Breeze vibration synchronous monitoring system and method based on split type deployment

Abstract

The invention discloses an aeolian vibration synchronous monitoring system and method based on split type deployment, and belongs to the technical field of intelligent monitoring of power transmission lines. A breeze vibration synchronous monitoring system is formed by an edge intelligent terminal, vibration sensors installed on an A-phase lead, a B-phase lead and a C-phase lead and a meteorological sensing unit, self-adaptive integral error removal is carried out on collected data in combination with data characteristics, and displacement signals corresponding to the three-phase lead are obtained respectively; in combination with a leakage compensation model, power compensation is carried out on the displacement signal, the corresponding aeolian vibration amplitude is determined, and the authenticity of the aeolian vibration amplitude is further guaranteed; meanwhile, real-time wind speed data and real-time wind direction data are obtained, the breeze vibration amplitude is calibrated according to the real-time wind speed data, the real-time wind direction data and a preset breeze vibration correlation model, and the accuracy of the data is verified. The problem that existing aeolian vibration monitoring accuracy needs to be improved is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent monitoring of power transmission lines, and in particular to a system and method for synchronously monitoring breeze vibration based on split deployment. Background Art

[0002] The statements in this section merely mention background art related to the present invention and do not necessarily constitute prior art.

[0003] Breeze vibration monitoring of transmission lines is one of the key technologies to ensure the safe operation of the power grid. Its core lies in preventing potential risks such as conductor fatigue breakage and hardware wear caused by high-frequency, small-amplitude vibrations through real-time data collection and analysis.

[0004] Common breeze vibration monitoring methods include bending amplitude method and inverted vibration measurement method. These two methods use a cantilever beam mechanical structure with integrated strain gauges to measure the vibration amplitude and frequency at a distance of 89mm from the wire clamp cutout. Since the monitoring point itself is in a high-frequency vibration environment, the mechanical structure will deform and fail after long-term operation, resulting in inaccurate monitoring data and monitoring equipment failure. It is impossible to achieve synchronous sampling of ns-level vibrations at multiple monitoring points and meteorological environment (wind speed, wind direction), resulting in reduced measurement accuracy and inaccurate data. At the same time, the collected data is directly applied without considering the impact of various factors such as noise, temperature changes, sensor drift, and slight changes in gravity on data collection.

[0005] In addition, the existing technologies mostly monitor the transmission line as a whole, without considering the vibration differences between the phases of the transmission line, resulting in the accuracy of breeze vibration monitoring needs to be improved. Summary of the invention

[0006] In order to address the deficiencies in the prior art, the present invention provides a breeze vibration synchronous monitoring system and method based on split deployment. The split-deployed vibration sensors are used to reduce the impact of changes in the monitoring structure on the monitoring accuracy. At the same time, the impact of environmental factors on sensor monitoring and the correlation between vibration monitoring and meteorology, spatial distribution of conductors and other factors are considered to improve the accuracy of breeze vibration monitoring.

[0007] In a first aspect, the present invention provides a breeze vibration synchronous monitoring system based on split deployment;

[0008] A breeze vibration synchronous monitoring system based on split deployment, comprising:

[0009] A vibration sensor, wherein a plurality of the vibration sensors are respectively arranged on the three-phase conductors of the power line, and the vibration sensor comprises a separate vibration sensing unit respectively arranged at both ends of the tension clamp outlet, and is used to collect acceleration data at both ends of the tension clamp outlet;

[0010] A meteorological sensing unit for collecting real-time wind speed data and real-time wind direction data;

[0011] An edge intelligent terminal for obtaining multiple sets of acceleration data, performing differential operations respectively to obtain corresponding acceleration continuous-time signals; based on the acceleration continuous-time signals, performing adaptive integration error removal in combination with data characteristics to obtain corresponding displacement signals respectively; combining a leakage compensation model to perform power compensation on the displacement signals to determine corresponding micro-vibration amplitudes; obtaining real-time wind speed data and real-time wind direction data, and calibrating the micro-vibration amplitudes in combination with a preset micro-vibration correlation model.

[0012] In some embodiments, performing adaptive integration error removal in combination with data characteristics based on the acceleration continuous-time signals to obtain corresponding displacement signals respectively includes:

[0013] Calculating the standard deviation and average value of the acceleration continuous-time signals, and preprocessing the acceleration continuous-time signals according to the velocity continuous-time signals, the standard deviation and the average value;

[0014] Performing a bias elimination operation and an integration operation on the preprocessed acceleration continuous-time signals in sequence to obtain a velocity time series; performing a bias elimination operation and an integration operation on the velocity time series in sequence to obtain an initial displacement signal;

[0015] Performing fitting by the Gaussian column main elimination method to eliminate the signal trend in the initial displacement signal and obtain a displacement signal.

[0016] In some embodiments, performing fitting by the Gaussian column main elimination method to eliminate the trend signal in the initial displacement signal and obtain a displacement signal includes:

[0017] Constructing a polynomial trend model of the initial displacement signal, and fitting the polynomial trend model by the Gaussian column main elimination method to obtain a trend signal sequence;

[0018] Determining the displacement signal according to the difference between the initial displacement signal and the trend signal sequence.

[0019] In some embodiments, combining the leakage compensation model to perform power compensation on the displacement signal to determine the corresponding micro-vibration amplitude includes:

[0020] Performing a Fourier transform on the displacement signal to determine the corresponding frequency and amplitude of the displacement signal;

[0021] Compensating the amplitude based on the frequency and a preset leakage compensation model to determine the corresponding micro-vibration amplitude.

[0022] In some embodiments, constructing the leakage compensation model includes:

[0023] Constructing a vibration amplitude attenuation coefficient polynomial model according to the frequency offset between the main frequency point frequency and the adjacent leakage point frequency;

[0024] Constructing a leakage compensation model according to the energy distribution of the main frequency point and the leakage point; combining the leakage compensation model and the vibration amplitude attenuation coefficient polynomial model, and using the frequency offset to determine the leakage compensation factor and update the leakage compensation model.

[0025] In some embodiments, the leakage compensation model is expressed as:

[0026] A real =A s *P correced (f0) / P0;

[0027]

[0028] In the formula, A real represents the amplitude of the gentle breeze vibration, A s represents the amplitude, f0 represents the main frequency point frequency, P0 represents the main frequency point power, Δf i represents the i-th frequency offset, and k i represents the i-th leakage compensation factor.

[0029] In some embodiments, constructing the gentle breeze vibration correlation model specifically includes: using the wind speed, the angle between the wind direction and the power line, and the amplitude of the gentle breeze vibration response to construct the gentle breeze vibration correlation model.

[0030] In some embodiments, the edge intelligent terminal is further configured to obtain the power information data of the separated vibration sensing unit, process the power information data, the real-time wind direction data, and the real-time wind speed data through a preset wind speed and wind direction power-sampling model, obtain the sampling frequency and the waveform sampling duration parameters, and transmit them to the separated vibration sensing unit.

[0031] In some embodiments, processing the power information data, the real-time wind direction data, and the real-time wind speed data through a preset wind speed and wind direction power-sampling model to obtain the sampling frequency and the waveform sampling duration parameters specifically includes: according to the comparison result between the battery power, the charging current and the preset threshold range, combining the comparison result between the wind speed data and the wind direction data and the preset threshold, and outputting the corresponding sampling frequency and the waveform sampling duration parameters.

