change vector. The first or second concentration index is used as the concentration index for modal change.

[0011] In one possible implementation of this application, when the change vector does not satisfy the preset boundary conditions, a second concentration index is calculated based on the change vector, including: Calculate the sum of the change vectors of each modal data to obtain the total change; When the total change does not meet the preset boundary conditions, the change vector is normalized to obtain the probability distribution vector. Calculate the Shannon information entropy of the probability distribution vector and use the Shannon information entropy as the second concentration index.

[0012] In one possible implementation of this application, risk assessment of the structural state of communication towers is performed based on the equivalent icing quality index and the modal variation concentration index, including: When the equivalent icing quality index is less than or equal to the first preset threshold, the structural state of the communication tower is determined to be ic-free. When the equivalent icing quality index is greater than the second preset threshold and the modal change concentration index is greater than or equal to the third preset threshold, the structural state of the communication tower is determined to be uniform icing. When the equivalent icing quality index is greater than the second preset threshold and the modal change concentration index is less than the third preset threshold, the structural state of the communication tower is determined to be asymmetric icing state. When a communication tower is in an asymmetric icing state and galloping self-excited vibration is detected, the structural state of the communication tower is determined to be in a state of dynamic instability.

[0013] In one possible implementation of this application, when the communication tower is in an asymmetric icing state and galloping self-excited vibration is detected, the communication tower is determined to be in a dynamic instability state, including: When the communication tower is in an asymmetric icing state, obtain the vibration energy and vibration frequency within the current analysis period; When the vibration energy is greater than the preset high energy threshold and the deviation between the vibration main frequency and the target instantaneous modal frequency is less than the preset locking tolerance, it is determined that the triggered galloping self-excited vibration has been detected and the communication tower is in a state of dynamic instability.

[0014] In one possible implementation of this application, the method for determining the target instantaneous modal frequency includes: Extract the maximum value of the vector from the change vectors of each mode; Based on the optimal modal evolution path, determine the instantaneous mode corresponding to the maximum value of the vector; The frequency corresponding to the instantaneous mode is taken as the target instantaneous mode frequency.

[0015] To achieve the above objectives, a communication tower tilt attitude monitoring device based on an inertial sensor is also provided. The device includes: The acquisition module is used to acquire acceleration time-series data collected by sensors on the communication tower within a preset time period; The determination module is used to determine the reference modal characteristics and the instantaneous modal characteristics at multiple times based on acceleration time series data. The reference modal characteristics include the set of reference modal frequencies and the set of reference modal damping ratios. The processing module is used to determine the equivalent icing quality index and the modal evolution cost matrix based on instantaneous modal features and baseline modal features. The modal evolution cost matrix is ​​calculated and processed through the optimal allocation algorithm to obtain the optimal modal evolution path. The calculation module is used to calculate the mode change concentration index based on the optimal mode evolution path and the mode evolution cost matrix. The risk assessment module is used to assess the structural status of communication towers based on the equivalent icing quality index and modal change concentration index, so as to monitor the tilt attitude of the communication towers.

[0016] This application has, but is not limited to, the following technical effects: By acquiring acceleration time-series data collected by sensors on the communication tower within a preset time period, baseline modal characteristics and instantaneous modal characteristics at multiple time points are determined based on the acceleration time-series data. Based on the instantaneous modal characteristics and baseline modal characteristics, the equivalent icing quality index and modal evolution cost matrix are determined. Then, an optimal matching algorithm is used to calculate and process the modal evolution cost matrix to obtain the optimal modal evolution path. Using the optimal modal evolution path and the modal evolution cost matrix, the modal change concentration index is calculated. Finally, the equivalent icing quality index and the modal change concentration index are used to assess the structural state of the communication tower for risk assessment. In this application, to monitor the tilt attitude of communication towers, the evolution path of each modal characteristic from the baseline state is tracked and quantified. The basis for risk assessment is shifted from the vague vibration phenomenon itself to the quantifiable deterioration of specific structural modal characteristics as an internal cause of galloping. The modal change concentration index is calculated by using the optimal modal evolution path and the modal evolution cost matrix. Then, the structural state of the communication tower is risk-determined by using the equivalent icing quality index and the modal change concentration index. The external vibration phenomenon is correlated with the evolution process of the internal structural characteristics, thereby effectively distinguishing between fluttering and galloping and providing timely early warning of the dynamic instability of the communication tower. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the first embodiment of the communication tower tilt attitude monitoring method based on inertial sensors according to this application; Figure 2 This is a schematic diagram of the overall processing flow involved in the inertial sensor-based communication tower tilt attitude monitoring method of this application; Figure 3 This is a schematic diagram of the device structure of the hardware operating environment involved in the embodiments of this application. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0019] This application provides a method for monitoring the tilt attitude of a communication tower based on an inertial sensor. In the first embodiment of this method, referring to... Figure 1 ,include: Step S10: Obtain acceleration time-series data collected by sensors on the communication tower within a preset time period.

