Oil-immersed transformer transportation vibration data processing and protection linkage method

Abstract

The present application relates to the technical field of oil-immersed transformer, and proposes an oil-immersed transformer transportation vibration data processing and protection linkage method, which comprises the following steps: adopting wavelet transform to decompose the preprocessed vibration signal, extracting the approximation coefficient and the detail coefficient, and separating the vibration components from the vehicle excitation and the oil impact according to the target frequency range; reconstructing the separated vibration components, and extracting the characteristic frequency and amplitude through Fourier transform; based on the characteristic frequency and amplitude, combining the coil displacement amount, the core looseness and the bolt pre-tightening force attenuation rate, determining the parameter weight through the analytic hierarchy process, and calculating the comprehensive damage level index; according to the comprehensive damage level index, triggering the graded early warning, and linkage controlling the vehicle suspension stiffness, the vehicle speed or the driving path to avoid the resonance or the oil impact risk; recording the whole cycle monitoring data, combining the transportation route and the vehicle state information, constructing the vibration influence database to carry out the path optimization and the structure improvement analysis.

Description

Technical Field

[0001] This invention relates to the field of oil-immersed transformer technology, and in particular to a method for processing and protecting against vibration data during the transport of oil-immersed transformers. Background Technology

[0002] During road transportation, oil-immersed transformers are inevitably subjected to complex vibrations and impacts caused by factors such as uneven road surfaces and vehicle start-stop cycles. These vibrations can easily lead to internal coil displacement, loosening of core laminations, weakening of bolt preload, and violent sloshing of insulating oil, which in turn can cause insulation damage, structural deformation, and even functional failures, seriously threatening power grid safety.

[0003] In terms of data analysis, existing methods usually rely on simple threshold alarms or overall spectrum analysis of vibration signals. It is difficult to effectively separate vibration components from different sources such as vehicle excitation and oil impact from mixed vibration signals, resulting in inaccurate feature extraction and inability to quantitatively assess the damage degree of specific components.

[0004] In addition, existing protection measures are mostly reactive, that is, an alarm is issued after the vibration exceeds the limit, and manual intervention is required to slow down or check. They lack a real-time linkage mechanism with vehicle control systems (such as suspension and vehicle speed), resulting in delayed warnings and an inability to proactively intervene to avoid risks.

[0005] Therefore, there is an urgent need for a comprehensive method that can achieve all-round condition perception, accurate damage quantification assessment, and real-time linkage protection with transport vehicles to improve the transportation safety and reliability of oil-immersed transformers. Summary of the Invention

[0006] To address the aforementioned shortcomings, the present invention aims to propose a method for the integrated processing and protection of vibration data during the transportation of oil-immersed transformers. This method aims to improve the accuracy of vibration monitoring, risk warning capabilities, and active protection effects during the transportation of oil-immersed transformers, thereby significantly reducing the transportation damage rate and extending the service life of the transformers.

[0007] To achieve this objective, the present invention adopts the following technical solution:

[0008] A method for processing and linking transport vibration data of oil-immersed transformers with protection, the method comprising:

[0009] Step S1: Collect vibration signals using sensors placed on the transformer, and store the vibration signals as time-domain data including timestamps;

[0010] Step S2: Preprocess the vibration signal, including denoising, baseline correction, extraction of valid data segments, and normalization.

[0011] Step S3: Decompose the preprocessed vibration signal using wavelet transform, extract approximation coefficients and detail coefficients, and separate the vibration components from vehicle excitation and oil impact according to the target frequency range;

[0012] Step S4: Reconstruct the separated vibration components and extract the characteristic frequencies and amplitudes through Fourier transform;

[0013] Step S5: Based on the characteristic frequency and amplitude, combined with the coil displacement, core looseness and bolt preload attenuation rate, the weight of each parameter is determined by the analytic hierarchy process (AHP), and the comprehensive damage level index is calculated.

[0014] Step S6: Based on the comprehensive damage level index, trigger a graded warning and control the vehicle suspension stiffness, vehicle speed, or driving path to avoid the risk of resonance or oil shock.

[0015] Step S7: Record the full-cycle monitoring data, and combine it with the transportation route and vehicle status information to construct a vibration impact database for path optimization and structural improvement analysis.

[0016] Preferably, step S3 includes:

[0017] The preprocessed vibration signal is decomposed into N-level wavelet functions using a preset wavelet function to obtain the Nth level approximation coefficients and the 1st to Nth level detail coefficients. The number of decomposition levels N is determined according to the target separation signal frequency range, which includes the low-frequency range corresponding to vehicle excitation and the mid-to-high frequency range corresponding to oil impact.

[0018] The low-frequency components corresponding to the vehicle excitation vibration are extracted from the Nth-level approximation coefficients, and the extracted approximation coefficients are subjected to soft thresholding to remove high-frequency interference, thus obtaining the vehicle excitation signal coefficients.

[0019] Target detail coefficients with frequencies in the mid-to-high frequency range are selected from the detail coefficients, and soft thresholding is applied to the target detail coefficients to remove noise, thereby obtaining the oil impact signal coefficients.

[0020] Based on the vehicle excitation signal coefficient and the oil impact signal coefficient, wavelet inverse transform is performed to obtain the vehicle excitation vibration signal and the oil impact vibration signal.

[0021] Preferably, step S4 includes:

[0022] Fast Fourier transform is performed on the vehicle excitation vibration signal and the oil impact vibration signal respectively to obtain the corresponding spectra of the vehicle excitation vibration signal and the oil impact vibration signal;

[0023] From the spectrum of the vehicle excitation vibration signal, the frequency corresponding to the peak value is extracted as the vehicle excitation characteristic frequency; from the spectrum of the oil impact vibration signal, the frequency corresponding to the peak value is extracted as the oil impact characteristic frequency.

[0024] The absolute values ​​of the peak values ​​are extracted from the time-domain signals of the vehicle excitation vibration signal and the oil impact vibration signal, respectively, as the amplitudes of the vibration components, to obtain the vehicle excitation amplitude and the oil impact amplitude. Alternatively, the amplitude values ​​of the spectral peaks corresponding to the vehicle excitation characteristic frequency and the oil impact characteristic frequency are extracted from the spectrum of the vehicle excitation vibration signal and the spectrum of the oil impact vibration signal, respectively, to obtain the vehicle excitation amplitude and the oil impact amplitude.

[0025] Preferably, step S5 includes:

[0026] Obtain the first measured values ​​and safety thresholds for coil displacement, core looseness, and bolt preload attenuation rate;

[0027] The first measured value is normalized to the interval [0, 1], where 0 represents no damage and 1 represents that the value has reached the fatal damage threshold.

[0028] The weights of coil displacement, core looseness, and bolt preload attenuation rate were determined using the analytic hierarchy process (AHP).

[0029] The Damage Index (DLI) is calculated using the weighted summation formula:

[0030] ;

[0031] in, , , These represent the weights of coil displacement, core looseness, and bolt preload attenuation rate, respectively. , , These represent the normalized values ​​of coil displacement, core looseness, and bolt preload attenuation rate, respectively.

[0032] Based on historical damage data, DLI was divided into five damage levels using the K-means clustering algorithm, and corresponding early warning measures were set for each level.

[0033] By training a vibration pattern library using machine learning algorithms, real-time vibration data is compared with historical fault data to predict potential failure risk types.

[0034] Preferably, step S6 includes:

[0035] When the comprehensive damage level index reaches the preset warning threshold, the audible and visual alarm device is triggered to issue a warning and display the specific risk location and risk level.

