Lightweight Power Cable Material Property Database and Analysis Platform

Abstract

The present invention discloses a lightweight power cable material property database and analysis platform, which relates to the technical field of power cable material property data analysis, and includes a compression configuration initialization module, a mutation time section identification module, a compression impact assessment module, a compression strategy adjustment module, and a strategy adaptive optimization module; by using a behavior recognition method to identify the mutation points existing in the stored fatigue data, and determining the corresponding time range centered on the mutation points, and calibrating it as the mutation time section; analyzing the fatigue data of the lightweight power cable in the mutation time section, and evaluating the influence degree of the current storage compression strategy of the fatigue data on the accuracy of fatigue life prediction. The present invention solves the problem of key information loss in the compression process of lightweight power cable fatigue data in the mutation section, realizes the accurate evaluation and adaptive optimization of the compression strategy, thereby improving the accuracy of life prediction and the intelligent level of data management.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis of the characteristics of power cable materials, and particularly to a lightweight power cable material characteristic database and analysis platform. Background Art

[0002] A lightweight power cable is a cable specifically designed to reduce weight while maintaining or enhancing electrical conductivity, mechanical strength, and environmental resistance. It is widely used in high-end weight-sensitive fields such as aerospace, electric vehicles, high-speed trains, and intelligent manufacturing. Its core lies in the use of new lightweight and high-strength materials, such as aluminum alloy conductors, carbon nanotube conductors, composite polymer insulation materials, etc., to reduce the overall mass and improve energy efficiency and reliability. The characteristic data of lightweight power cable materials refer to the detailed parameters used to describe the electrical (such as conductivity, dielectric constant), mechanical (such as tensile strength, fatigue resistance), thermal (such as high-temperature resistance, thermal expansion coefficient), and environmental adaptability (such as corrosion resistance, moisture resistance) of these cable materials. These data directly determine the application suitability and reliability of the cable. However, due to the significant differences in the performance requirements of cable materials in different application scenarios and the influence of various factors on the material characteristics (such as alloy composition, nanomaterial structure, manufacturing process), the traditional laboratory measurement and empirical material selection methods can no longer meet the modern high-efficiency R & D requirements. Therefore, it is particularly important to establish a lightweight power cable material characteristic database and analysis platform. Through advanced database storage technology, a large amount of material data can be systematically managed, and combined with artificial intelligence, big data analysis, and simulation calculation technologies, in-depth mining, optimal matching, and trend prediction of material performance can be carried out, thereby accelerating the R & D process of new lightweight power cables, improving the accuracy and reliability of material selection, and ultimately promoting the development and industrial application of lightweight power cable technology.

[0003] The existing lightweight power cable material property database and analysis technology mainly rely on a multi-level data storage architecture and an intelligent analysis engine to achieve efficient data management and material optimization. First, at the database storage layer, the platform adopts a hybrid storage architecture that combines relational databases (such as MySQL, PostgreSQL) and NoSQL databases (such as MongoDB, Cassandra). Among them, relational databases are mainly used to store structured basic material information (such as material composition, manufacturing process, supplier data, etc.), while NoSQL databases are used to store massive amounts of unstructured data such as experimental data, simulation results, and material microstructure images. At the same time, the platform also combines distributed storage technologies (such as Hadoop HDFS, Apache Parquet) to support the efficient reading, writing, and management of large-scale data. Second, at the data processing and analysis layer, the platform integrates AI-driven machine learning algorithms (such as random forest, support vector machine SVM, deep neural network DNN), which can learn the influencing factors of material properties from historical data and predict the material properties under specific formulation or process conditions through regression analysis and optimization algorithms. In addition, based on finite element analysis (FEA) and computational materials science methods (such as density functional theory DFT, molecular dynamics simulation MD), the platform can conduct virtual simulation tests of power cables under different working environments to evaluate the stability of materials under conditions such as high temperature, high pressure, and vibration. Finally, at the decision support layer, the platform presents the analysis results through visualization tools (such as Tableau, PowerBI, Matplotlib) and provides an intelligent material selection recommendation system based on multi-objective optimization algorithms to help R & D personnel quickly screen the optimal material combination, reduce experimental costs, and accelerate the R & D and industrial application of new lightweight power cables.