[0020] As an example, the inertial sensor-based communication tower tilt attitude monitoring method can be applied to an inertial sensor-based communication tower tilt attitude monitoring device. This device belongs to an inertial sensor-based communication tower tilt attitude monitoring system. A schematic diagram of the overall processing flow of this system is shown below. Figure 2 As shown, the specific modules and processing steps are as follows: Module 1: Benchmark Model Management Module; Functional attributes: The function of this module is to establish and continuously optimize the analytical benchmark representing the ice-free state of the tower. It runs in parallel throughout the system lifecycle to ensure the long-term accuracy and adaptability of the benchmark.

[0021] Internal processing: Includes steps S1.1 (initial calibration) and S1.2 (continuous optimization). S1.1 is executed when the system is first started to generate an initial baseline. In subsequent operation, S1.2 is executed periodically to update the baseline when the no-icing condition is met.

[0022] Input: Acceleration time series data, ambient temperature data.

[0023] Output: Writes or updates baseline modal features (set of baseline modal frequencies) to or updates the "Baseline Modal Feature Library". Reference modal damping ratio set ) and its variance , ).

[0024] Module 2: Instantaneous Feature Recognition Module; Functional attributes: This module is the starting point of the main analysis process and is responsible for extracting dynamic features that represent the current state of the structure from the data of the current analysis cycle.

[0025] Internal processing: Execute step S2.1.

[0026] Input: Acceleration time series data for the current analysis period.

[0027] Output: Instantaneous modal features ( , ), which is an instantaneous variable and is only valid within the current period.

[0028] Module 3: Modal Evolution Analysis Module; Functional attributes: Through a computational process deeply tied to the physical mechanism of galloping, the evolution path of modes is tracked and the degree of risk concentration is quantified.

[0029] Internal processing includes steps S2.2 (constructing the cost matrix and tracing the optimal evolution path) and S2.3 (calculating modal changes and quantifying concentration indices).

[0030] Input: Instantaneous modal features (set of instantaneous modal frequencies) provided by the "Instantaneous Feature Recognition Module" Instantaneous modal damping ratio set ).

[0031] Global baseline variables read from the "Benchmark Modal Feature Library" , , ).

[0032] Output: Modal variation concentration index .

[0033] Target lock frequency : That is, the instantaneous frequency corresponding to the mode that changes most drastically, which is tracked by the optimal evolution path.

[0034] Module Four: Risk Level Determination Module; Functional attributes: This module is the decision-making center of the system. It integrates all analytical indicators and outputs the final risk level based on a finite structural state that corresponds to the structural evolution process.

[0035] Internal processing: Execute steps S3.1 (structural state definition) and S3.2 (state transition logic determination).

[0036] Input: Provided by the "Modal Evolution Analysis Module" and target lock frequency Vibration energy calculated in real time based on current acceleration data and vibration dominant frequency Equivalent icing quality index calculated based on baseline modal characteristics .

[0037] Output: A structured status report.

[0038] As an example, the sensor could be an inertial sensor. During data acquisition, upon the device's initial power-on, one or more of its built-in inertial sensors (e.g., mounted at two-thirds of the tower's height) immediately begin collecting data for an initial fixed duration. (For example, 10 minutes, or a preset time period) of three-axis acceleration time series data.

[0039] Step S20: Based on acceleration time series data, determine the reference modal characteristics and the instantaneous modal characteristics at multiple times. The reference modal characteristics include the reference modal frequency set and the reference modal damping ratio set.

[0040] As an example, by performing multimodal parameter identification processing on the collected acceleration time series data, the acceleration is decomposed into modal frequency and modal damping ratio. The reference modal features serve as the reference data for global analysis, while the instantaneous modal features are modal feature data collected in real time. The two are obtained in the same way, the difference being that they are extracted from data from different time periods.

[0041] As an example, the reference modal frequency set is a set of reference modal frequencies of different orders, and the reference modal damping ratio set is the same.

[0042] Step S20 includes: The acceleration time series data is processed by a preset recognition algorithm to obtain the modal frequency and modal damping ratio.

[0043] As an example, a modal parameter identification algorithm is applied to the acceleration time series data. For example, the preset identification algorithm is frequency domain decomposition (FDD), which is a well-known technique in the field. It identifies the modal parameters of the structure by calculating the cross power spectral density matrix of the acceleration signal and performing singular value decomposition. Alternatively, other modal analysis methods well known to those skilled in the art, such as stochastic subspace identification (SSI), can also be used. By using this algorithm to perform parameter identification processing on the acceleration time series data, the modal frequencies and modal damping ratios can be obtained.

[0044] The data set corresponding to the preset number of modal frequencies and modal damping ratios before sorting, the frequency variance vector and the damping ratio variance vector corresponding to the data set, are used as the reference modal features.

[0045] As an example, extracting the preceding modalities from the identified modalities , N is the preset quantity (for example, The device identifies modes with significant energy and stores their corresponding modal frequencies and modal damping ratios as initial reference modal characteristics in the device's non-volatile memory, along with the reference vibration energy. , The root mean square (RMS) value of the acceleration signal in the un-iced state.