[0036] When a resonance risk is identified, the vehicle's mass m and the dangerous frequency are obtained in real time. Calculate the natural frequency of the target to be avoided. ,in One of the following conditions must be met: or ;

[0037] The natural frequency of vertical vibration of the vehicle body It is determined by the suspension stiffness k and the vehicle mass m, and satisfies the following relationship: ,use replace The target suspension stiffness is obtained by reverse calculation. : ;

[0038] Obtain the actual suspension stiffness And calculate the stiffness adjustment amount. , According to the stiffness adjustment amount Adjust the vehicle suspension stiffness.

[0039] Preferably, step S6 includes:

[0040] Calculate the comprehensive index of oil sloshing Satisfies the expression:

[0041] ;

[0042] in, Indicates the oil impact pressure. Indicates the swaying acceleration. Indicates the frequency of shaking. , , These represent the weights of oil impact pressure, sloshing acceleration, and sloshing frequency, respectively, and satisfy the following conditions: , , , These represent the danger thresholds corresponding to oil impact pressure, sloshing acceleration, and sloshing frequency, respectively.

[0043] According to the comprehensive index of oil sloshing The comparison result with the preset threshold is used for hierarchical control. If If the speed is less than or equal to the safety threshold, the current speed and driving path will be maintained.

[0044] like If the speed is greater than the safety threshold and less than or equal to the warning threshold, the current speed will be finely adjusted.

[0045] like If the speed exceeds the warning threshold, speed-path linkage control will be activated: the vehicle speed will be forcibly reduced to a safe speed range, and the absolute value of acceleration will be limited to a preset acceleration threshold until... Falling back to or below the safe state threshold;

[0046] Several alternative routes are planned using GPS, and a comprehensive score is calculated for each route:

[0047] ;

[0048] in, Indicates the road surface safety value. Indicates time efficiency and road surface friendliness. Average road roughness coefficient based on path Average slope The proportion of curves Perform the calculation:

[0049] ;

[0050] in, This represents the maximum acceptable value of the average road surface roughness coefficient. This indicates the maximum acceptable value for the average slope. , and These represent the average road surface roughness coefficient, average slope, and percentage of curves, respectively. The weight, , and These settings are all based on the degree of influence on oil sloshing.

[0051] Based on estimated travel time and shortest travel time The ratio is calculated and satisfies the following relationship: , They represent , and The weighting coefficients, Indicates the longest travel time;

[0052] Choose the route with the highest overall score as the optimal driving route.

[0053] Preferably, step S7 includes:

[0054] Record vibration monitoring data, road surface grade and vehicle driving status information of the transformer throughout the entire cycle from shipment to unloading, and construct a vibration impact database;

[0055] Based on the vibration impact database, the correlation between different road conditions and vibration damage was analyzed. A comprehensive damage index was used based on transportation data. Quantifying component damage, the comprehensive damage index The calculation satisfies the expression:

[0056] ;

[0057] in, Indicates fatigue damage. Indicates structural deformation and damage. Indicates functional failure or damage. This represents the coupling coefficient calibrated experimentally, from which fatigue damage is calculated based on Miner's fatigue accumulation theory. : ,in, Represents the reference acceleration The corresponding baseline fatigue life, Indicates the fatigue index of a material. Indicates the vibration frequency. Indicates the resonant frequency of the component. Indicates vibration acceleration. Indicates time, Represents the frequency sensitivity coefficient;

[0058] Structural deformation damage Satisfying the relation: ,in, Represents the elastic limit acceleration. Indicates the acceleration due to plastic deformation failure. Indicates the reference deformation time;

[0059] Functional failure damage Satisfying the relation: in, Indicates the functional sensitivity coefficient;

[0060] Damage threshold determined based on component material and structure. ,Will By substituting the components into the comprehensive damage model, the vibration tolerance limit of the components is obtained by reverse calculation, and optimization suggestions for improving the transformer structure are generated.

[0061] One of the above technical solutions has the following advantages or beneficial effects:

[0062] This invention acquires vibration signals using sensors placed on a transformer and stores them as time-domain data with timestamps, achieving comprehensive acquisition of raw vibration information and providing a complete data foundation for subsequent processing. By performing denoising, baseline correction, truncation of effective data segments, and normalization on the vibration signals, it improves signal quality, reduces interference, and provides a standardized data source for subsequent feature extraction. By employing wavelet transform to decompose the signal and separating vehicle excitation and oil impact components based on the target frequency range, it accurately separates components from different sources in the mixed vibration signal. Finally, by reconstructing the separated vibration components and extracting features using Fourier transform, it achieves further improvements. By accurately acquiring key vibration characteristics representing the state of each component through the identification of frequency and amplitude, and by using the analytic hierarchy process (AHP) to determine weights and calculate a comprehensive damage level index, the system achieves quantitative assessment and risk level classification of internal transformer damage. Furthermore, by triggering graded early warnings based on damage levels and coordinating the control of suspension stiffness, vehicle speed, and route, the system achieves real-time, proactive, and adaptive protection against transportation vibration risks. Finally, by recording monitoring data throughout the entire process and combining it with transportation routes and vehicle conditions to construct a vibration impact database, the system provides data support and decision-making basis for transportation route optimization and transformer structural improvement. Attached Figure Description

[0063] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0064] Figure 1 This is a flowchart of the method for processing and protecting against vibration data during transportation of oil-immersed transformers provided in an embodiment of the present invention;

[0065] Figure 2 This is a flowchart illustrating the graded early warning and response process of the oil-immersed transformer transportation vibration data processing and protection linkage method provided in this embodiment of the invention.

[0066] Figure 3 This is a flowchart of the resonance risk monitoring and vibration isolation adjustment process of the oil-immersed transformer transportation vibration data processing and protection linkage method provided in the embodiments of the present invention. Detailed Implementation

[0067] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0068] In this invention, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus 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 apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0069] A method for processing and linking transport vibration data of oil-immersed transformers with protection, such as Figure 1 As shown, in a preferred embodiment of the present invention, the method for processing and linking transport vibration data of an oil-immersed transformer includes:

[0070] Step S1: Collect vibration signals using sensors placed on the transformer, and store the vibration signals as time-domain data including timestamps;

[0071] It should be noted that a sensor is a device that can sense physical quantities and convert them into electrical signals; for example, an accelerometer can sense vibration acceleration. Vibration signals are information about changes in physical quantities that reflect the vibration state of an object. Time-domain data refers to a continuous sequence of signal values ​​recorded at a certain sampling frequency on a time axis. Timestamps are used to mark the specific time of data acquisition for subsequent analysis of vibration changes over time. In this step, the sensor's role is to convert the vibration experienced by the transformer during transportation into a processable electrical signal. Time-domain data and timestamps provide a digital representation and temporal location of the vibration signal, laying the foundation for subsequent data processing and analysis.

[0072] Understandably, vibration signal acquisition is the initial step in the entire vibration data processing and protection method, aiming to obtain real-time vibration information of the transformer during transportation. By sensing vibration through sensors and converting it into electrical signals, then storing it as time-domain data, the temporal characteristics of the vibration can be completely recorded, providing raw data support for subsequent signal processing and damage assessment. The accuracy and completeness of vibration signal acquisition directly affect the accuracy and reliability of subsequent steps.

[0073] Specifically, multiple high-precision acceleration sensors can be installed in key parts of the transformer (such as coils, core, and casing). These sensors can detect vibration acceleration in different directions in real time. During transportation, the data acquisition system synchronously collects vibration signals from each sensor according to a set sampling frequency (such as 100Hz, 200Hz, etc.) and adds a precise timestamp to the data at each sampling point. The collected vibration signals are converted from analog to digital and stored in data storage devices in formats such as .txt or .csv to ensure data integrity and traceability. For example, a distributed data acquisition system can be used, connecting multiple sensors to a central data acquisition module. The module performs signal conversion through an analog-to-digital converter and uses an embedded system or computer for data storage and management.

[0074] Step S2: Preprocess the vibration signal, including denoising, baseline correction, extraction of valid data segments, and normalization.