[0004] The existing technology has the following deficiencies:

[0005] In long-term operating high-voltage power equipment, lightweight power cables need to withstand repeated bending, stretching, or compression. These cyclic loads will gradually cause microcracks to form and expand inside the material, ultimately affecting the fatigue life of the cable. To accurately evaluate the fatigue performance of the material, the database needs to store a large amount of fatigue test data for a long time and restore the complete fatigue curve during query. However, due to the large amount of data and limited storage space, the database usually stores the fatigue data in a compressed manner. During the fatigue process, the material properties do not decay uniformly but have mutation points (such as the rapid crack propagation stage). These key points are crucial for accurately predicting the fatigue life. If the database uses a fixed-ratio compression algorithm when compressing the data and fails to identify these mutation points, key fatigue information may be lost during data restoration, resulting in misjudgment of the degradation trend of the material properties. The existing lightweight power cable material characteristic database and analysis technology lack the ability to identify and differentially process the mutation points in the fatigue curve, and cannot ensure the integrity of the key features that support life prediction after the fatigue data is compressed to maintain the accuracy of key data. When engineers query the fatigue curve, the data provided by the database may have an increased error in fatigue life prediction, thus affecting the actual use evaluation of the material. Furthermore, this will lead to an increase in the need for experimental verification. Engineers need to conduct additional experiments to correct the database data, increasing the R & D cost, and ultimately affecting the optimization design and popularization application of new materials, delaying the R & D progress of new lightweight cables.

[0006] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure. Therefore, it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0007] The object of the present invention is to provide a lightweight power cable material characteristic database and analysis platform to solve the problems in the above background art.

[0008] To achieve the above object, the present invention provides the following technical solution: A lightweight power cable material characteristic database and analysis platform, including a compression configuration initialization module, a mutation time section identification module, a compression impact evaluation module, a compression strategy adjustment module, and a strategy adaptive optimization module;

[0009] The compression configuration initialization module analyzes the fatigue behavior characteristics of the lightweight power cable and establishes a corresponding fatigue data storage compression configuration strategy based on the material type, service conditions, and historical fatigue data;

[0010] The mutation time section identification module identifies the mutation points existing in the stored fatigue data through a behavior recognition method, determines the corresponding time range centered on the mutation points, and calibrates it as the mutation time section;

[0011] The compression impact assessment module analyzes the fatigue data of the lightweight power cable within the mutation time segment, and evaluates the impact degree of the current storage compression strategy of the fatigue data on the prediction accuracy of the fatigue life;

[0012] The compression strategy adjustment module adjusts the data compression strategy of the mutation time segment according to the evaluation results for different impact degrees;

[0013] The strategy self-adaptive optimization module combines the fatigue life prediction results and the experimental verification feedback, tracks and traces back the storage strategy of the mutation time segment, adjusts and updates the compression processing logic according to the prediction error, and continuously monitors the change trend of the fatigue performance of the lightweight power cable.

[0014] Preferably, in the mutation time segment identification module, based on the change trend of the stored fatigue data of the lightweight power cable, a behavior identification method including first derivative mutation detection, curvature change analysis, and local fitting residual anomaly identification is adopted to detect the non-linear change points in the fatigue curve to determine the position of the mutation point, and with this mutation point as the center, combined with the data sampling period and the performance change density, a time window with a preset range is extended forward and backward, and the covered time range is calibrated as the mutation time segment.

[0015] Preferably, in the compression impact assessment module, a fatigue life prediction model is constructed based on the fatigue data of the lightweight power cable within the mutation time segment, and life predictions are respectively made for the original fatigue data before compression and the fatigue data after compression reduction to generate life assessment analysis information, and preprocessing is performed after generation; life prediction result information and model fitting modulus information are extracted from the preprocessed life assessment analysis information, and analysis is performed after extraction to respectively generate a prediction offset sensitivity index and a compression reduction error index; a compression impact assessment model is constructed for the generated prediction offset sensitivity index and compression reduction error index, and a compression impact coefficient is generated through weighted summation; a preset compression impact coefficient threshold interval is determined, and after determination, it is compared with the generated compression impact coefficient, and the impact degree of the current storage compression strategy of the fatigue data on the prediction accuracy of the fatigue life is evaluated according to the comparison result.

[0016] Preferably, the acquisition logic of the prediction offset sensitivity index is as follows:

[0017] Extract the life prediction result information from the preprocessed life assessment analysis information, specifically including the fatigue life prediction values generated by using the original fatigue data before compression at different moments within the mutation time segment as the input of the fatigue life prediction model and the fatigue life prediction values generated by using the fatigue data after compression reduction as the input of the fatigue life prediction model, and respectively label them as and , Indicates within the mutation time segment The fatigue life prediction value generated by using the original fatigue data before time compression as the input of the fatigue life prediction model, Indicates within the mutation time segment The fatigue life prediction value generated by using the fatigue data after decompression and restoration as the input of the fatigue life prediction model, , is a positive integer;

[0018] Calculate the prediction offset sensitivity index, and the specific calculation formula is as follows:

[0019]

[0020] In the formula, is the prediction offset sensitivity index.