[0046] As an example, an initial set of reference modal frequencies is recorded as follows: ,in, For the first The frequencies of the first reference modes; an initial set of reference mode damping ratios is recorded as follows: ,in For the first Damping ratio of the first reference mode.

[0047] Specifically, since the initial calibration is based on only a single data window and cannot calculate a statistically significant variance, in order to ensure the numerical stability of the weight calculation in subsequent steps and avoid division by zero errors, the variance of the baseline modal characteristics is set to a set of preset, relatively large conservative values ​​and stored in non-volatile memory.

[0048] An initial reference mode frequency variance vector is recorded as follows: .

[0049] The initial damping ratio variance vector of a reference mode is recorded as follows: .

[0050] After completing the above steps, a complete set of global baseline variables is obtained and can be used in subsequent steps. The initial calibration phase ends immediately and the system switches to real-time monitoring mode.

[0051] The acceleration time series data within the current analysis period are processed for parameter identification to obtain instantaneous modal characteristics at multiple moments. The duration of the current analysis period is shorter than the duration of the preset time period.

[0052] As an example, the current analysis period could be a 10-minute sliding time window, with T as the duration. step (It can be 1 minute) for step-by-step updates.

[0053] No specific limitations are imposed, but the duration of the current analysis period is less than the duration of the preset time period.

[0054] As an example, a preset recognition algorithm is used to perform parameter recognition processing on the acceleration time series data within the current analysis period, and the preceding parameters are extracted from the recognition results. indivual( Can be equal to or not equal to The energy-significant modes are identified, and two classes of instantaneous variable / instantaneous modal features are generated that are only valid within the current analysis period. One set of instantaneous modal frequencies is recorded as follows: ,in, For the first The observed current modal frequencies; a set of instantaneous modal damping ratios is recorded as ,in The first value was obtained by calculating the identified single-degree-of-freedom spectral peak using the half-power bandwidth method. The observed current modal damping ratio.

[0055] Step S30: Based on the instantaneous modal characteristics and the baseline modal characteristics, determine the equivalent icing quality index and the modal evolution cost matrix. Then, calculate and process the modal evolution cost matrix using the optimal allocation algorithm to obtain the optimal modal evolution path.

[0056] As an example, the equivalent icing quality index can be a parameter index that reflects the overall icing severity of the tower, calculated based on the decrease in modal frequency. The modal evolution cost matrix includes a basic matching matrix. The elements in the modal evolution cost matrix quantify the dissimilarity between modes and nonlinearly amplify the decrease in damping ratio. The basic matching matrix is ​​used to lock which instantaneous mode corresponds to which reference mode. The basic matching matrix is ​​used as the input of the optimal allocation algorithm.

[0057] As an example, after constructing the modal evolution cost matrix, the completed cost matrix will be... (A transient variable valid only within this step) is used as input to the optimal assignment algorithm. A combinatorial optimization algorithm capable of solving the assignment problem is applied to find a mapping from the baseline mode set to the transient mode set that minimizes the sum of the total costs of the matched pairs. The optimal assignment algorithm can be the Jonker-Volgenant algorithm, whose output is the optimal mode evolution path. ,in Indicates the first The reference mode evolves into the first state in the current state. A transient mode.

[0058] Step S30 includes: Extract the frequency variance vector and damping ratio variance vector from the baseline modal features.

[0059] Based on the frequency variance vector and the damping ratio variance vector, the frequency change weight and the damping change weight are calculated.

[0060] As an example, the modal evolution cost matrix has the following dimensions: For the first One reference mode ( From 1 to ) and its matching number A number of instantaneous modes ( From 1 to ), its corresponding cost element Used for quantization from the reference mode Evolving to instantaneous modes The "cost" is determined by extracting the frequency variance vector and damping ratio variance vector from the reference modal features after storing the reference modal features in non-volatile memory. Based on these two types of parameters, the frequency change weight and damping change weight are calculated. This weight ensures that the more stable the modal parameter is in the non-icing state, the greater its contribution to the total cost value when it changes.

[0061] As an example, taking the i-th reference mode as an example, the frequency change weights The calculation method can be: in, Indicates the first The frequency variance vector of each reference mode For example, a pre-defined, minimal positively stable term to prevent division by zero errors. .

[0062] As an example, taking the i-th reference mode as an example, the damping change weight The calculation method can be: in, Indicates the first The damping ratio variance vector of each reference mode.

[0063] Based on the frequency change weight, damping change weight, instantaneous modal features, and baseline modal features, a modal evolution cost matrix is ​​constructed.

[0064] As an example, the basic matching matrix is ​​constructed based on the frequency-based Mahalanobis distance.

[0065] formula: Execute: Input the optimal allocation algorithm (such as the Hungarian algorithm / KM algorithm) and output the optimal matching path (i.e., determine which instantaneous mode j corresponds to the reference mode i).