[0075] It should be noted that denoising is the process of removing noise components from a signal to improve signal quality. Baseline correction corrects the baseline drift of the signal, restoring it to near the true zero point. Extracting the effective data segment refers to extracting the valid portion containing useful information from the original data. Normalization scales the data proportionally to make it fall within a specific range (e.g., [-1,1]). In this step, denoising reduces the impact of external interference on the vibration signal, baseline correction ensures signal accuracy, extracting the effective data segment removes invalid data, and normalization facilitates subsequent algorithm processing and comparison between different signals.

[0076] It is understandable that vibration signals may be affected by external interference (such as electromagnetic interference, environmental noise, etc.) and sensor characteristics (such as drift, nonlinearity, etc.) during the acquisition process, leading to a decrease in signal quality. Preprocessing steps can improve the signal-to-noise ratio, correct the signal baseline, extract effective data segments containing useful information, and normalize the data to a uniform range. This provides high-quality, standardized data for subsequent wavelet decomposition and feature extraction, improving the accuracy and reliability of the entire vibration data processing method.

[0077] Specifically, the collected vibration signals are first denoised using methods such as mean filtering, median filtering, and wavelet denoising. For example, when using mean filtering, a window size (e.g., 5-10 sampling points) can be set to perform a moving average on the signal to smooth it and reduce interference from high-frequency random noise. Next, baseline correction is performed. If baseline drift exists, the mean of the signal can be calculated and subtracted to restore the baseline to near zero. Then, valid data segments are extracted, and the vibration characteristics during transportation are analyzed to determine the time range of vehicle start-up and stop phases, excluding these and retaining only the vibration signals from the stable transportation phase as valid data segments. Finally, normalization is performed, mapping the vibration acceleration value of each sampling point in the extracted valid data segments to the range [-1, 1], so that subsequent wavelet decomposition and feature extraction algorithms can process the data more effectively.

[0078] Step S3: Decompose the preprocessed vibration signal using wavelet transform, extract approximation coefficients and detail coefficients, and separate the vibration components from vehicle excitation and oil impact according to the target frequency range;

[0079] It should be noted that wavelet transform is a time-frequency analysis method that can decompose a signal into sub-signals of different frequencies and time scales. Approximation coefficients represent the low-frequency components of the signal, reflecting its overall trend; detail coefficients represent the high-frequency components, reflecting local variations. Vehicle-excited vibration refers to transformer vibration caused by the movement of transport vehicles, typically with a low frequency range (e.g., 0.5-10Hz); oil impact vibration refers to the impact vibration generated by the sloshing of oil inside the transformer on its components, typically with a high frequency range (e.g., 50-500Hz). In this step, the wavelet transform decomposes the preprocessed vibration signal into sub-signals with different frequency components. By extracting approximation and detail coefficients, the two different vibration sources—vehicle excitation and oil impact—can be separated, providing a foundation for subsequent vibration feature extraction.

[0080] Understandably, the preprocessed vibration signal contains mixed information from various vibration components, including vehicle excitation and oil impact. Wavelet transform decomposition can expand the signal at different frequencies and time scales, separating the vibration components from different sources. Vehicle-excited vibration has a lower frequency, primarily reflecting the overall impact of vehicle movement on the transformer; oil impact vibration has a higher frequency, primarily reflecting the impact of internal oil sloshing on components. Separating these two vibration components allows for a more accurate analysis and assessment of the transformer's vibration state during transportation, providing a targeted basis for subsequent damage assessment and protective control.

[0081] For example, assuming the preprocessed vibration signal sampling frequency is 200Hz, the target vehicle excitation vibration frequency range is 0.5-10Hz, and the oil impact vibration frequency range is 50-500Hz, the db4 wavelet function is selected, and the decomposition level N=5 is calculated according to the frequency resolution formula. The vibration signal is decomposed into 5 db4 wavelets, yielding the 5th-level approximation coefficient A5 and the 1st to 5th-level detail coefficients D1-D5. The low-frequency components of the vehicle excitation vibration are extracted from A5 and subjected to soft thresholding to obtain the vehicle excitation signal coefficient A5'; the target detail coefficients with frequencies between 50-500Hz are selected from D3-D5 and subjected to soft thresholding to obtain the oil impact signal coefficients D3'-D5'. Inverse wavelet transforms are then performed on A5' and D3'-D5' respectively to obtain the vehicle excitation vibration signal and the oil impact vibration signal, achieving effective separation of the mixed vibration signals.

[0082] Step S4: Reconstruct the separated vibration components and extract the characteristic frequencies and amplitudes through Fourier transform;

[0083] It should be noted that reconstruction refers to recombining the coefficients of the separated vibration components into a complete signal using inverse wavelet transform. Fourier transform is a mathematical transformation method that converts a time-domain signal into a frequency-domain signal, revealing the different frequency components and their amplitudes within the signal. Characteristic frequencies are frequencies in the vibration signal where energy is concentrated or have significant characteristics, reflecting the main frequency characteristics of the vibration; amplitude refers to the magnitude of the vibration signal at the characteristic frequency, reflecting the intensity of the vibration. In this step, the purpose of reconstruction is to restore the separated vehicle excitation and oil impact vibration components into complete vibration signals for frequency domain analysis; Fourier transform is then used to extract characteristic frequencies and amplitudes from the reconstructed signal, providing key characteristic parameters for subsequent damage assessment.

[0084] Understandably, the coefficients of the vehicle excitation and oil impact vibration signals obtained after wavelet decomposition, although separating vibration components from different sources, exist in the form of wavelet domain coefficients, making direct frequency domain analysis and feature extraction difficult. Through inverse wavelet transform reconstruction, these coefficients can be restored to time-domain signals, allowing them to be processed by Fourier transform. Fourier transform converts time-domain signals into frequency-domain signals, clearly displaying the amplitude of each frequency component, thereby extracting characteristic frequencies and amplitudes. These characteristic parameters reflect the main frequency characteristics and intensity of the vibration signal, serving as important criteria for assessing the degree of transformer vibration damage.

[0085] For example, the vehicle excitation signal coefficients A5' obtained in step S3 are subjected to inverse wavelet transform to obtain the vehicle excitation vibration signal. Perform inverse wavelet transform on the oil impact signal coefficients D3'-D5' to obtain the oil impact vibration signal. Then, regarding and Perform Fast Fourier Transform on each to obtain the spectrum. (f) and (f). In In (f), a peak with a large amplitude was found at f=5Hz, which was identified as the vehicle excitation characteristic frequency. =5Hz; in In (f), a peak with a relatively large amplitude was found to occur at f=120Hz, which was identified as the characteristic frequency of oil impact. =120Hz. Meanwhile, in the time domain signal... In the middle, the absolute value of the peak value is taken. =0.8 m / s²; in the spectrum diagram In (f), the amplitude of the spectral peak at f=5Hz is taken as =0.8m / s². Similarly, in the time-domain signal x_impact(t), the absolute value of the peak value is taken as 0.8m / s². =1.2 m / s²; in the spectrum diagram In (f), the amplitude of the spectral peak at f=120Hz is taken as =1.2m / s².

[0086] Step S5: Based on the characteristic frequency and amplitude, combined with the coil displacement, core looseness and bolt preload attenuation rate, the weight of each parameter is determined by the analytic hierarchy process (AHP), and the comprehensive damage level index is calculated.