[0032] In a second aspect, the present invention provides a method for synchronously monitoring the gentle breeze vibration based on a split deployment;

[0033] A method for synchronously monitoring the gentle breeze vibration based on a split deployment includes:

[0034] Obtain multiple sets of acceleration data and perform differential operations on them respectively to obtain corresponding continuous-time acceleration signals; wherein, the acceleration data is collected by split vibration sensing units arranged at both ends of the outlet of the strain clamp.

[0035] Based on the continuous-time acceleration signals, perform adaptive integral error removal in combination with data characteristics to obtain corresponding displacement signals respectively; in combination with a leakage compensation model, perform power compensation on the displacement signals to determine corresponding aeolian vibration amplitudes.

[0036] Obtain real-time wind speed data and real-time wind direction data, and calibrate the aeolian vibration amplitude in combination with a preset aeolian vibration correlation model; wherein, the real-time wind speed data and the real-time wind direction data are collected by a meteorological sensing unit.

[0037] In some embodiments, performing adaptive integral error removal in combination with data characteristics based on the continuous-time acceleration signals to obtain corresponding displacement signals respectively includes:

[0038] Calculate the standard deviation and average value of the continuous-time acceleration signals, and perform preprocessing on the continuous-time acceleration signals according to the continuous-time velocity signals, the standard deviation, and the average value.

[0039] Perform a bias elimination operation and an integration operation on the preprocessed continuous-time acceleration signals in sequence to obtain a velocity time series; perform a bias elimination operation and an integration operation on the velocity time series in sequence to obtain an initial displacement signal.

[0040] Perform fitting through the Gaussian column pivoting method to eliminate the signal trend in the initial displacement signal and obtain the displacement signal.

[0041] In some embodiments, performing fitting through the Gaussian column pivoting method to eliminate the trend signal in the initial displacement signal and obtain the displacement signal includes:

[0042] Construct a polynomial trend model of the initial displacement signal, and fit the polynomial trend model through the Gaussian column pivoting method to obtain a trend signal sequence.

[0043] Determine the displacement signal according to the difference between the initial displacement signal and the trend signal sequence.

[0044] In some embodiments, the combining the leakage compensation model to perform power compensation on the displacement signal and determine the corresponding aeolian vibration amplitude includes:

[0045] Perform a Fourier transform on the displacement signal to determine the frequency and amplitude corresponding to the displacement signal.

[0046] Perform compensation on the amplitude based on the frequency and a preset leakage compensation model to determine the corresponding aeolian vibration amplitude.

[0047] In some embodiments, constructing the leakage compensation model includes:

[0048] Constructing a vibration amplitude attenuation coefficient polynomial model according to the frequency offset between the main frequency point frequency and the adjacent leakage point frequencies;

[0049] Constructing a leakage compensation model according to the energy distributions of the main frequency point and the leakage points; combining the leakage compensation model and the vibration amplitude attenuation coefficient polynomial model, and using the frequency offset to determine the leakage compensation factor and update the leakage compensation model.

[0050] In some embodiments, the leakage compensation model is expressed as:

[0051]

[0052] wherein, f0 represents the main frequency point frequency, P0 represents the main frequency point power, Δf i represents the i-th frequency offset, and k i represents the i-th leakage compensation factor.

[0053] In some embodiments, constructing the aeolian vibration correlation model specifically includes: using the wind speed, the angle between the wind direction and the power line, and the amplitude of the aeolian vibration response to construct the aeolian vibration correlation model.

[0054] In some embodiments, it further includes: obtaining the power information data of the discrete vibration sensing unit, processing the power information data, the real-time wind direction data, and the real-time wind speed data through a preset wind speed-wind direction-power-sampling model, and obtaining the sampling frequency and the waveform sampling duration parameters and transmitting them to the discrete vibration sensing unit.

[0055] In some embodiments, processing the power information data, the real-time wind direction data, and the real-time wind speed data through a preset wind speed-wind direction-power-sampling model to obtain the sampling frequency and the waveform sampling duration parameters specifically includes: according to the comparison results of the battery power and the charging current with a preset threshold range, and combining the comparison results of the wind speed data and the wind direction data with a preset threshold, outputting the corresponding sampling frequency and the waveform sampling duration parameters.

[0056] Compared with the prior art, the beneficial effects of the present invention are:

[0057] 1. The technical solution provided by the present invention deploys the vibration sensing units in a split manner. The vibration sensing units interact with each other through wireless communication without physical connection, which is more convenient for installation and maintenance. It avoids the problem of cable and structure breakage caused by the vibration of the monitoring points in the integrated connection, reduces the impact on the monitored vibration amplitude frequency, improves the accuracy of the sampled data at the monitoring points, ensures the reliability of the device itself, and simultaneously monitors the vibration of the three-phase conductors, ensuring the comprehensiveness of the aeolian vibration monitoring.

[0058] 2. The technical solution provided by the present invention processes the data through the trend removal method of removing anomalies, eliminating biases, and self-adapting, and establishing a compensation model for FFT spectrum leakage, eliminating the interference caused by external factors to the data collected by the vibration sensor, improving the accuracy and authenticity of the collected data, and avoiding system misjudgment.

[0059] 3. The technical solution provided by the present invention constructs a wind speed, wind direction, power - sampling model, and adaptively adjusts the sampling frequency and the sampling waveform length based on the instantaneous data of wind speed and wind direction, the current power, and the charging state, so as to reduce the overall operating power consumption of the system as much as possible, thereby reducing the configuration of the device power system capacity (the capacity configuration of the battery + solar panel), greatly reducing the overall weight of the device, making the device more portable, and minimizing the addition of hardware at the vibration monitoring point position, which affects the accuracy of sampling at the monitoring point. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The specification drawings forming a part of the present invention are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation to the present invention.

[0061] Figure 1 It is a deployment schematic diagram of the vibration sensor provided by the embodiment of the present invention;

[0062] Figure 2 It is a data transmission schematic diagram of the aeolian vibration synchronous monitoring system based on split deployment provided by the embodiment of the present invention;

[0063] Figure 3 It is a data processing schematic diagram of the wind speed, wind direction, power - sampling model provided by the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0064] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used in the present invention have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0065] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular forms are also intended to include the plural forms. In addition, it should be understood that the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units need not be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0066] Without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other.

[0067] Embodiment 1

[0068] The existing monitoring of aeolian vibration is restricted by mechanical structures, external environmental factors, etc., resulting in the accuracy and authenticity of the monitoring effect to be improved. Therefore, the present invention provides a synchronous monitoring system for aeolian vibration based on split deployment. Through the cooperation of vibration sensors, meteorological sensing units and edge intelligent terminals arranged in a split manner, synchronous monitoring of aeolian vibration of three-phase conductors is realized and the monitoring accuracy is improved.