[0066] As an example, after determining the correspondence between the baseline mode i and the instantaneous mode j, constructing the final mode evolution cost matrix first requires calculating the value of each cost element in the matrix, starting with the i-th element. The first reference mode and the second The cost element corresponding to each instantaneous mode For example, cost element The calculation method can be: In the baseline update phase, the average frequency weights of all modes are calculated. and average damping weight ,set up , It is a preset positive gain coefficient used to amplify the effect of damping drop, for example, , This indicates the statistical significance of the damping change, even if the damping variance is large (leading to weight). (Smaller), this item also reflects the basic fluctuation situation, when the instantaneous damping ratio is relatively small. Less than the reference damping ratio When the parameter of the exponential term is positive, its value increases exponentially, thus significantly increasing the total scalar value. exp is the natural constant. This represents the reference mode frequency in the i-th reference mode feature. This represents the instantaneous modal frequency in the j-th instantaneous modal feature; after determining each cost element in the matrix, the modal evolution cost matrix can be constructed.

[0067] The equivalent icing quality index is calculated based on the ratio between the instantaneous modal characteristics and the baseline modal characteristics.

[0068] As an example, the equivalent icing quality index The calculation method can be: Where N: the total number of modal orders involved in the calculation (i.e., the number of calibrated reference modes), i: the index variable of the modal order, ranging from 1 to N, f base,i : The i-th reference mode frequency, representing the structural frequency in the un-iced state, f obs,i The instantaneous modal frequency of the i-th order is the frequency of f. base,i The corresponding instantaneous frequency.

[0069] As an example, the instantaneous modal frequencies of each order are checked before calculation. .like <γ (For example, if γ=0.5, that is, the frequency drops by more than 50%), it is judged as an anomaly or a serious fault. In this case, the mode is removed and does not participate in the exponent calculation; or its frequency is regarded as unchanged to prevent a single outlier from causing the overall exponent to be overwhelmed.

[0070] After step S30, the following steps are also included: When the equivalent icing mass index meets the preset conditions, the acceleration time series data in the current analysis period are processed by the preset identification algorithm to obtain the candidate reference mode characteristics of the current analysis period.

[0071] As an example, this embodiment iteratively updates the baseline modal characteristics within the current analysis period. The preset condition can be an optimization trigger condition. Within the current analysis period, the system first determines whether the optimization trigger condition is met. This condition is a composite logical judgment that must be satisfied simultaneously: Condition 1: The average temperature of the current cycle is higher than a preset temperature threshold (e.g., 5°C) to preliminarily determine that there is no icing environment.

[0072] Condition 2: Based on the currently stored set of reference mode frequencies Calculated equivalent icing quality index The value is less than a preset low threshold (e.g., 0.02) to confirm that the current tower body is basically free of ice.

[0073] If the above optimization triggering conditions are not met, this step will not perform any operation in this cycle. If the optimization triggering conditions are met, the system will consider the current analysis cycle to be a potentially better ice-free state. In this case, the system will execute a modal parameter identification algorithm on the acceleration data of the current analysis cycle to obtain a set of candidate baseline modal features (candidate modal frequencies) for the current analysis cycle. Candidate modal damping ratio ).

[0074] By pre-setting a health metric function, the baseline health metric of the baseline modal features and the candidate health metric of the candidate baseline modal features are calculated.

[0075] As an example, a health metric can be calculated by comparing candidate baseline modality features with currently stored baseline modality features using a pre-defined health metric function. ,in For the first First-order modal frequency, The weights are preset, which can be 1 / N. A higher health metric typically indicates that the structure is closer to an ice-free state. The system calculates the baseline health metric of the currently stored baseline modal features. Candidate health metrics and candidate baseline modal features .

[0076] When the candidate health metric is greater than the baseline health metric, the candidate baseline modal feature replaces the baseline modal feature.

[0077] As an example, if The system performs a baseline update operation, which uses the candidate... and Overwrite the original non-volatile memory and Meanwhile, to obtain a more accurate variance, the system can recalculate the statistical variance based on all data windows that meet the optimization triggering conditions within a recent period (e.g., the past 24 hours), and use it to overwrite the frequency variance vector in memory. And damping ratio variance vector .

[0078] Step S40: Based on the optimal modal evolution path and the modal evolution cost matrix, the modal change concentration index is calculated.

[0079] As an example, those skilled in the art often regard the high amplitude vibration of iron towers as a risk in general. The essence of the technical problem lies in the failure to delve into the mechanism from the phenomenon level, that is, the inability to distinguish between forced vibration directly caused by external wind loads and self-excited vibration caused by the deterioration of the structure's own characteristics (especially the decrease in damping). Therefore, the design principle of the embodiments of this application is not to track the vibration amplitude, but to directly quantify the evolution process of the structural modal characteristics. By constructing a cost function that simulates the self-excited mechanism of galloping aerodynamics, and combining it with information entropy theory, an index that can characterize whether the changes in modal characteristics are dangerously "concentrated" on a few modes is finally obtained, thereby fundamentally identifying the structural preconditions for self-excited vibration (galloping).