[0087] It should be noted that coil displacement refers to the degree of positional shift of the transformer coil under vibration; core looseness refers to the degree of loosening of the core laminations under vibration; and bolt preload attenuation rate refers to the proportion of reduction in bolt preload during vibration. The Analytic Hierarchy Process (AHP) is a method for determining the weights of multiple indicators. By constructing a judgment matrix and performing consistency checks, the relative importance weights of each indicator are determined. The comprehensive damage level index is a quantitative indicator of the degree of transformer vibration damage, comprehensively considering the influence of multiple damage parameters. In this step, characteristic frequency and amplitude reflect the characteristics and intensity of vibration, while coil displacement, core looseness, and bolt preload attenuation rate represent the damage manifestations of different internal components of the transformer under vibration. Determining the weights of these parameters through the AHP can reasonably reflect their contribution to the overall damage level of the transformer; the comprehensive damage level index integrates multiple damage parameters into a unified quantitative indicator, facilitating a comprehensive assessment and early warning of transformer damage.

[0088] It is understandable that during the transportation of a transformer, different components will be affected by vibrations of different degrees, resulting in different damage manifestations. The characteristic frequency and amplitude can reflect the characteristics and intensity of the vibration, while the coil displacement, core looseness, and bolt pre-tightening force attenuation rate are the direct manifestations of the damage caused by these vibrations to the internal components of the transformer. By using the analytic hierarchy process to determine the weights of each damage parameter, their relative importance to the overall damage degree of the transformer can be quantified, thus constructing a comprehensive damage level index. This index can comprehensively evaluate the damage degree of the transformer during transportation, providing a scientific basis for subsequent hierarchical early warning and protection control.

[0089] Step S6: According to the comprehensive damage level index, trigger hierarchical early warning and联动 control the vehicle suspension stiffness, vehicle speed, or driving path to avoid resonance or oil fluid impact risks;

[0090] It should be noted that hierarchical early warning is a mechanism that triggers corresponding early warning measures according to different levels of damage. The vehicle suspension stiffness refers to the ability of the vehicle suspension system to resist deformation, and adjusting the suspension stiffness can change the vibration characteristics of the vehicle. The vehicle speed and driving path are operating parameters of the transport vehicle that affect the vibration degree of the transformer. By controlling the vehicle speed and optimizing the driving path, the vibration impact can be reduced. In this step, the comprehensive damage level index is used to judge the damage degree of the transformer, trigger corresponding-level early warnings, and remind the transport personnel to take measures; at the same time, by联动 controlling the vehicle suspension stiffness, vehicle speed, or driving path, the risks of resonance or oil fluid impact can be actively avoided, reducing the damage to the transformer during transportation.

[0091] It is understandable that when the comprehensive damage level index reaches a certain level, it indicates that the transformer may face a greater risk of vibration damage, and timely measures need to be taken for early warning and protection. Hierarchical early warning can trigger corresponding-level early warning signals and measures according to different damage levels, enabling the transport personnel to timely understand the damage status of the transformer and take corresponding actions. At the same time, through the联动 control with the vehicle system, according to the real-time vibration situation and damage assessment results, automatically adjusting the vehicle suspension stiffness, controlling the vehicle speed, or optimizing the driving path can effectively avoid risks such as resonance and oil fluid impact, reduce the vibration damage of the transformer during transportation, and improve the safety and reliability of transportation.

[0092] In one embodiment, as Figure 2 It should be noted that the "联动" in the original text seems to be a specific term in the relevant context, but it is not a common English word. Here, I have used "联动" directly as required. If there is a more accurate English expression for this specific concept in your actual situation, you can replace it accordingly.As shown, step S6 demonstrates the triggering of an early warning based on the comprehensive damage level index. It begins with real-time data transmission from sensors to the backend. Parameter analysis extracts the risk values ​​of each component, providing accurate data for subsequent early warning and response. This corresponds to step S6, where the comprehensive damage level index is acquired in real-time to determine the risk status. The extracted risk values ​​are compared with a dynamic threshold library to determine the risk level, triggering Level 1, Level 2, and Level 3 warnings respectively. These levels correspond to the warnings triggered when the comprehensive damage level index reaches a preset threshold in step S6. In Level 1 warnings, a mild response is triggered, including text prompts on the terminal indicating the risk location, expedited sampling in the backend, and recording of trend curves. In Level 2 warnings, a moderate response is triggered, such as a yellow audible and visual alarm, push notifications of processing suggestions, asynchronous alarm transfer to the transportation monitoring center, and coordinated protective preparedness. In Level 3 warnings, an emergency response is triggered, such as a red audible and visual alarm, forced pop-up display, activation of the vehicle emergency mode interface, and forced initiation of vehicle emergency intervention. After a Level 1 warning, it is determined whether the risk has escalated. If so, the process jumps to the corresponding higher-level procedure; otherwise, monitoring continues until the risk subsides. Under a Level 2 warning, a safe zone is determined. If the risk is confirmed to be resolved, the data is stored. If the risk remains, the corresponding handling procedure continues. This demonstrates that step S6 continuously and dynamically assesses the risk during transportation and adjusts the warning and handling levels in real time according to the actual situation to ensure accurate control and effective response to the risk.

[0093] Step S7: Record the full-cycle monitoring data, and combine it with the transportation route and vehicle status information to construct a vibration impact database for path optimization and structural improvement analysis.

[0094] It should be noted that the full-cycle monitoring data refers to the vibration data and related information continuously recorded throughout the entire transportation process of the transformer, from its initial shipment to unloading. The transportation route includes information such as the path and road conditions (e.g., road surface grade, gradient, curves). Vehicle status information involves operating parameters such as vehicle speed, acceleration, and suspension stiffness. The vibration impact database is a system for storing and managing this data for subsequent analysis and mining. In this step, the full-cycle monitoring data comprehensively records the vibration experience of the transformer during transportation; the transportation route and vehicle status information provide the environment and conditions under which vibration occurs. The construction of the vibration impact database aims to integrate this data to provide data support for subsequent path optimization and structural improvement analysis.

[0095] Understandably, by recording full-cycle monitoring data of transformers during transportation, along with related transportation routes and vehicle status information, a comprehensive understanding of the vibration environment and conditions experienced by the transformers can be obtained. This data reflects the vibration characteristics and damage under different transportation stages, road conditions, and vehicle operating states. Constructing this data into a vibration impact database facilitates data mining and analysis, thereby identifying high-risk road sections, optimizing transportation route planning, and providing a basis for improving transformer structures. This helps improve the safety and reliability of transformer transportation and reduce transportation damage rates.

[0096] In one embodiment, the present invention, through a data processing and risk prediction module, utilizes a multi-source heterogeneous data fusion algorithm and machine learning model to achieve a vibration signal feature extraction accuracy exceeding 95%, and can predict failure risks, such as heat sink resonance fracture and coil short circuit, up to 10 minutes in advance, thus gaining time for real-time protection. Secondly, the real-time protection and damage control module addresses issues such as resonance risk and excessive oil flow impact by adjusting vehicle suspension stiffness and coordinating vehicle speed and path control, effectively reducing resonance amplitude by 40%-60% and oil sloshing intensity by 30%-50%, significantly improving transformer structural stability and reducing oil leakage risk by over 80%. Furthermore, the transportation optimization and product improvement module, leveraging the transportation process data traceability and optimization module, analyzes large batches of data to identify the impact weight of vibration damage under different road conditions, providing a scientific basis and quantitative indicators for transportation route planning and product structure improvement, such as enhancing heat sink connection strength and optimizing coil fixing methods, achieving a two-way improvement in transportation safety and product reliability, increasing the product's vibration tolerance during transportation by 20%-30%. Overall, the present invention comprehensively enhances risk warning capabilities and active protection effects throughout the entire embodiment, significantly reduces transportation damage rates, and significantly extends the service life of transformers, thus possessing extremely high engineering application value.

[0097] Preferably, step S3 includes:

[0098] The preprocessed vibration signal is decomposed into N-level wavelet functions using a preset wavelet function to obtain the Nth level approximation coefficients and the 1st to Nth level detail coefficients. The number of decomposition levels N is determined according to the target separation signal frequency range, which includes the low-frequency range corresponding to vehicle excitation and the mid-to-high frequency range corresponding to oil impact.