[0069] Next, in combination with Figures 1-3 , a synchronous monitoring system for aeolian vibration based on split deployment disclosed in this embodiment will be described in detail. The synchronous monitoring system for aeolian vibration based on split deployment includes vibration sensors, meteorological sensing units and edge intelligent terminals. The edge intelligent terminals are communicatively connected to the vibration sensors and meteorological sensing units through LoRa communication channels; the number of vibration sensors is 3, which are respectively deployed on the A, B, and C phase conductors; each single vibration sensor includes split vibration sensing units respectively deployed at both ends of the outlet of the strain clamp (the position with the largest bending amplitude); the split vibration sensing units are used to collect acceleration data at both ends of the outlet of the strain clamp, the meteorological sensing unit is used to collect real-time wind speed data and real-time wind direction data, and the edge intelligent terminal is used to obtain multiple groups of acceleration data and perform differential operations respectively to obtain corresponding acceleration continuous-time signals; based on the acceleration continuous-time signals, adaptive integration error removal is performed in combination with data characteristics to respectively obtain corresponding displacement signals; in combination with a leakage compensation model, power compensation is performed on the displacement signals to determine corresponding aeolian vibration amplitudes; real-time wind speed data and real-time wind direction data are obtained, and the aeolian vibration amplitudes are calibrated in combination with a preset aeolian vibration correlation model.

[0070] Based on this, by processing the three groups of acceleration data collected by the three vibration sensors, the corresponding aeolian vibration amplitudes of the three-phase conductors are obtained.

[0071] In this embodiment, the distance between the two separate vibration sensing units is 89 mm. The separate vibration sensing unit includes a MEMS accelerometer and an MCU. The MCU is communicatively connected to the MEMS accelerometer through a data line. The separate vibration sensing unit and the edge intelligent terminal are wirelessly communicatively connected through a LoRa module.

[0072] Next, in combination with the working method of the micro-vibration synchronous monitoring system based on split deployment described in this embodiment, the micro-vibration synchronous monitoring system based on split deployment will be further described.

[0073] As an implementation manner, before the separate vibration sensing unit works, it further includes:

[0074] The edge intelligent terminal sends a timing instruction to the separate vibration sensing unit through the LoRa communication channel. After receiving the timing instruction, the separate vibration sensing unit completes timing with an accuracy of 100 ns through a single Beidou timing chip. After the separate vibration sensing unit completes timing, it sends a timing completion flag to the edge intelligent terminal. If the timing is not completed within the timeout (default 3 minutes), the edge intelligent terminal will re-specify and send an instruction to the separate vibration sensing unit that has not completed timing until the timing of the 6 separate vibration sensing units and itself is completed.

[0075] In order to save costs and achieve low power consumption of the device, it is necessary to minimize the computing resources and power consumption of the MCU as much as possible. Therefore, it is necessary to achieve a short sampling time, and the sampled data just meets the analysis resource requirements. Therefore, in this embodiment, the sampling rate f sample is set to 2000 HZ, the sampling time is 1.2 seconds, and there are a total of 2400 points.

[0076] During sampling, the edge intelligent terminal issues a command and monitors the LORA module that wakes up the vibration sensor in the air. The LORA module wakes up the MCU of the separate vibration sensing unit by means of external IO port triggering. The low-power MCU starts the MEMS accelerometers of the two ends of the vibration sensing unit according to the instruction command.

[0077] Based on this, trigger sampling of the separate vibration sensing unit is achieved through LORA paging in the air to reduce power consumption.

[0078] Further, obtaining multiple groups of acceleration data and performing differential operations respectively to obtain corresponding acceleration continuous time signals; based on the acceleration continuous time signals, adaptive integration error removal is performed in combination with data characteristics to obtain corresponding displacement signals specifically including:

[0079] Step 1: Perform differential operations on the acceleration data at both ends of the outlet position of the strain clamp to obtain the acceleration continuous time signal a(t).

[0080] Step 2: Calculate the standard deviation and mean value of the acceleration continuous-time signal, and preprocess the acceleration continuous-time signal according to the velocity continuous-time signal, standard deviation, and mean value.

[0081] Specifically, first, calculate the standard deviation σa of the acceleration continuous-time signal a(t), and calculate the mean value of a(t) Then, determine whether a(t) satisfies If it is satisfied, the data sampling is normal and no smoothing is required, and go to Step 3; if it is not satisfied, perform smoothing processing according to the data points before and after the current signal. The smoothing formula is as follows:

[0082] a(t) = (8a(t - 1) + 5a(t + 1)) / 13.

[0083] Through the smoothing process in Step 2, obvious abnormal signals in the acceleration continuous-time signal can be removed.

[0084] Step 3: Perform a debiasing operation and an integration operation on the preprocessed acceleration continuous-time signal in sequence to obtain a velocity time series.

[0085] Remove the DC component (debias) of the acceleration continuous-time signal to reduce the trend during integral calculation; specifically, the removal method is to subtract the mean value of this time series from a(t) Thus, a new acceleration time series is obtained, expressed as:

[0086]

[0087] Integrate a(t) to obtain a velocity time series v(t), expressed as:

[0088] v(t) = ∫a(t)dt.

[0089] Step 4: Perform the same debiasing operation and integration operation as in Step 3 on the velocity time series in sequence to obtain an initial displacement signal s(t).

[0090] Step 5: Perform fitting by the Gaussian column main elimination method to eliminate the signal trend in the initial displacement signal and obtain a displacement signal.

[0091] Since s(t) has undergone double integration and there is a large integral cumulative error in the signal, it is necessary to effectively remove the integral trend to obtain a displacement time series that can more accurately reflect the phenomenon of aeolian vibration; therefore, in this embodiment, an adaptive integral error removal method for automatically removing the trend according to the characteristics of the input data is provided. The specific process is as follows:

[0092] First, design a polynomial trend model for fitting the initial displacement signal, expressed as:

[0093] s trend (t)=c0 + c1t + c2t 2 +…+c n t n 。

[0094] Here, the fitting target of the discrete data is: S*C = Y, where C is the coefficient vector to be solved, C = [c0, c1, c2,..., c n T , Y is the signal vector Y = [s1, s2,..., s m T , S is the fitting matrix, which is expressed as follows:

[0095]

[0096] Perform initial settings before removing the trend. According to experimental experience and model analysis, in this embodiment, the maximum fitting order is defined as 8, and the minimum fitting order is 2; define an error threshold ∈ to measure whether the fitting residual meets the requirements. Gradually increase the order, starting from the minimum fitting order 2, and gradually increase the fitting order. Each time, calculate the trend signal s trend (t) and the residual residual = ||Y - S*C||2. If the residual is less than the error threshold ∈, stop increasing the order; set a constraint condition. If the residual still does not meet the requirements when reaching the maximum fitting order 8, return the fitting result of the highest order.

[0097] Then, in order to solve S*C = Y, obtain the coefficient C through the Gaussian column main elimination method; specifically, construct an augmented matrix [S|Y], which is expressed as:

[0098]

[0099] Select the pivot element. Select the element with the largest absolute value in each column as the pivot element and exchange rows; during the elimination process, for the k-th column, make all elements after the k + 1-th row become 0; back-substitute to solve, and solve for C starting from the last row using the back-substitution method; output the fitting polynomial order n, the fitting coefficient C, and s trend (t), then the displacement signal sequence after removing the trend is expressed as

[0100] s detrend (t)=s(t)-s trend (t).

[0101] Furthermore, in combination with the leakage compensation model, perform power compensation on the displacement signal to determine the corresponding micro-vibration amplitude, including:

[0102] S1. Use the fast Fourier transform to calculate the displacement signal s detrend ​​(t) frequency and amplitude A s .