[0080] Furthermore, since the destructive nature of galloping is usually caused by the loss of control of one or a few key modes, obtaining only the changes in each mode is not enough to make a risk assessment. It is also necessary to further quantify whether these changes are "concentrated" in a few modes. Shannon entropy is an ideal tool for measuring the degree of concentration of a distribution. A low entropy value directly corresponds to a highly concentrated distribution.

[0081] Therefore, based on the optimal modal evolution path and the modal evolution cost matrix, the modal change concentration index is calculated.

[0082] Step S40 includes: Based on the optimal modal evolution path and the modal evolution cost matrix, calculate the change vector of each mode.

[0083] As an example, the dimension of this change vector is... Each of its elements That is, the first The cost of a baseline mode along its optimal evolution path directly quantifies the drastic change of that mode from the baseline state to the current state. , This represents the optimal modal evolution path.

[0084] When the change vector satisfies the preset boundary conditions, the first concentration index is calculated based on the change vector.

[0085] As an example, calculate the total change. The total change in the change vector of each mode is compared with a preset parameter threshold. If the total change is less than the preset parameter threshold, the change vector is determined to satisfy the preset boundary condition, indicating that the current structure has no significant change compared to the baseline state. The preset parameter threshold can be a very small positive threshold (e.g., ...). ).

[0086] As an example, the first concentration index is used as the modal change concentration index in this case. The calculation method can be: Where N represents the modal order.

[0087] When the change vector does not meet the preset boundary conditions, the second concentration index is calculated based on the change vector.

[0088] As an example, when the change vector does not meet the preset boundary conditions, the final risk quantification index, also known as the second concentration index, is obtained by calculating its normalized Shannon information entropy based on the change vector.

[0089] The step of calculating the second concentration index based on the change vector when the change vector does not meet the preset boundary conditions includes: The total change is obtained by summing the change vectors of each modality data.

[0090] When the total change does not meet the preset boundary conditions, the change vector is normalized to obtain the probability distribution vector.

[0091] Calculate the Shannon information entropy of the probability distribution vector and use the Shannon information entropy as the second concentration index.

[0092] As an example, if the above-mentioned preset boundary conditions are not met, then the change vector will be... After normalization, a probability distribution vector is obtained. Its elements are ,in, This represents the total change.

[0093] Then calculate the Shannon information entropy of this probability distribution to obtain the second concentration index / modal change concentration index. : Among them, it is agreed that when hour, The value of the item is 0, which is close to of The value indicates that the total change is highly concentrated in one or a few modes, indicating a high-risk state of galloping caused by asymmetric icing.

[0094] The first or second concentration index is used as the concentration index for modal change.

[0095] As an example, in different implementation scenarios, the corresponding first concentration index or second concentration index is used as the modal change concentration index, where the modal change concentration index is a dimensionless scalar.

[0096] Step S50: Based on the equivalent icing quality index and the modal change concentration index, the structural status of the communication tower is assessed for risk, so as to monitor the tilt attitude of the communication tower.

[0097] As an example, the tilting attitude of the communication tower monitored in this application includes monitoring of static tilt caused by static loads, as well as monitoring of the risk of "dynamic attitude instability," of which galloping is the most severe manifestation. Moreover, the most fundamental difficulty in this application lies in the fact that high-amplitude vibration phenomena are inherently ambiguous, making it impossible to directly determine their benign or malignant nature. Therefore, the core idea of ​​this application is to utilize a finite state machine that corresponds one-to-one with the structural evolution process, combining indicators that reveal changes in the structure's intrinsic characteristics with directly observable vibration phenomena, thereby making a clear and evidence-based judgment on galloping risk and achieving effective early warning of dynamic attitude instability.

[0098] Step S50 includes steps S51 to S54: Step S51: When the equivalent icing quality index is less than or equal to the first preset threshold, the structural state of the communication tower is determined to be ic-free.

[0099] As an example, the first preset threshold could be a preset low threshold. .For example, A value of 0.02 is acceptable. The absence of icing indicates that the tower is not covered with ice or has very little icing, and its dynamic characteristics are within the calibrated reference range.

[0100] Step S52: When the equivalent icing quality index is greater than the second preset threshold and the modal change concentration index is greater than or equal to the third preset threshold, the structural state of the communication tower is determined to be uniform icing.

[0101] As an example, a uniform icing state indicates that the tower has significant icing, but the icing distribution is relatively uniform and has not caused significant asymmetric changes in the structural dynamics. The main risk at this time is an increase in static load.

[0102] As an example, the second preset threshold could be a preset high threshold. (For example, 0.05), the third preset threshold can be a preset high-entropy threshold. (For example, The higher the modal variation concentration index, the more evenly the modal variation is distributed across all modes, and the better the structural symmetry is maintained.

[0103] Step S53: When the equivalent icing quality index is greater than the second preset threshold and the modal change concentration index is less than the third preset threshold, the structural state of the communication tower is determined to be asymmetric icing state.