[0099] The low-frequency components corresponding to the vehicle excitation vibration are extracted from the Nth-level approximation coefficients, and the extracted approximation coefficients are subjected to soft thresholding to remove high-frequency interference, thus obtaining the vehicle excitation signal coefficients.

[0100] Target detail coefficients with frequencies in the mid-to-high frequency range are selected from the detail coefficients, and soft thresholding is applied to the target detail coefficients to remove noise, thereby obtaining the oil impact signal coefficients.

[0101] Based on the vehicle excitation signal coefficient and the oil impact signal coefficient, wavelet inverse transform is performed to obtain the vehicle excitation vibration signal and the oil impact vibration signal.

[0102] It should be noted that "wavelet function" refers to a class of basis functions with tight support and regularity, used for time-frequency localization analysis of non-stationary signals, such as Db4 wavelet or Sym5 wavelet. Its function is to separate mixed vibration signals into different frequency components through multi-scale decomposition. "Approximation coefficient" represents the low-frequency part of the signal, corresponding to slow changes or basic excitation components in vibration, such as vehicle vibration. "Detail coefficient" represents the high-frequency part of the signal, corresponding to rapid changes or impact components in vibration, such as oil sloshing. "Soft thresholding" is a wavelet denoising method that suppresses noise interference by setting coefficients below a certain threshold to zero and reducing coefficients above the threshold by the threshold.

[0103] Understandably, this step aims to accurately separate vibration components from different physical sources from a mixed vibration signal. The principle is to utilize the multi-resolution analysis capability of wavelet transform in the time and frequency domain to decompose the signal into different levels of approximations and details according to frequency components. Then, based on the typical frequency characteristics of vehicle excitation and oil impact (such as vehicle excitation concentrated in 0.5-10Hz and oil impact concentrated in 50-500Hz), target coefficients are selected. Soft thresholding is then used to enhance useful components and suppress noise, ultimately reconstructing a pure single-source vibration signal. This provides an accurate and separated data foundation for subsequent feature extraction and damage assessment.

[0104] Specifically, firstly based on the sampling frequency (For example, 1000Hz) and the target frequency range are calculated to decompose the number of layers N. This requires that the upper limit of the approximation coefficient frequency of the Nth layer be ≤10Hz (upper limit of the vehicle excitation frequency band) and the lower limit of the detail coefficient frequency of the 1st layer be ≥500Hz (upper limit of the oil impact frequency band). The calculation formula is: the frequency range of the nth layer approximation coefficient. Detail coefficient frequency range Subsequently, the preprocessed signal x(t) is decomposed into N levels using the db4 wavelet to obtain approximate coefficients. (Low frequency) and detail coefficient to (High frequency); then to Soft thresholding is performed (the threshold is adaptively calculated based on the noise variance), and the vehicle excitation signal coefficient A' is obtained by retaining the ≤10Hz components. N，Simultaneously, the oil impact coefficient is obtained by selecting D3-D5 (corresponding to the 50-500Hz frequency band) from the detail coefficients and performing soft thresholding. The remaining detail coefficients (such as D1-D2 corresponding to high-frequency interference >500Hz) are directly discarded. Finally, A'N and The vehicle excitation vibration signal is reconstructed by performing an inverse wavelet transform. (t) and oil impact vibration signal (t), (t)= , (t)= .

[0105] Preferably, step S4 includes:

[0106] Fast Fourier transform is performed on the vehicle excitation vibration signal and the oil impact vibration signal respectively to obtain the corresponding spectra of the vehicle excitation vibration signal and the oil impact vibration signal;

[0107] From the spectrum of the vehicle excitation vibration signal, the frequency corresponding to the peak value is extracted as the vehicle excitation characteristic frequency; from the spectrum of the oil impact vibration signal, the frequency corresponding to the peak value is extracted as the oil impact characteristic frequency.

[0108] The absolute values ​​of the peak values ​​are extracted from the time-domain signals of the vehicle excitation vibration signal and the oil impact vibration signal, respectively, as the amplitudes of the vibration components, to obtain the vehicle excitation amplitude and the oil impact amplitude. Alternatively, the amplitude values ​​of the spectral peaks corresponding to the vehicle excitation characteristic frequency and the oil impact characteristic frequency are extracted from the spectrum of the vehicle excitation vibration signal and the spectrum of the oil impact vibration signal, respectively, to obtain the vehicle excitation amplitude and the oil impact amplitude.

[0109] It should be noted that "Fast Fourier Transform" is a mathematical algorithm that converts a time-domain signal into a frequency-domain signal, used to analyze the frequency components of a signal. For example, the FFT algorithm can decompose a vibration signal into sinusoidal components of different frequencies. "Spectrum" is the amplitude distribution of a signal in the frequency domain. For example, the amplitude-frequency curve obtained by FFT is used to visually display the intensity of each frequency component. "Characteristic frequency" is the main frequency point where energy is concentrated in the signal. For example, the peak frequency representing the inherent vibration of the suspension system in a vehicle excitation signal. "Amplitude" is a quantitative indicator of vibration intensity. For example, the peak amplitude in the time domain signal or the amplitude corresponding to a specific frequency in the frequency domain is used to evaluate the magnitude of vibration energy and to accurately extract key vibration features from mixed signals.

[0110] Understandably, this step uses Fourier transform to convert the reconstructed vibration signal from the time domain to the frequency domain. Utilizing the advantage that frequency domain analysis makes it easier to identify dominant frequency components, characteristic frequencies and amplitudes representing vehicle excitation and oil impact are extracted from the spectrum. The principle is that vibrations from different sources have different distributions in the frequency domain—vehicle excitation is concentrated in the low-frequency range (e.g., 0-20Hz), while oil impact is distributed in the mid-to-high frequency range (e.g., 20-200Hz). Through spectrum analysis of the separated signals, their respective dominant frequencies can be accurately located. Step S4 solves the problem of feature confusion in mixed signals, providing accurate input parameters for subsequent damage assessment.

[0111] Preferably, step S5 includes:

[0112] Obtain the first measured values ​​and safety thresholds for coil displacement, core looseness, and bolt preload attenuation rate;

[0113] The first measured value is normalized to the interval [0, 1], where 0 represents no damage and 1 represents that the value has reached the fatal damage threshold.

[0114] The weights of coil displacement, core looseness, and bolt preload attenuation rate were determined using the analytic hierarchy process (AHP).

[0115] The Damage Index (DLI) is calculated using the weighted summation formula:

[0116] ;

[0117] in, , , These represent the weights of coil displacement, core looseness, and bolt preload attenuation rate, respectively. , , These represent the normalized values ​​of coil displacement, core looseness, and bolt preload attenuation rate, respectively.

[0118] Based on historical damage data, DLI was divided into five damage levels using the K-means clustering algorithm, and corresponding early warning measures were set for each level.

[0119] By training a vibration pattern library using machine learning algorithms, real-time vibration data is compared with historical fault data to predict potential failure risk types.

[0120] It should be noted that "coil displacement" refers to the offset distance of the transformer's internal coil relative to its standard position, such as a millimeter-level value measured by a displacement sensor. Its function is to assess the risk of insulation damage by quantifying the degree of displacement. "Core looseness" refers to the degree of looseness between the transformer core laminations, such as a decibel value measured and converted by a vibration acceleration sensor. Its function is to reflect the integrity of the core structure. "Bolt preload attenuation rate" refers to the proportion of preload loss of the transformer's fixing bolts under vibration, such as a percentage value monitored by a bolt stress sensor. Its function is to assess the reliability of mechanical connections. "Analytic Hierarchy Process" is a multi-criteria decision-making method that combines qualitative and quantitative methods. Its function is to scientifically determine the weight of each damage parameter by constructing a judgment matrix. "K-means clustering algorithm" is an unsupervised machine learning algorithm. Its function is to automatically classify damage indices into different levels based on data similarity.