[0103] It is caused by the mismatch between the basic assumptions of the discrete Fourier transform (DFT) and the actual signal. Specifically, the essence of spectral leakage is that after the signal is truncated, it no longer satisfies the periodicity assumption, resulting in the spectrum "leaking" from one frequency point to other frequency points. For example, if the frequency and the frequency resolution are not aligned, and the true frequency of the signal is 5 Hz, but the FFT resolution corresponding to the sampling points is 4.8 Hz or 5.2 Hz, this will cause leakage.

[0104] FFT is a fast Fourier transform for efficient calculation. Due to the fact that its resolution cannot be infinitely high, the values obtained after FFT calculation will be inaccurate after spectral leakage occurs, and the amplitude A s is the characteristic quantity of aeolian vibration - the initial aeolian vibration amplitude, and the A obtained through FFT calculation s will have varying degrees of energy attenuation due to calculation reasons, resulting in an amplitude smaller than the actual value; therefore, in this embodiment, leakage compensation is performed through the following steps.

[0105] S2. Compensate the aeolian vibration amplitude based on a preset leakage compensation model to determine the corresponding aeolian vibration amplitude; expressed as:

[0106] A real = A s * P correced (f0) / P0;

[0107]

[0108] Δf i = |f i - f0|;

[0109] In the formula, f0 represents the main frequency point frequency, P0 represents the main frequency point power, and Δf i represents the i-th frequency offset, f i represents the i-th frequency point frequency, and k i represents the i-th leakage compensation factor.

[0110] Here, the main frequency point frequency and the main frequency point power are the values obtained through experiments when constructing the leakage compensation model.

[0111] In this step, only the aeolian vibration amplitude A needs to be compensated through the pre-constructed leakage compensation model s to redistribute the leaked power components back to the main frequency point. After compensation, the amplitude of the main frequency point is increased, the leaked energy is effectively compensated back to the main frequency point, and the spectrum is closer to the true value.

[0112] Further, constructing the leakage compensation model specifically includes: constructing a vibration amplitude attenuation coefficient polynomial model according to the frequency offset between the main frequency point frequency and the adjacent leakage point frequencies; constructing the leakage compensation model according to the energy distribution of the main frequency point and the leakage points; combining the leakage compensation model and the vibration amplitude attenuation coefficient polynomial model, and using the frequency offset to determine the leakage compensation factor and update the leakage compensation model.

[0113] Specifically, first, according to the monitoring requirements of aeolian vibration, different sine signals from 0 to 200 HZ are simulated and input, with the vibration amplitude being a fixed amplitude and increasing by 1 HZ each time. The calculated amplitude of each group of signals after FFT calculation is recorded through software. According to the FFT results obtained from the experiment, there is a certain offset between the main frequency point f0 and its adjacent leakage points f1, f2, …. Let the frequency offset be expressed as:

[0114] Δf i =|f i -f0|.

[0115] Then, establish the vibration amplitude attenuation coefficient polynomial model in the experimental environment

[0116] y=k0x + k1x 2 + k2x 3 +...;

[0117] where x is the frequency offset Δf i , y describes the attenuation ratio of the power amplitude with the offset, and k0, k1, k2 need to be obtained through experimental fitting.

[0118] Further, to reduce spectral leakage, a window is added to the signal before performing FFT, w[n] = 0.54 - 0.46cos(2πn / (N - 1)). This step can reduce the influence of the spectral leakage phenomenon, but this phenomenon still exists.

[0119] According to the above simulation experiment, assuming that there is leakage at a certain frequency point f0, the leaked spectral power will spread to adjacent frequency points f1, f2, …. Estimate the leakage distribution ratio based on the leakage characteristics after windowing (such as the frequency domain response of the window function). In most cases, the leakage is mainly concentrated in 2 - 3 frequency points outside the main lobe. The range of the main lobe can be calculated through the window function theory, and the weight ratio of the leakage energy distribution is determined. Let the power compensation formulas for the main frequency point f0 and its adjacent leakage points f1, f2, … be

[0120] P corrected (f0) = P0 + k1P1 + k2P2 +...;

[0121] where k n is the leakage weight factor, which is an empirical weight used to adjust the contribution of adjacent leakage points.

[0122] Bring the frequency offset into the pre-set attenuation model in the experiment. For each leakage point f i , calculate its frequency offset, expressed as:

[0123] x i = Δf i = |f i - f0|;

[0124] Substitute x i into the vibration amplitude attenuation model to obtain the attenuation ratio y of the corresponding leakage point i used to describe the leakage contribution weight of this point to the main frequency point; expressed as:

[0125] y i = k0x i + k1x i 2 + k2x i 3 +...;

[0126] The k in the leakage compensation formula i is directly related to the vibration attenuation ratio y i . Without loss of generality, directly set k i = y i , that is, the value of the leakage compensation factor k i comes from the vibration amplitude attenuation model. Therefore, for each leakage point f i , the final compensation factor is:

[0127] k i = k0Δf i + k1(Δf i ) 2 + k2(Δf i ) 3 +...

[0128] Combine the power contributions and compensation factors of all leakage points to obtain the final leakage compensation model, expressed as:

[0129] A real = A s * P correced (f0) / P0;

[0130]

[0131] In summary, a leakage compensation model applicable to the light wind monitoring scenario can be obtained, which can be directly called during actual light wind monitoring to ensure the real-time nature of data monitoring.

[0132] Furthermore, the construction of the aeolian vibration correlation model is specifically as follows: The aeolian vibration correlation model is constructed by using the wind speed, the angle between the wind direction and the power line, and the amplitude of the aeolian vibration response.

[0133] The wind speed determines the energy of air flow, thus affecting the vibration intensity and excitation frequency of structures or equipment. Aeolian vibration is usually caused by low-speed wind. When the wind speed is low, the aerodynamic force generated is relatively small, and the vibration amplitude of the structure is also small. However, as the wind speed increases, the aerodynamic effects become more significant, which may trigger stronger vibrations, especially near the resonance frequency of the structure. The wind direction determines the distribution and influence of the wind force on the structure. Different wind directions may cause uneven distribution of the wind force on the structure, thus affecting the vibration mode and intensity.

[0134] The wind pressure is proportional to the square of the wind speed. Therefore, the greater the wind speed, the significantly increased aerodynamic load and excitation force generated by the wind on the object, thus triggering stronger vibrations. For aeolian vibration, it can be assumed that the excitation force generated by the wind speed on the structure is the main factor affecting the vibration intensity. The relationship between the vibration intensity and the excitation force can be represented by a simplified linear model:

[0135] F wind = kv 2 ;

[0136] where k is a constant, depending on the shape and surface characteristics of the vibration monitoring unit structure.

[0137] The influence of the wind direction on the vibration intensity is more complex because the wind direction determines the direction of the wind force and the distribution of the air flow on the surface of the object. The wind direction will affect the way the object is affected by the wind, resulting in different vibration modes. We simplify the analysis of the model. Assuming that the angle between the wind direction and the windward surface of the object is θ, the effective wind pressure will be affected by the wind direction angle. The component of the wind force can be expressed as:

[0138] F effective = F wind *cos(θ);

[0139] where F wind = kv 2 is the total wind force excitation, θ is the angle between the wind direction and the normal of the object surface. When the wind direction is directly facing the object, that is, θ is 0 degree, the influence of the wind force on the object is the greatest; when the wind direction is parallel to the object surface, that is, θ is 90 degrees, the effective action of the wind force is the smallest. Different wind directions will change the action point of the wind force on the object, thereby affecting the vibration intensity and mode.