[0104] As an example, asymmetric icing indicates that the tower is severely and unevenly iced, resulting in a high concentration of modal variations in a few modes. This corresponds to the structural precondition for galloping, indicating that the tower is in the "awaiting" stage of dynamic instability. A low modal variation concentration index is a key internal factor in determining asymmetry.

[0105] Step S54: When the communication tower is in an asymmetric icing state and a self-excited vibration triggered by galloping is detected, the structural state of the communication tower is determined to be in a dynamic unstable state.

[0106] As an example, the dynamic instability state indicates that, given an already asymmetric icing structure, external wind excitation has successfully triggered galloping self-excited vibration. The structure is currently in a dynamic attitude instability process. The prerequisite for determining that a communication tower has entered this state is that the system's state in the previous analysis period was asymmetric icing.

[0107] Step S54 includes: When the communication tower is in an asymmetric icing state, the vibration energy and vibration frequency within the current analysis period are obtained.

[0108] As an example, a status determination is performed once at the end of each analysis cycle. To perform this determination, in addition to the equivalent icing quality index... and modal variation concentration index In addition, the following two indicators need to be calculated in real time: vibration energy. This can be done by bandpass filtering (before bandwidth coverage) on the acceleration time series data of the current period. The frequency range of the first modal is determined, and then the root mean square value of the filtered signal is calculated. The dominant vibration frequency is... This could involve performing power spectral density analysis on the bandpass filtered signal to extract the frequency corresponding to the maximum energy peak.

[0109] When the vibration energy is greater than the preset high energy threshold and the deviation between the vibration main frequency and the target instantaneous modal frequency is less than the preset locking tolerance, it is determined that the triggered galloping self-excited vibration has been detected and the communication tower is in a state of dynamic instability.

[0110] As an example, a preset high energy threshold could be a reference vibrational energy. The value is 10 times the value of the tower. When two conditions are met, the communication tower is determined to be in a state of dynamic instability.

[0111] Condition 1: High-energy vibration: Greater than a preset high energy threshold This threshold can be defined as a multiple of the average vibration energy under the non-icing state calibrated in step S1, for example, 10 times.

[0112] Condition 2: Precise frequency locking: Vibration dominant frequency Locking is performed near the frequency of the tracked mode that changes most drastically, that is, near the target instantaneous modal frequency. The determination method is that the relative deviation between the vibration dominant frequency and the target instantaneous modal frequency is less than the preset locking tolerance, which can be 5%.

[0113] The methods for determining the instantaneous modal frequency of the target include: Extract the maximum value of the vector from the change vectors of each mode, and determine the instantaneous mode corresponding to the maximum value of the vector based on the optimal mode evolution path.

[0114] The frequency corresponding to the instantaneous mode is taken as the target instantaneous mode frequency.

[0115] Specifically, firstly, from the vector of modal changes of each order... Find the index corresponding to the maximum value in the table: Then, the optimal modal evolution path is used. Find the instantaneous mode corresponding to the maximum value of the vector: The frequency corresponding to the instantaneous mode is taken as the target instantaneous mode frequency. Finally, the dominant vibration frequency is checked. With the target instantaneous modal frequency Is the relative deviation less than a preset locking tolerance? (For example, 5%), the calculation method is as follows: When this condition is met, the state is determined to be dynamically unstable. If the current state is dynamically unstable, but the vibrational energy... It has dropped to the preset high energy threshold. The state then transitions from dynamic instability to asymmetric icing, indicating that galloping has stopped, but the structure remains in a dangerous state where galloping can be easily triggered again.

[0116] As an example, let S0 be the state without ice, S1 be the state with uniform ice, S2 be the state with asymmetric ice, and S3 be the state with galloping.

[0117] The transition rules for various structural states are as follows: From any state to S0: If the equivalent icing mass index First preset threshold The new states all transition to S0.

[0118] From S0 to S1: If the equivalent icing quality index Second preset threshold And the modal variation concentration index Third preset threshold The state transitions from S0 to S1.

[0119] From S0 or S1 to S2: If and The state transitions to S2. From S2 to S1: If the current state is S2, and... Rise and greater than The state transitions from S2 to S1.

[0120] From S2 to S3 (core judgment): If the current state is S2 and the following two galloping phenomenon criteria are met at the same time, the state will transition from S2 to S3.

[0121] Criterion 1: High-energy vibration Greater than a preset high energy threshold .

[0122] Criterion 2: Precise frequency locking: dominant vibration frequency It is locked near the frequency of the mode that changes most drastically, as tracked in step S2.

[0123] After each state determination, the device will generate and output a document containing a timestamp, the current state code (S0-S3), and key evidence data (such as...). , , Structured status report.