[0121] Understandably, by establishing a comprehensive damage assessment model, damage parameters with different physical meanings and dimensions can be uniformly quantified, solving the problem that existing technologies cannot accurately assess the degree of damage to specific components. The principle is to reduce dimensional differences through normalization, determine the contribution weight of each parameter to the overall damage through the analytic hierarchy process, and finally obtain the comprehensive damage index through weighted summation, providing a scientific basis for subsequent graded early warning and protection linkage, and realizing the leap from single-parameter threshold alarm to multi-parameter fusion assessment.

[0122] In one embodiment, raw measurement data, including coil displacement, core looseness, and bolt preload decay rate, are acquired from various sensors installed on the transformer. These measured values ​​are then compared with pre-set safety thresholds and normalized, mapping each parameter value to the [0,1] interval. Next, the analytic hierarchy process (AHP) is used to construct a judgment matrix through expert scoring, calculating the weight coefficients of each parameter. Finally, a weighted summation formula is applied. Calculate the comprehensive damage index; based on historical damage data, use the K-means algorithm to automatically cluster the DLI values ​​into five damage levels, and establish the following level classification criteria: Damage level Level 1 corresponds to the DLI range [0, 0.2), indicating no significant damage, all parameters are within the safe range, and continuous monitoring measures are taken; Damage level Level 2 corresponds to the DLI range [0.2, 0.4), indicating minor damage, a single parameter slightly exceeds the threshold, and targeted inspection measures (such as bolt retightening) are taken; Damage level Level 3 corresponds to the DLI range [0.4, 0.6), indicating moderate damage, two or more parameters exceed the standard, and measures such as suspending transportation and partial repair are taken; Damage level Level 4 corresponds to the DLI range [0.6, 0.8), indicating severe damage, close to the fatal threshold, and measures such as comprehensive overhaul and replacement of damaged parts are taken; Damage level Level 5 corresponds to the DLI range [0.8, 1.0], indicating fatal damage, structural integrity is compromised, and measures such as emergency shutdown and scrap assessment are taken.

[0123] For example, when the monitored coil displacement is 1.5 mm (threshold 2.0 mm), the core looseness is 75 dB (threshold 100 dB), and the bolt preload decay rate is 45% (threshold 60%), the normalized result is... =0.75， =0.75, =0.75; Assuming the weighting coefficient is =0.4, =0.3, =0.3, then

[0124] The vibration is classified as Level 4, and the system will initiate early warning measures for comprehensive overhaul. At the same time, the machine learning module identifies that the current vibration characteristics are 85% similar to the "coil insulation wear" pattern in historical data, predicting that there may be a risk of insulation damage.

[0125] Preferably, step S6 includes:

[0126] When the comprehensive damage level index reaches the preset warning threshold, the audible and visual alarm device is triggered to issue a warning and display the specific risk location and risk level.

[0127] When a resonance risk is identified, the vehicle's mass m and the dangerous frequency are obtained in real time. Calculate the natural frequency of the target to be avoided. ,in One of the following conditions must be met: or ;

[0128] The natural frequency of vertical vibration of the vehicle body It is determined by the suspension stiffness k and the vehicle mass m, and satisfies the following relationship: ,use replace The target suspension stiffness is obtained by reverse calculation. : ;

[0129] Obtain the actual suspension stiffness And calculate the stiffness adjustment amount. , According to the stiffness adjustment amount Adjust the vehicle suspension stiffness.

[0130] Among them, "vehicle mass m" refers to the total mass of the transport vehicle and its load (including the transformer), which is usually provided by real-time measurement from an on-board weight sensor. Its function is to serve as a basic parameter for calculating the natural frequency of the vehicle's vibration system; "dangerous frequency" "This refers to the specific frequencies identified through spectrum analysis that may cause resonance in transformer components; its function is to serve as a target reference for frequency avoidance." "Target natural frequency" "The target value of the system's natural frequency is set to avoid resonance and maintain a safe distance from dangerous frequencies; "suspension stiffness k" refers to the vehicle's suspension system's ability to resist deformation. It is a key parameter that affects the vehicle's vertical vibration frequency, and its adjustment is the core means of achieving active vibration avoidance.

[0131] Understandably, step S6 aims to achieve real-time identification and active suppression of resonance risk. The principle is that when the comprehensive damage index warning indicates that there may be resonance risk, the system calculates the target natural frequency and target suspension stiffness based on the vehicle vibration model (single degree of freedom system) and resonance avoidance principle, through the real-time acquisition of vehicle body mass and dangerous frequency. Then, by adjusting the stiffness of the suspension system, the vibration characteristics of the whole vehicle are actively changed, so that the system avoids the dangerous frequency area, thereby suppressing the growth of resonance amplitude from the source and avoiding the transformer from being damaged faster due to continuous resonance.

[0132] In one embodiment, the system first continuously monitors the Comprehensive Damage Level Index (DLI). When the DLI exceeds a preset threshold for resonance risk (e.g., corresponding to Level 3 moderate damage), an audible and visual alarm is immediately triggered to warn the driver, and the specific risk location (e.g., "Core Resonance Risk Level 3") is displayed on the human-machine interface. Subsequently, the system calls the real-time monitoring module to obtain the current vehicle mass m (measured by a weight sensor) and the dangerous frequency identified by the previous signal processing module. (For example, the peak frequency discovered through FFT analysis of the core vibration signal). Then, based on the resonance avoidance principle, the target natural frequency is set. Requires it to cooperate with Maintain a relative safe distance of 10%. or Then, according to the formula for the natural frequency of a single-degree-of-freedom vibration system... ,Will use By substitution, the required target suspension stiffness can be obtained through inverse solving. Then, the current actual stiffness is read from the suspension control unit via the vehicle bus (such as the CAN bus). And calculate the stiffness adjustment amount. Finally, Sending the data to the active suspension control system to perform stiffness adjustment, another implementation example... Figure 3 As shown, all of these reflect the adjustment of stiffness.

[0133] Preferably, step S6 includes:

[0134] Calculate the comprehensive index of oil sloshing Satisfies the expression:

[0135] ;

[0136] in, Indicates the oil impact pressure. Indicates the swaying acceleration. Indicates the frequency of shaking. , , These represent the weights of oil impact pressure, sloshing acceleration, and sloshing frequency, respectively, and satisfy the following conditions: , , , These represent the danger thresholds corresponding to oil impact pressure, sloshing acceleration, and sloshing frequency, respectively.

[0137] According to the comprehensive index of oil sloshing The comparison result with the preset threshold is used for hierarchical control. If If the speed is less than or equal to the safety threshold, the current speed and driving path will be maintained.

[0138] like If the speed is greater than the safety threshold and less than or equal to the warning threshold, the current speed will be finely adjusted.

[0139] like If the speed exceeds the warning threshold, speed-path linkage control will be activated: the vehicle speed will be forcibly reduced to a safe speed range, and the absolute value of acceleration will be limited to a preset acceleration threshold until... Falling back to or below the safe state threshold;

[0140] Several alternative routes are planned using GPS, and a comprehensive score is calculated for each route:

[0141] ;

[0142] in, Indicates the road surface safety value. Indicates time efficiency and road surface friendliness. Average road roughness coefficient based on path Average slope The proportion of curves Perform the calculation:

[0143] ;

[0144] in, This represents the maximum acceptable value of the average road surface roughness coefficient. This indicates the maximum acceptable value for the average slope. , and These represent the average road surface roughness coefficient, average slope, and percentage of curves, respectively. The weight, , and These settings are all based on the degree of influence on oil sloshing.