[0140] Specifically, in the actual situation of transmission line monitoring, the wind speed and wind direction act together on the aeolian vibration intensity of the object. Considering the wind speed and wind direction, the aeolian vibration correlation model is expressed as:

[0141]

[0142] In the formula, X vibration represents the amplitude (intensity) of the aeolian vibration response, A represents a constant related to factors such as structural stiffness, mass, and damping, w n represents the natural frequency of the structure, θ represents the angle between the wind direction and the normal of the object surface, and v represents the wind speed.

[0143] Exemplarily, in the working process of the aeolian vibration synchronous monitoring system based on split deployment, the collected real-time wind direction data and real-time wind speed data are brought into the aeolian vibration correlation model to obtain the amplitude of the aeolian vibration response, and it is judged whether the difference between the obtained amplitude of the aeolian vibration response is within a preset range. If so, the collected amplitude of the aeolian vibration is accurate.

[0144] As an implementation manner, the edge intelligent terminal is further configured to obtain the power information data of the separated vibration sensing unit, process the power information data, the real-time wind direction data, and the real-time wind speed data through a preset wind speed and wind direction power-sampling model, obtain the sampling frequency and the waveform sampling duration parameters, and transmit them to the separated vibration sensing unit.

[0145] Specifically, according to the comparison result of the battery power and the charging current with a preset threshold range, combined with the comparison result of the wind speed data and the wind direction data with a preset threshold, the corresponding sampling frequency and waveform sampling duration parameters are output.

[0146] Exemplarily, when the wind speed is 0, the sampling call is not initiated; when the battery power is greater than 80%, and the wind speed is greater than 0 m / s, sampling is performed according to the standard sampling frequency and the number of waveform sampling points; when the battery power is in the range of 80% - 50% and the charging current is greater than 100 mA, and the wind speed is greater than 0 m / s, sampling is performed according to the standard sampling frequency and the number of waveform sampling points; when the battery power is in the range of 80% - 50% and the charging current is greater than 50 mA and less than 100 mA, and the wind speed is greater than 0 m / s, the sampling frequency is reduced by 10% and the waveform sampling length is reduced by 10% for sampling; when the battery power is in the range of 80% - 50% and the charging current is less than 50 mA, and the wind speed is greater than 0 m / s, the sampling frequency is reduced by 20% and the waveform sampling length is reduced by 20% for sampling; when the battery power is in the range of 50% - 30% and the charging current is greater than 100 mA, and the wind speed is greater than 0 m / s, the sampling frequency is reduced by 10% and the waveform sampling length is reduced by 10% for sampling; when the battery power is in the range of 50% - 30% and the charging current is greater than 50 mA and less than 100 mA, and the wind speed is greater than 0 m / s, the sampling frequency is reduced by 30% and the waveform sampling length is reduced by 30% for sampling; when the battery power is in the range of 10% - 30% and the charging current is greater than 100 mA, and the wind speed is greater than 0 m / s, the sampling frequency is reduced by 30% and the waveform sampling length is reduced by 30% for sampling; when the battery power is in the range of 10% - 30% and the charging current is greater than 50 mA and less than 100 mA, and the wind speed is greater than 0 m / s, the sampling frequency is reduced by 40% and the waveform sampling length is reduced by 40% for sampling; when the battery power is in the range of 10% - 20% and the wind speed is greater than 5 m / s, sampling is triggered, and the sampling frequency is reduced by 40% and the waveform sampling length is reduced by 40% for sampling; when the battery power is less than 10%, when the wind speed is greater than 5 m / s and the included angle between the wind direction and the conductor is greater than 30°, sampling is triggered, and the sampling frequency is reduced by 40% and the waveform sampling length is reduced by 40% for sampling.

[0147] The weight of the vibration sensor is particularly important in the technical specifications of this system. The standard stipulates that the maximum does not exceed 1 kg; in fact, the lighter the weight, the smaller the impact on the monitoring point and the more accurate the measurement data. The power supply system of the vibration sensor adopts a power supply scheme of battery + solar panel, which accounts for a relatively large proportion in the overall weight of the device. Through the low-power design of the software and hardware of the device, the overall system power consumption of the device is reduced, the dependence on the power supply system capacity is reduced, and the structure is designed to be lightweight and miniaturized.

[0148] Embodiment 2

[0149] Based on the split-type deployed micro-vibration synchronous monitoring system described in Embodiment 1, this embodiment provides a split-type deployed micro-vibration synchronous monitoring method, which is deployed on an edge intelligent terminal; the method includes the following steps:

[0150] Obtain multiple groups of acceleration data and perform differential operations on them respectively to obtain corresponding continuous-time acceleration signals; among them, the acceleration data is collected by a split vibration sensing unit arranged at both ends of the outlet of the strain clamp.

[0151] Based on the continuous-time acceleration signals, perform adaptive integration error removal in combination with the data characteristics to obtain corresponding displacement signals respectively; in combination with the leakage compensation model, perform power compensation on the displacement signals to determine the corresponding aeolian vibration amplitudes.

[0152] Obtain real-time wind speed data and real-time wind direction data, and calibrate the aeolian vibration amplitude according to the real-time wind speed data and the real-time wind direction data in combination with a preset aeolian vibration correlation model; among them, the real-time wind speed data and the real-time wind direction data are collected by a meteorological sensing unit.

[0153] Further, obtain multiple groups of acceleration data and perform differential operations on them respectively to obtain corresponding continuous-time acceleration signals; based on the continuous-time acceleration signals, perform adaptive integration error removal in combination with the data characteristics to obtain corresponding displacement signals specifically including:

[0154] Step 1: Perform a differential operation on the acceleration data at both ends of the outlet position of the strain clamp to obtain a continuous-time acceleration signal a(t).

[0155] Step 2: Calculate the standard deviation and average value of the continuous-time acceleration signal, and preprocess the continuous-time acceleration signal according to the continuous-time velocity signal, standard deviation and average value.

[0156] Specifically, first, calculate the standard deviation σa of the continuous-time acceleration signal a(t), and calculate the average value of a(t) Then, judge whether a(t) satisfies If it is satisfied, the data sampling is normal and no smoothing is required, and step 3 is executed; if it is not satisfied, perform smoothing processing according to the data points before and after the current signal, and the smoothing formula is as follows:

[0157] a(t) = (8a(t - 1) + 5a(t + 1)) / 13.

[0158] Through the smoothing process in step 2, obvious abnormal signals in the continuous-time acceleration signal can be removed.

[0159] Step 3: Perform a debiasing operation and an integration operation on the preprocessed continuous-time acceleration signal in sequence to obtain a velocity time series.

[0160] Remove the DC component (debias) of the continuous-time acceleration signal to reduce the trend during integral calculation; specifically, the removal method is to subtract the average value of this time series from a(t) Thus, a new acceleration time series is obtained, expressed as:

[0161]

[0162] Integrate a(t) to obtain the velocity time series v(t), expressed as:

[0163] v(t) = ∫a(t)dt.

[0164] Step 4: Perform the same detrending operation and integration operation on the velocity time series as in Step 3 in sequence to obtain the initial displacement signal s(t).

[0165] Step 5: Perform fitting by Gaussian column pivoting elimination method to eliminate the signal trend in the initial displacement signal and obtain the displacement signal.