[0124] This application provides a method for monitoring the tilt attitude of communication towers based on inertial sensors. It acquires acceleration time-series data collected by sensors on the communication tower within a preset time period. Based on the acceleration time-series data, it determines the baseline modal characteristics and instantaneous modal characteristics at multiple time points. Based on the instantaneous modal characteristics and the baseline modal characteristics, it determines the equivalent icing quality index and the modal evolution cost matrix. Then, using an optimal matching algorithm, it calculates and processes the modal evolution cost matrix to obtain the optimal modal evolution path. Using the optimal modal evolution path and the modal evolution cost matrix, it calculates the modal change concentration index. Finally, using the equivalent icing quality index and the modal change concentration index, it monitors the tilt attitude of the communication tower. This application uses structural state risk assessment to monitor the tilt attitude of communication towers. By tracking and quantifying the evolution path of each modal characteristic from the baseline state, the basis for risk assessment is shifted from the fuzzy vibration phenomenon itself to the quantifiable deterioration of specific structural modal characteristics as an internal cause of galloping. The modal change concentration index is calculated through the optimal modal evolution path and the modal evolution cost matrix. Then, the structural state of the communication tower is assessed for risk using the equivalent icing quality index and the modal change concentration index. This process correlates external vibration phenomena with the evolution process of the structure's internal characteristics, thereby effectively distinguishing between fluttering and galloping and providing timely early warning of dynamic instability of the communication tower.

[0125] Reference Figure 3 , Figure 3 This is a schematic diagram of the device structure of the hardware operating environment involved in the embodiments of this application.

[0126] like Figure 3 As shown, the communication tower tilt attitude monitoring device based on inertial sensors may include: a processor 1001, a memory 1003, and a communication bus 1002. The communication bus 1002 is used to realize the connection and communication between the processor 1001 and the memory 1003.

[0127] Optionally, the inertial sensor-based communication tower tilt attitude monitoring device may also include a user interface, a network interface, a camera, RF (Radio Frequency) circuitry, sensors, a WiFi module, etc. The user interface may include a display screen and an input submodule such as a keyboard; optional user interfaces may also include standard wired or wireless interfaces. The network interface may include standard wired or wireless interfaces (such as a Wi-Fi interface).

[0128] Those skilled in the art will understand that Figure 3The structure of the communication tower tilt attitude monitoring device based on inertial sensors shown does not constitute a limitation on the communication tower tilt attitude monitoring device based on inertial sensors. It may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0129] like Figure 3 As shown, the memory 1003, serving as a storage medium, may include an operating system, a network communication module, and a communication tower tilt attitude monitoring program based on inertial sensors. The operating system is a program that manages and controls the hardware and software resources of the communication tower tilt attitude monitoring device based on inertial sensors, supporting the operation of the communication tower tilt attitude monitoring program based on inertial sensors and other software and / or programs. The network communication module is used to enable communication between the various components within the memory 1003, as well as communication with other hardware and software in the communication tower tilt attitude monitoring system based on inertial sensors.

[0130] exist Figure 3 In the communication tower tilt attitude monitoring device based on inertial sensors shown, the processor 1001 is used to execute the communication tower tilt attitude monitoring program based on inertial sensors stored in the memory 1003 to implement the steps of the above-mentioned communication tower tilt attitude monitoring method based on inertial sensors.

[0131] The specific implementation of the communication tower tilt attitude monitoring device based on inertial sensors in this application is basically the same as the embodiments of the communication tower tilt attitude monitoring method based on inertial sensors described above, and will not be repeated here.

[0132] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0133] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0134] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, including instructions used to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0135] The above are merely preferred embodiments of this application and do not limit the scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the scope of protection of this application.

[0136] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0137] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for monitoring the tilt attitude of communication towers based on inertial sensors, characterized in that, The method includes: Acquire acceleration time-series data collected by sensors on the communication tower within a preset time period; Based on the acceleration time series data, reference modal characteristics and instantaneous modal characteristics at multiple times are determined. The reference modal characteristics include a set of reference modal frequencies and a set of reference modal damping ratios. Based on the instantaneous modal features and the baseline modal features, the equivalent icing quality index and the modal evolution cost matrix are determined. The modal evolution cost matrix is ​​then calculated and processed using an optimal allocation algorithm to obtain the optimal modal evolution path. Based on the optimal modal evolution path and the modal evolution cost matrix, the modal change concentration index is calculated. Based on the equivalent icing quality index and the modal change concentration index, the structural state of the communication tower is assessed for risk, so as to monitor the tilt attitude of the communication tower.

2. The method for monitoring the tilt attitude of communication towers based on inertial sensors as described in claim 1, characterized in that, The determination of the baseline modal characteristics and instantaneous modal characteristics at multiple time points based on the acceleration time-series data includes: The acceleration time series data is processed by a preset recognition algorithm to obtain the modal frequency and modal damping ratio. The data set corresponding to a preset number of modal frequencies and modal damping ratios before sorting, the frequency variance vector and the damping ratio variance vector corresponding to the data set, are used as the reference modal features; The acceleration time series data within the current analysis period are processed for parameter identification to obtain instantaneous modal characteristics at multiple moments. The duration of the current analysis period is shorter than the duration of the preset time period.