[0145] Based on estimated travel time and shortest travel time The ratio is calculated and satisfies the following relationship: , They represent , and The weighting coefficients, Indicates the longest travel time;

[0146] Choose the route with the highest overall score as the optimal driving route.

[0147] It should be noted that the "Comprehensive Oil Sloshing Index" is a dimensionless comprehensive evaluation index used to quantify the degree of danger of oil sloshing. Its function is to integrate multi-source monitoring information (pressure, acceleration, frequency) into a straightforward risk value; "Oil Impact Pressure Q" refers to the impact pressure of the oil on the inner wall of the transformer tank, measured by a pressure sensor, reflecting the energy magnitude of the oil sloshing; "Sloshing Acceleration"... "This refers to the acceleration of oil particles, measured by an accelerometer, reflecting the intensity of oil sloshing; "Sloshing frequency" "This refers to the characteristic frequency of oil sloshing, obtained through spectral analysis, reflecting the periodic characteristics of oil sloshing;" weight , , "These are the contribution coefficients of each parameter, allocated according to their degree of influence on the hazards of oil sloshing, satisfying the normalization condition;" Hazard threshold , , "These are limit values ​​determined based on the transformer's structural strength and insulation requirements, used to normalize monitoring parameters. Road surface friendliness" It directly reflects the influence of road physical properties on oil sloshing. (Road safety value) is a comprehensive safety factor based on historical accident data or road grade assessment, used to reflect the basic safety level of the route; time efficiency Its function is to balance transportation efficiency in decision-making, avoiding the selection of excessively long detours due to over-susceptibility, by comparing the estimated travel time T with the shortest travel time. The ratio relationship ( Quantify transportation efficiency, among which For the longest permitted driving time, It is used to balance the timeliness and safety of transportation.

[0148] Understandably, step S6 aims to achieve quantitative assessment and proactive prevention of oil sloshing risk. Its principle lies in constructing a multi-parameter sloshing index to assess the danger level of oil sloshing in real time, and taking graded response measures based on the assessment results: when the risk is low, maintain the current state; when the risk is moderate, fine-tune the vehicle speed; when the risk is high, activate the linkage control between vehicle speed and path, directly reduce the sloshing intensity by reducing the vehicle speed limit acceleration, and select the driving path most friendly to oil sloshing based on a comprehensive evaluation of multiple factors, thereby suppressing the generation and development of oil sloshing from the source and avoiding accelerated aging of the insulation structure due to continuous impact.

[0149] For example, suppose that currently detected a = 3.5 m / s² =8m / s²), f=4Hz ( =10Hz), weight taken =0.5, =0.3, =0.2, then

[0150] S = 0.5 × (8 / 15) + 0.3 × (3.5 / 8) + 0.2 × (4 / 10) = 0.267 + 0.131 + 0.08 = 0.478, which is a safe state; maintain the current speed and path. If the monitored values ​​change to Q = 12 kPa, a = 6 m / s², f = 8 Hz, then...

[0151] S = 0.5 × (12 / 15) + 0.3 × (6 / 8) + 0.2 × (8 / 10) = 0.4 + 0.225 + 0.16 = 0.785, indicating a warning state. The system will slightly adjust the vehicle speed from 70 km / h to 65 km / h. If the monitored values ​​further deteriorate to Q = 14 kPa, a = 7.5 m / s², f = 9 Hz, then S = 0.928, indicating a dangerous state. The system will forcibly reduce the vehicle speed to 50 km / h, limiting the acceleration to no more than 0.2 m / s², and simultaneously calculate the scores for each alternative path: Path 1 (Highway): =0.8, =0.9, =0.95, =0.875; Route 2 (National Highway): =0.6, =0.7, =0.8, =0.69; Route 3 (Provincial Highway): =0.5, =0.6, =0.7, =0.59, and finally the highest-scoring path 1 was selected as the driving route.

[0152] Preferably, step S7 includes:

[0153] Record vibration monitoring data, road surface grade and vehicle driving status information of the transformer throughout the entire cycle from shipment to unloading, and construct a vibration impact database;

[0154] Based on the vibration impact database, the correlation between different road conditions and vibration damage was analyzed. A comprehensive damage index was used based on transportation data. Quantifying component damage, the comprehensive damage index The calculation satisfies the expression:

[0155] ;

[0156] in, Indicates fatigue damage. Indicates structural deformation and damage. Indicates functional failure or damage. This represents the coupling coefficient calibrated experimentally, from which fatigue damage is calculated based on Miner's fatigue accumulation theory. : ,in, Represents the reference acceleration The corresponding baseline fatigue life, Indicates the fatigue index of a material. Indicates the vibration frequency. Indicates the resonant frequency of the component. Indicates vibration acceleration. Indicates time, Represents the frequency sensitivity coefficient;

[0157] Structural deformation damage Satisfying the relation: ,in, Represents the elastic limit acceleration. Indicates the acceleration due to plastic deformation failure. Indicates the reference deformation time;

[0158] Functional failure damage Satisfying the relation: in, Indicates the functional sensitivity coefficient;

[0159] Damage threshold determined based on component material and structure. ,Will By substituting the components into the comprehensive damage model, the vibration tolerance limit of the components is obtained by reverse calculation, and optimization suggestions for improving the transformer structure are generated.

[0160] The calculation model for the comprehensive damage index D integrates multiple damage mechanisms to accurately quantify the cumulative damage to components. Coupling coefficient. Used to regulate fatigue damage Structural deformation and damage and functional failure damage The nonlinear synergistic effect among these three factors reflects the coupling relationship in which multiple types of damage mutually exacerbate each other. Fatigue damage Based on Miner's fatigue accumulation theory, the baseline fatigue life is calculated. and reference acceleration The material fatigue index serves as a reference standard for the fatigue characteristics of materials. Frequency sensitivity coefficient characterizes the sensitivity of a material to stress levels. The quantized vibration frequency deviates from the component's resonant frequency. The effect of time on either aggravating or mitigating fatigue damage. Structural deformation damage. Through elastic limit acceleration and acceleration of plastic deformation failure To define the elastic and plastic deformation stages of a material, the reference deformation time This provides a timescale standard for deformation accumulation. Functional failure damage. Using an exponential model, the functional sensitivity coefficient The sensitivity of vibration acceleration A, frequency F, and time t to functional degradation effects such as loosening of connectors and seal failure is determined. Ultimately, a preset damage threshold is used. This critical safety boundary allows for the inverse application of the comprehensive damage model to determine the vibration parameter tolerance values ​​of components under extreme damage conditions, thus providing a quantitative design basis for product structure improvement. The aim is to achieve precise quantification and tolerance assessment of transformer transportation vibration damage. Its principle lies in accumulating full-cycle data by constructing a vibration impact database and establishing a comprehensive damage model that integrates multiple damage mechanisms. This model precisely quantifies the cumulative damage degree of components from three dimensions: fatigue accumulation, structural deformation, and functional failure. Based on the damage threshold, the vibration tolerance limit of the component is inferred, providing a data-driven scientific basis for product structure improvement and achieving closed-loop optimization from transportation protection to product design.