[0166] Since s(t) has undergone double integration and there is a large integral cumulative error in the signal, it is necessary to effectively remove the integral trend to obtain a displacement time series that can more accurately reflect the phenomenon of aeolian vibration; therefore, in this embodiment, an adaptive integral error removal method for automatically removing the trend according to the characteristics of the input data is provided. The specific process is as follows:

[0167] First, design a polynomial trend model for fitting the initial displacement signal, expressed as:

[0168] s trend (t) = c0 + c1t + c2t 2 +…+ c n t n .

[0169] Here, the fitting target for discrete data is: S*C = Y, where C is the coefficient vector to be solved C = [c0, c1, c2,..., c n T , Y is the signal vector Y = [s1, s2,..., s m T , and S is the fitting matrix, expressed as follows:

[0170]

[0171] Before removing the trend, perform initial settings. According to experimental experience and model analysis, in this embodiment, the maximum fitting order is defined as 8 and the minimum fitting number is 2; define an error threshold ∈ to measure whether the fitting residual meets the requirements. Gradually increase the order, starting from the minimum fitting number 2, and gradually increase the fitting order. Each time, calculate the fitting trend signal s trend ​​(t) and the residual residual = ||Y - S*C||2. If the residual is less than the error threshold ∈, stop increasing the order; establish a constraint condition. If the residual still does not meet the requirements when the maximum fitting order 8 is reached, return the fitting result of the highest order.

[0172] Then, to solve S*C = Y, obtain the coefficient C through Gauss column pivoting elimination method; specifically, construct an augmented matrix [S|Y], expressed as:

[0173]

[0174] Select the pivot element. Select the element with the largest absolute value in each column as the pivot element and exchange rows; during the elimination process, for the k-th column, make all elements after the k + 1-th row zero; back-substitute to solve, and solve for C starting from the last row using the back-substitution method; output the fitting polynomial order n, the fitting coefficient C, and s trend (t), then the detrended displacement signal sequence is expressed as

[0175] s detrend (t) = s(t) - s trend (t).

[0176] Furthermore, in combination with the leakage compensation model, perform power compensation on the displacement signal to determine the corresponding Aeolian vibration amplitude, including:

[0177] S1. Use the fast Fourier transform to calculate the frequency and amplitude A of the displacement signal s detrend (t). s .

[0178] It is caused by the mismatch between the basic assumptions of the discrete Fourier transform (DFT) and the actual signal. Specifically, the essence of spectral leakage is that after the signal is truncated, it no longer satisfies the periodicity assumption, resulting in the spectrum "leaking" from one frequency point to other frequency points. For example, the frequency and the frequency resolution are not aligned. If the true frequency of the signal is 5 Hz, but the FFT resolution corresponding to the sampling points is 4.8 Hz or 5.2 Hz, this will cause leakage.

[0179] FFT is a fast Fourier transform for efficient calculation. Due to the factor that its resolution cannot be infinitely high, the values obtained after FFT calculation will be inaccurate after spectral leakage, and the amplitude A s is the Aeolian vibration characteristic quantity - the initial Aeolian vibration amplitude. The A obtained through FFT calculation s will have different degrees of energy attenuation due to calculation reasons, resulting in an amplitude smaller than the actual value; therefore, in this embodiment, leakage compensation is performed through the following steps.

[0180] S2. Compensate the amplitude of aeolian vibration based on a preset leakage compensation model to determine the corresponding amplitude of aeolian vibration, expressed as:

[0181] A real =A s *P correced (f0) / P0;

[0182]

[0183] Δf i =|f i -f0|;

[0184] In the formula, f0 represents the main frequency point frequency, P0 represents the main frequency point power, Δf i represents the i-th frequency offset, f i represents the i-th frequency point frequency, k i represents the i-th leakage compensation factor.

[0185] Here, the main frequency point frequency and the main frequency point power are the values obtained through experiments when constructing the leakage compensation model.

[0186] In this step, only the amplitude of aeolian vibration A s needs to be compensated through the pre-constructed leakage compensation model, and the leaked power component is redistributed back to the main frequency point. After compensation, the amplitude of the main frequency point is increased, the leaked energy is effectively compensated back to the main frequency point, and the spectrum is closer to the true value.

[0187] Furthermore, the specific construction of the leakage compensation model is as follows: According to the frequency offset between the main frequency point frequency and the adjacent leakage point frequencies, construct a polynomial model of the vibration amplitude attenuation coefficient; According to the energy distribution of the main frequency point and the leakage points, construct a leakage compensation model; Combine the leakage compensation model and the polynomial model of the vibration amplitude attenuation coefficient, and use the frequency offset to determine the leakage compensation factor and update the leakage compensation model.

[0188] Specifically, first, according to the monitoring requirements of aeolian vibration, simulate and input different sine signals from 0 to 200 HZ with a fixed vibration amplitude incrementing by 1 HZ each time, and record the calculated amplitude of each group of signals after FFT calculation through software. According to the FFT results obtained from the experiment, there is a certain offset between the main frequency point f0 and its adjacent leakage points f1, f2, …, and the frequency offset is set to be expressed as:

[0189] Δf i =|f i -f0|.

[0190] Then, establish a polynomial model of the vibration amplitude attenuation coefficient in the experimental environment

[0191] y=k0x + k1x 2+k2x 3 +...;

[0192] where x is the frequency offset Δf i , y describes the attenuation ratio of the power amplitude with the offset, and k0, k1, k2 need to be obtained by experimental fitting.

[0193] Furthermore, to reduce spectral leakage, the signal is windowed before performing FFT, w[n] = 0.54 - 0.46cos(2πn / (N - 1)). This step can reduce the influence of the spectral leakage phenomenon, but this phenomenon still exists.

[0194] According to the above simulation experiment, assuming that there is leakage at a certain frequency point f0, the leaked spectral power will spread to adjacent frequency points f1, f2, …. Estimate the leakage distribution ratio based on the leakage characteristics after windowing (such as the frequency domain response of the window function). In most cases, the leakage is mainly concentrated in 2 - 3 frequency points outside the main lobe. The range of the main lobe can be obtained through theoretical calculation of the window function, and the weight ratio of the leaked energy distribution is determined. Let the compensation formula for the power P1, P2, … of the main frequency point f0 and its adjacent leakage points f1, f2, … be

[0195] P corrected (f0) = P0 + k1P1 + k2P2 +...;

[0196] where k n is the leakage weight factor, which is an empirical weight used to adjust the contribution of adjacent leakage points.

[0197] Substitute the frequency offset into the attenuation model preset in the experiment. For each leakage point f i , calculate its frequency offset, expressed as:

[0198] x i = Δf i = |f i - f0|;

[0199] Substitute x i into the vibration amplitude attenuation model to obtain the attenuation ratio yi of the corresponding leakage point, which is used to describe the leakage contribution weight of this point to the main frequency point; expressed as:

[0200] y i = k0x i + k1x i 2 + k2x i 3 +...;

[0201] The k i in the leakage compensation formula is directly related to the vibration attenuation ratio y i . Without loss of generality, directly set ki = y i , that is, the leakage compensation factor k i The value of comes from the vibration amplitude attenuation model. Therefore, for each leakage point f i , the final compensation factor is:

[0202] k i = k0Δf i + k1(Δf i ) 2 + k2(Δf i ) 3 +...