3. The method for monitoring the tilt attitude of communication towers based on inertial sensors as described in claim 1, characterized in that, The determination of the equivalent icing quality index and the mode evolution cost matrix based on the instantaneous modal features and the baseline modal features includes: Extract the frequency variance vector and damping ratio variance vector from the reference modal features; Based on the frequency variance vector and the damping ratio variance vector, the frequency change weight and the damping change weight are calculated. Based on the frequency change weight, damping change weight, instantaneous modal features, and baseline modal features, a modal evolution cost matrix is ​​constructed; The equivalent icing quality index is calculated based on the ratio between the instantaneous modal features and the baseline modal features.

4. The method for monitoring the tilt attitude of communication towers based on inertial sensors as described in claim 1, characterized in that, After determining the equivalent icing quality index and the modal evolution cost matrix based on the instantaneous modal features and the baseline modal features, the process further includes: When the equivalent icing mass index meets the preset conditions, the acceleration time series data in the current analysis period is processed by a preset identification algorithm to obtain the candidate reference mode features of the current analysis period. The baseline health metric of the baseline modal feature and the candidate health metric of the candidate baseline modal feature are calculated by using a preset health metric function. When the candidate health metric is greater than the baseline health metric, the candidate baseline modal feature replaces the baseline modal feature.

5. The method for monitoring the tilt attitude of communication towers based on inertial sensors as described in claim 1, characterized in that, The modal change concentration index is calculated based on the optimal modal evolution path and the modal evolution cost matrix, including: Based on the optimal modal evolution path and the modal evolution cost matrix, calculate the change vector of each mode; When the change vector satisfies the preset boundary conditions, the first concentration index is calculated based on the change vector. When the change vector does not meet the preset boundary conditions, a second concentration index is calculated based on the change vector. The first concentration index or the second concentration index is used as the modal change concentration index.

6. The method for monitoring the tilt attitude of communication towers based on inertial sensors as described in claim 5, characterized in that, When the change vector does not meet the preset boundary conditions, a second concentration index is calculated based on the change vector, including: The sum of the change vectors of each modal data is calculated to obtain the total change. When the total change does not meet the preset boundary conditions, the change vector is normalized to obtain the probability distribution vector. Calculate the Shannon information entropy of the probability distribution vector and use the Shannon information entropy as the second concentration index.

7. The method for monitoring the tilt attitude of communication towers based on inertial sensors as described in claim 5, characterized in that, The risk assessment of the structural status of communication towers based on the equivalent icing quality index and the modal variation concentration index includes: When the equivalent icing quality index is less than or equal to the first preset threshold, the structural state of the communication tower is determined to be an ice-free state. When the equivalent icing quality index is greater than the second preset threshold and the modal change concentration index is greater than or equal to the third preset threshold, the structural state of the communication tower is determined to be uniform icing. When the equivalent icing quality index is greater than the second preset threshold and the modal change concentration index is less than the third preset threshold, the structural state of the communication tower is determined to be asymmetric icing state. When the communication tower is in an asymmetric icing state and a self-excited vibration triggered by galloping is detected, it is determined that the structural state of the communication tower is in a state of dynamic instability.

8. The method for monitoring the tilt attitude of communication towers based on inertial sensors as described in claim 7, characterized in that, When the communication tower is in an asymmetric icing state and galloping self-excited vibration is detected, it is determined that the communication tower is in a dynamic instability state, including: When the communication tower is in an asymmetric icing state, the vibration energy and vibration frequency within the current analysis period are obtained; When the vibration energy is greater than a preset high energy threshold and the deviation between the vibration main frequency and the target instantaneous modal frequency is less than a preset locking tolerance, it is determined that a galloping self-excited vibration has been detected and the communication tower is in a state of dynamic instability.

9. The method for monitoring the tilt attitude of communication towers based on inertial sensors as described in claim 8, characterized in that, The methods for determining the target instantaneous modal frequency include: Extract the maximum value of the vector from the change vectors of each mode; Based on the optimal mode evolution path, determine the instantaneous mode corresponding to the maximum value of the vector; The frequency corresponding to the instantaneous mode is taken as the target instantaneous mode frequency.

10. A communication tower tilt attitude monitoring device based on inertial sensors, characterized in that, The device includes: The acquisition module is used to acquire acceleration time-series data collected by sensors on the communication tower within a preset time period; The determination module is used to determine the reference modal characteristics and the instantaneous modal characteristics at multiple times based on the acceleration time series data. The reference modal characteristics include a set of reference modal frequencies and a set of reference modal damping ratios. The processing module is used to determine the equivalent icing quality index and the mode evolution cost matrix based on the instantaneous mode features and the reference mode features, and to calculate and process the mode evolution cost matrix through the optimal allocation algorithm to obtain the optimal mode evolution path. The calculation module is used to calculate the mode change concentration index based on the optimal mode evolution path and the mode evolution cost matrix. The risk assessment module is used to assess the structural status of the communication tower based on the equivalent icing quality index and the modal change concentration index, so as to monitor the tilt attitude of the communication tower.

Images