[0161] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0162] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for processing and linking transport vibration data of oil-immersed transformers with protection, characterized in that, The method for processing and protecting against vibration data during transportation of oil-immersed transformers includes: Step S1: Collect vibration signals using sensors placed on the transformer, and store the vibration signals as time-domain data including timestamps; Step S2: Preprocess the vibration signal, including denoising, baseline correction, extraction of valid data segments, and normalization. Step S3: Decompose the preprocessed vibration signal using wavelet transform, extract approximation coefficients and detail coefficients, and separate the vibration components from vehicle excitation and oil impact according to the target frequency range; Step S4: Reconstruct the separated vibration components and extract the characteristic frequencies and amplitudes through Fourier transform; Step S5: Based on the characteristic frequency and amplitude, combined with the coil displacement, core looseness and bolt preload attenuation rate, the weight of each parameter is determined by the analytic hierarchy process (AHP), and the comprehensive damage level index is calculated. Step S6: Based on the comprehensive damage level index, trigger a graded warning and control the vehicle suspension stiffness, vehicle speed, or driving path to avoid the risk of resonance or oil shock. Step S7: Record the full-cycle monitoring data, and combine it with the transportation route and vehicle status information to construct a vibration impact database for path optimization and structural improvement analysis; Step S7 includes: Record vibration monitoring data, road surface grade and vehicle driving status information of the transformer throughout the entire cycle from shipment to unloading, and construct a vibration impact database; Based on the vibration impact database, the correlation between different road conditions and vibration damage was analyzed. A comprehensive damage index was used based on transportation data. Quantifying component damage, the comprehensive damage index The calculation satisfies the expression: ； in, Indicates fatigue damage. Indicates structural deformation and damage. Indicates functional failure or damage. This represents the coupling coefficient calibrated experimentally, from which fatigue damage is calculated based on Miner's fatigue accumulation theory. : ,in, Represents the reference acceleration The corresponding baseline fatigue life, Indicates the fatigue index of a material. Indicates the vibration frequency. Indicates the resonant frequency of the component. Indicates vibration acceleration. Indicates time, Represents the frequency sensitivity coefficient; Structural deformation damage Satisfying the relation: ,in, Represents the elastic limit acceleration. Indicates the acceleration due to plastic deformation failure. Indicates the reference deformation time; Functional failure damage Satisfying the relation: in, Indicates the functional sensitivity coefficient; Damage threshold determined based on component material and structure. ,Will By substituting the components into the comprehensive damage model, the vibration tolerance limit of the components is obtained by reverse calculation, and optimization suggestions for improving the transformer structure are generated.

2. The method for processing and protecting against vibration during transportation of oil-immersed transformers according to claim 1, characterized in that, Step S3 includes: The preprocessed vibration signal is decomposed into N-level wavelet functions using a preset wavelet function to obtain the Nth level approximation coefficients and the 1st to Nth level detail coefficients. The number of decomposition levels N is determined according to the target separation signal frequency range, which includes the low-frequency range corresponding to vehicle excitation and the mid-to-high frequency range corresponding to oil impact. The low-frequency components corresponding to the vehicle excitation vibration are extracted from the Nth-level approximation coefficients, and the extracted approximation coefficients are subjected to soft thresholding to remove high-frequency interference, thus obtaining the vehicle excitation signal coefficients. Target detail coefficients with frequencies in the mid-to-high frequency range are selected from the detail coefficients, and soft thresholding is applied to the target detail coefficients to remove noise, thereby obtaining the oil impact signal coefficients. Based on the vehicle excitation signal coefficient and the oil impact signal coefficient, wavelet inverse transform is performed to obtain the vehicle excitation vibration signal and the oil impact vibration signal.

3. The method for processing and protecting against vibration during transportation of oil-immersed transformers according to claim 2, characterized in that, Step S4 includes: Fast Fourier transform is performed on the vehicle excitation vibration signal and the oil impact vibration signal respectively to obtain the corresponding spectra of the vehicle excitation vibration signal and the oil impact vibration signal; From the spectrum of the vehicle excitation vibration signal, the frequency corresponding to the peak value is extracted as the vehicle excitation characteristic frequency; from the spectrum of the oil impact vibration signal, the frequency corresponding to the peak value is extracted as the oil impact characteristic frequency. The absolute values ​​of the peak values ​​are extracted from the time-domain signals of the vehicle excitation vibration signal and the oil impact vibration signal, respectively, as the amplitudes of the vibration components, to obtain the vehicle excitation amplitude and the oil impact amplitude. Alternatively, the amplitude values ​​of the spectral peaks corresponding to the vehicle excitation characteristic frequency and the oil impact characteristic frequency are extracted from the spectrum of the vehicle excitation vibration signal and the spectrum of the oil impact vibration signal, respectively, to obtain the vehicle excitation amplitude and the oil impact amplitude.

4. The method for processing and protecting against vibration during transportation of oil-immersed transformers according to claim 1, characterized in that, Step S5 includes: Obtain the first measured values ​​and safety thresholds for coil displacement, core looseness, and bolt preload attenuation rate; The first measured value is normalized to the interval [0, 1], where 0 represents no damage and 1 represents that the value has reached the fatal damage threshold. The weights of coil displacement, core looseness, and bolt preload attenuation rate were determined using the analytic hierarchy process (AHP). The Damage Index (DLI) is calculated using the weighted summation formula: ； in, , , These represent the weights of coil displacement, core looseness, and bolt preload attenuation rate, respectively. , , These represent the normalized values ​​of coil displacement, core looseness, and bolt preload attenuation rate, respectively. Based on historical damage data, DLI was divided into five damage levels using the K-means clustering algorithm, and corresponding early warning measures were set for each level. By training a vibration pattern library using machine learning algorithms, real-time vibration data is compared with historical fault data to predict potential failure risk types.

5. The method for processing and protecting against vibration during transportation of oil-immersed transformers according to claim 1, characterized in that, Step S6 includes: When the comprehensive damage level index reaches the preset warning threshold, the audible and visual alarm device is triggered to issue a warning and display the specific risk location and risk level. When a resonance risk is identified, the vehicle's mass m and the dangerous frequency are obtained in real time. Calculate the natural frequency of the target to be avoided. ,in One of the following conditions must be met: or ; The natural frequency of vertical vibration of the vehicle body It is determined by the suspension stiffness k and the vehicle mass m, and satisfies the following relationship: ,use replace The target suspension stiffness is obtained by reverse calculation. : ; Obtain the actual suspension stiffness And calculate the stiffness adjustment amount. , According to the stiffness adjustment amount Adjust the vehicle suspension stiffness.

6. The method for processing and protecting against vibration during transportation of oil-immersed transformers according to claim 1, characterized in that, Step S6 includes: Calculate the comprehensive index of oil sloshing Satisfies the expression: ； in, Indicates the oil impact pressure. Indicates the swaying acceleration. Indicates the frequency of shaking. , , These represent the weights of oil impact pressure, sloshing acceleration, and sloshing frequency, respectively, and satisfy the following conditions: , , , These represent the danger thresholds corresponding to oil impact pressure, sloshing acceleration, and sloshing frequency, respectively. According to the comprehensive index of oil sloshing The comparison result with the preset threshold is used for hierarchical control. If If the speed is less than or equal to the safety threshold, the current speed and driving path will be maintained. like If the speed is greater than the safety threshold and less than or equal to the warning threshold, the current speed will be finely adjusted. like If the speed exceeds the warning threshold, speed-path linkage control will be activated: the vehicle speed will be forcibly reduced to a safe speed range, and the absolute value of acceleration will be limited to a preset acceleration threshold until... Falling back to or below the safe state threshold; Several alternative routes are planned using GPS, and a comprehensive score is calculated for each route: ； in, Indicates the road surface safety value. Indicates time efficiency and road surface friendliness. Average road roughness coefficient based on path Average slope The proportion of curves Perform the calculation: ； in, This represents the maximum acceptable value of the average road surface roughness coefficient. This indicates the maximum acceptable value for the average slope. , and These represent the average road surface roughness coefficient, average slope, and percentage of curves, respectively. The weight, , and These settings are all based on the degree of influence on oil sloshing. Based on estimated travel time and shortest travel time The ratio is calculated and satisfies the following relationship: , They represent , and The weighting coefficients, Indicates the longest travel time; Choose the route with the highest overall score as the optimal driving route.

Images