[0203] Combining the power contributions and compensation factors of all leakage points, the final leakage compensation model is obtained, expressed as:

[0204] A real = A s * P correced (f0) / P0;

[0205]

[0206] In summary, a leakage compensation model applicable to the light wind monitoring scenario can be obtained, which can be directly called during actual light wind monitoring to ensure the real-time nature of data monitoring.

[0207] Furthermore, the specific construction of the light wind vibration correlation model is as follows: using the angle between the wind speed, wind direction and the power line, and the amplitude of the light wind vibration response, the light wind vibration correlation model is constructed.

[0208] The wind speed determines the energy size of air flow, thus affecting the vibration intensity and excitation frequency of structures or equipment. Light wind vibration is usually caused by low-speed wind power. When the wind speed is low, the aerodynamic force generated is relatively small, and the vibration amplitude of the structure is also small. However, as the wind speed increases, the aerodynamic effects become more significant, which may trigger stronger vibrations, especially near the resonance frequency of the structure. The wind direction determines the distribution and influence of the wind force on the structure. Different wind directions may cause uneven distribution of the wind force on the structure, thus affecting the vibration mode and intensity.

[0209] The wind pressure is proportional to the square of the wind speed. Therefore, the greater the wind speed, the more significant the aerodynamic load and excitation force generated by the wind on the object, thus triggering stronger vibrations. For light wind vibration, it can be assumed that the excitation force generated by the wind speed on the structure is the main factor affecting the vibration intensity. The relationship between the vibration intensity and the excitation force can be represented by a simplified linear model:

[0210] F wind = kv 2 ;

[0211] Where k is a constant that depends on the shape and surface characteristics of the vibration monitoring unit structure.

[0212] The effect of wind direction on vibration intensity is more complicated because it determines the direction of wind force and the distribution of airflow on the surface of an object. Wind direction affects the way an object is affected by wind, resulting in different vibration modes. We simplify the model and assume that the angle of wind direction relative to the windward surface of the object is θ. The effective pressure of the wind will be affected by the wind direction angle. The components of wind force can be expressed as:

[0213] F effective =F wind *cos(θ);

[0214] Among them, F wind =kv 2 is the total wind excitation, θ is the angle between the wind direction and the normal line of the object surface. When the wind direction is facing the object, that is, θ is 0 degrees, the wind force has the greatest impact on the object; when the wind direction is parallel to the object surface, that is, θ is 90 degrees, the effective effect of the wind force is the smallest. Different wind directions will change the point of action of the wind force on the object, thereby affecting the intensity and mode of vibration.

[0215] Specifically, in the actual case of transmission line monitoring, wind speed and wind direction jointly affect the breeze vibration intensity of an object. Considering wind speed and wind direction, the breeze vibration correlation model is expressed as:

[0216]

[0217] In the formula, X vibration represents the amplitude (intensity) of the breeze vibration response, A represents a constant related to structural stiffness, mass, damping and other factors, and w n represents the natural frequency of the structure, θ represents the angle between the wind direction and the normal line of the object surface, and v represents the wind speed.

[0218] Exemplarily, during the operation of the breeze vibration synchronous monitoring system based on split deployment, the collected real-time wind direction data and real-time wind speed data are brought into the breeze vibration correlation model to obtain the amplitude of the breeze vibration response, and it is determined whether the difference between the amplitude and the obtained breeze vibration amplitude is within a preset range. If so, the collected breeze vibration amplitude is accurate.

[0219] As an implementation method, the edge intelligent terminal is also used to obtain power information data of the separate vibration sensing unit, process the power information data, real-time wind direction data and real-time wind speed data through a preset wind speed and wind direction power-sampling model, obtain the sampling frequency and waveform sampling duration parameters and transmit them to the separate vibration sensing unit.

[0220] Specifically, according to the comparison results of the battery power, charging current and the preset threshold range, combined with the comparison results of the wind speed data and wind direction data with the preset threshold, the corresponding sampling frequency and waveform sampling duration parameters are output.

[0221] Exemplarily, when the wind speed is 0, the sampling call is not initiated; when the battery power is greater than 80%, and the wind speed is greater than 0 m / s, sampling is performed according to the standard sampling frequency and waveform sampling points; when the battery power is in the range of 80% - 50%, and the charging current is greater than 100 mA, and the wind speed is greater than 0 m / s, sampling is performed according to the standard sampling frequency and waveform sampling points; when the battery power is in the range of 80% - 50%, and the charging current is greater than 50 mA and less than 100 mA, and the wind speed is greater than 0 m / s, the sampling frequency is reduced by 10% and the waveform sampling length is reduced by 10% for sampling; when the battery power is in the range of 80% - 50%, and the charging current is less than 50 mA, and the wind speed is greater than 0 m / s, the sampling frequency is reduced by 20% and the waveform sampling length is reduced by 20% for sampling; when the battery power is in the range of 50% - 30%, and the charging current is greater than 100 mA, and the wind speed is greater than 0 m / s, the sampling frequency is reduced by 10% and the waveform sampling length is reduced by 10% for sampling; when the battery power is in the range of 50% - 30%, and the charging current is greater than 50 mA and less than 100 mA, and the wind speed is greater than 0 m / s, the sampling frequency is reduced by 30% and the waveform sampling length is reduced by 30% for sampling; when the battery power is in the range of 10% - 30%, and the charging current is greater than 100 mA, and the wind speed is greater than 0 m / s, the sampling frequency is reduced by 30% and the waveform sampling length is reduced by 30% for sampling; when the battery power is in the range of 10% - 30%, and the charging current is greater than 50 mA and less than 100 mA, and the wind speed is greater than 0 m / s, the sampling frequency is reduced by 40% and the waveform sampling length is reduced by 40% for sampling; when the battery power is in the range of 10% - 20%, and the wind speed is greater than 5 m / s, sampling is triggered, and the sampling frequency is reduced by 40% and the waveform sampling length is reduced by 40% for sampling; when the battery power is less than 10%, when the wind speed is greater than 5 m / s, and the included angle between the wind direction and the conductor is greater than 30°, sampling is triggered, and the sampling frequency is reduced by 40% and the waveform sampling length is reduced by 40% for sampling.

[0222] In the above embodiments, the descriptions of each embodiment have their own focuses. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0223] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A breeze vibration synchronous monitoring system based on split deployment, characterized in that: include: A vibration sensor, wherein a plurality of the vibration sensors are respectively arranged on the three-phase conductors of the power line, and the vibration sensor comprises a separate vibration sensing unit respectively arranged at both ends of the tension clamp outlet, and is used to collect acceleration data at both ends of the tension clamp outlet; A meteorological sensor unit, wherein the meteorological sensor unit is used to collect real-time wind speed data and real-time wind direction data; An edge intelligent terminal, the edge intelligent terminal is used to obtain multiple groups of acceleration data and perform differential operations respectively to obtain corresponding acceleration continuous time signals; based on the acceleration continuous time signals, adaptive integral error removal is performed in combination with data characteristics to obtain corresponding displacement signals respectively; combined with a leakage compensation model, power compensation is performed on the displacement signal to determine the corresponding breeze vibration amplitude; Real-time wind speed data and real-time wind direction data are acquired, and the breeze vibration amplitude is calibrated in combination with a preset breeze vibration correlation model.

2. The breeze vibration synchronous monitoring system based on split deployment as claimed in claim 1 is characterized in that: Based on the acceleration continuous time signal, adaptive integral error removal is performed in combination with data characteristics to obtain corresponding displacement signals, including: Calculating a standard deviation and an average value of the acceleration continuous-time signal, and preprocessing the acceleration continuous-time signal according to the velocity continuous-time signal, the standard deviation and the average value; Performing debiasing and integration operations on the acceleration continuous time signal after preprocessing in sequence to obtain a velocity time series; performing debiasing and integration operations on the velocity time series in sequence to obtain an initial displacement signal; The Gaussian principal elimination method is used to perform fitting to eliminate the signal trend in the initial displacement signal and obtain the displacement signal; Furthermore, the Gaussian principal elimination method is used to perform fitting to eliminate the trend signal in the initial displacement signal, and the displacement signal is obtained including: Construct a polynomial trend model of the initial displacement signal, fit the polynomial trend model by Gaussian column principal elimination method, and obtain a trend signal sequence; The displacement signal is determined based on the difference between the initial displacement signal and the trend signal sequence.

3. The breeze vibration synchronous monitoring system based on split deployment as claimed in claim 1 is characterized in that: The method of combining the leakage compensation model to perform power compensation on the displacement signal and determining the corresponding breeze vibration amplitude includes: Performing Fourier transform on the displacement signal to determine the frequency and amplitude corresponding to the displacement signal; The amplitude is compensated based on the frequency and the preset leakage compensation model, and the corresponding breeze vibration amplitude is determined according to the main frequency point and the leakage point frequency; Furthermore, constructing the leakage compensation model includes: According to the frequency offset between the main frequency point and the adjacent leakage point frequency, a vibration amplitude attenuation coefficient polynomial model is constructed; According to the energy distribution of the main frequency point and the leakage point, a leakage compensation model is constructed; the leakage compensation model is combined with the vibration amplitude attenuation coefficient polynomial model, and the leakage compensation factor is determined by using the frequency offset and the leakage compensation model is updated; Furthermore, the leakage compensation model is expressed as: A real =A s *P correced (f0) / P0; In the formula, A real represents the breeze vibration amplitude, A s represents the amplitude, f0 represents the main frequency, P0 represents the main frequency power, Δf i represents the ith frequency offset, k i represents the i-th leakage compensation factor.

4. The breeze vibration synchronous monitoring system based on split deployment as claimed in claim 1 is characterized in that: The breeze vibration correlation model is constructed specifically by utilizing the wind speed, the angle between the wind direction and the power line and the amplitude of the breeze vibration response to construct the breeze vibration correlation model.

5. The breeze vibration synchronous monitoring system based on split deployment as claimed in claim 1, characterized in that: The edge intelligent terminal is also used to obtain the power information data of the separate vibration sensing unit, process the power information data, real-time wind direction data and real-time wind speed data through a preset wind speed and wind direction power-sampling model, obtain the sampling frequency and waveform sampling duration parameters and transmit them to the separate vibration sensing unit; Furthermore, the power information data, real-time wind direction data and real-time wind speed data are processed through the preset wind speed, wind direction and power-sampling model to obtain the sampling frequency and waveform sampling duration parameters: based on the comparison results of the battery power, charging current and the preset threshold range, combined with the comparison results of the wind speed data and wind direction data with the preset threshold, the corresponding sampling frequency and waveform sampling duration parameters are output.

6. A method for synchronously monitoring breeze vibration based on split deployment, characterized in that: include: Acquire multiple groups of acceleration data and perform differential operations respectively to obtain corresponding acceleration continuous time signals; wherein the acceleration data is collected by separate vibration sensing units arranged at both ends of the tension clamp outlet; Based on the acceleration continuous time signal, adaptive integral error removal is performed in combination with data characteristics to obtain corresponding displacement signals respectively; combined with the leakage compensation model, power compensation is performed on the displacement signal to determine the corresponding breeze vibration amplitude; Real-time wind speed data and real-time wind direction data are acquired, and the breeze vibration amplitude is calibrated in combination with a preset breeze vibration correlation model; wherein the real-time wind speed data and the real-time wind direction data are collected by a meteorological sensor unit.

7. The method for synchronously monitoring breeze vibration based on split deployment according to claim 6, characterized in that: Based on the acceleration continuous time signal, adaptive integral error removal is performed in combination with data characteristics to obtain corresponding displacement signals, including: Calculating a standard deviation and an average value of the acceleration continuous-time signal, and preprocessing the acceleration continuous-time signal according to the velocity continuous-time signal, the standard deviation and the average value; Performing debiasing and integration operations on the acceleration continuous time signal after preprocessing in sequence to obtain a velocity time series; performing debiasing and integration operations on the velocity time series in sequence to obtain an initial displacement signal; The Gaussian principal elimination method is used to perform fitting to eliminate the signal trend in the initial displacement signal and obtain the displacement signal; Furthermore, the Gaussian principal elimination method is used to perform fitting to eliminate the trend signal in the initial displacement signal, and the displacement signal is obtained including: Construct a polynomial trend model of the initial displacement signal, fit the polynomial trend model by Gaussian column principal elimination method, and obtain a trend signal sequence; The displacement signal is determined based on the difference between the initial displacement signal and the trend signal sequence.

8. The method for synchronously monitoring breeze vibration based on split deployment according to claim 6, characterized in that: The method of combining the leakage compensation model to perform power compensation on the displacement signal and determining the corresponding breeze vibration amplitude includes: Performing Fourier transform on the displacement signal to determine the frequency and amplitude corresponding to the displacement signal; The amplitude is compensated based on the frequency and the preset leakage compensation model to determine the corresponding breeze vibration amplitude; Furthermore, constructing the leakage compensation model includes: According to the frequency offset between the main frequency point and the adjacent leakage point frequency, a vibration amplitude attenuation coefficient polynomial model is constructed; According to the energy distribution of the main frequency point and the leakage point, a leakage compensation model is constructed; the leakage compensation model is combined with the vibration amplitude attenuation coefficient polynomial model, and the leakage compensation factor is determined by using the frequency offset and the leakage compensation model is updated; Furthermore, the leakage compensation model is expressed as: A real =A s *P correced (f0) / P0; In the formula, A real represents the breeze vibration amplitude, a s represents the amplitude, f0 represents the main frequency, P0 represents the main frequency power, Δf i represents the ith frequency offset, k i represents the i-th leakage compensation factor.

9. The method for synchronously monitoring breeze vibration based on split deployment according to claim 6, characterized in that: The breeze vibration correlation model is constructed specifically by utilizing the wind speed, the angle between the wind direction and the power line and the amplitude of the breeze vibration response to construct the breeze vibration correlation model.

10. The method for synchronously monitoring breeze vibration based on split deployment according to claim 6, characterized in that: Also includes: Obtaining the power information data of the separate vibration sensing unit, processing the power information data, real-time wind direction data and real-time wind speed data through a preset wind speed and wind direction power-sampling model, obtaining the sampling frequency and waveform sampling duration parameters and transmitting them to the separate vibration sensing unit; Furthermore, the power information data, real-time wind direction data and real-time wind speed data are processed through the preset wind speed, wind direction and power-sampling model to obtain the sampling frequency and waveform sampling duration parameters: based on the comparison results of the battery power, charging current and the preset threshold range, combined with the comparison results of the wind speed data and wind direction data with the preset threshold, the corresponding sampling frequency and waveform sampling duration parameters are output.

Images