[0021] Preferably, the acquisition logic of the decompression restoration error index is as follows:

[0022] Extract the model fitting modulus information from the preprocessed life evaluation and analysis information, specifically including the equivalent modulus values recorded in the original fatigue data before compression at different times within the mutation time segment, the equivalent modulus values recorded in the fatigue data after decompression and restoration, and the theoretical modulus values obtained by fitting the life prediction model, and label them as , and , Indicates within the mutation time segment The equivalent modulus value recorded in the original fatigue data before compression at time Indicates within the mutation time segment The equivalent modulus value recorded in the fatigue data after decompression and restoration at time Indicates within the mutation time segment The theoretical modulus value obtained by fitting the life prediction model at time , is a positive integer;

[0023] Calculate the decompression restoration error index, and the specific calculation formula is as follows:

[0024]

[0025] In the formula, is the decompression restoration error index, represents the maximum value between at time and 1, represents the maximum value between at time and 1.

[0026] Preferably, for the generated prediction offset sensitivity index and the compression and restoration error index Construct a compression impact evaluation model, and generate a compression impact coefficient through weighted summation. The specific calculation formula is as follows:

[0027]

[0028] In the formula, is the compression impact coefficient, and are the non-zero weight coefficients of the prediction offset sensitivity index and the compression and restoration error index respectively, and .

[0029] Preferably, determine a preset compression impact coefficient threshold interval , and after determination, compare it with the generated compression impact coefficient to evaluate the influence degree of the current fatigue data storage compression strategy on the fatigue life prediction accuracy according to the comparison result. The specific comparison and analysis are as follows:

[0030] If , the influence degree of the current fatigue data storage compression strategy on the fatigue life prediction accuracy is a low influence degree;

[0031] If , the influence degree of the current fatigue data storage compression strategy on the fatigue life prediction accuracy is a medium influence degree;

[0032] If , the influence degree of the current fatigue data storage compression strategy on the fatigue life prediction accuracy is a high influence degree.

[0033] Preferably, in the compression strategy adjustment module, according to the evaluation result, adjust the data compression strategy for the mutation time section for different influence degrees, specifically including:

[0034] If the evaluation result is a low influence degree, the adjustment of the data compression strategy for the mutation time section is specifically: keep the current compression strategy unchanged and maintain the original compression parameter configuration;

[0035] If the evaluation result is a medium influence degree, the adjustment of the data compression strategy for the mutation time section is specifically: reduce the compression ratio in the current compression parameters to within the preset reduction range of the original value, and enable the redundant data verification mechanism to enhance the restoration accuracy;

[0036] If the evaluation result is of high impact level, the specific adjustments to the data compression strategy for the mutation time segment are as follows: replace it with a precision - priority compression algorithm, enable the lossless compression mode, and turn off the automatic inheritance function of compression parameters.

[0037] In the above technical solution, the technical effects and advantages provided by the present invention are as follows:

[0038] 1. By introducing a mutation time segment recognition module and using behavior recognition methods such as first - order derivative mutation detection, curvature change analysis, and local fitting residual recognition, the present invention can accurately locate performance mutation points in fatigue data and construct a dynamic time window for calibration, thereby realizing the automatic recognition of non - linear behavior characteristics during the service process of lightweight power cables. Compared with the existing technology that relies on fixed rules or average strategies for data compression, this solution can ensure that in the key segment where fatigue performance changes sharply, the data compression strategy can timely identify the risk area and focus on processing it, avoiding the problem of misjudging the life caused by the destruction of key information during the compression process from the source, and greatly improving the intelligence and precision control ability of database compression processing.

[0039] 2. The present invention establishes a dual - evaluation mechanism composed of a prediction deviation sensitivity index and a compression reduction error index, constructs a compression impact evaluation model by combining the output results of the life prediction model and the physical modulus fidelity, and generates a compression impact coefficient through weighted calculation. This technical path can not only evaluate the impact of the compression strategy on the accuracy of life prediction from the dual dimensions of the "prediction layer" and the "physical layer", but also realize the quantitative comparison of the effects of different compression schemes, enabling the system to clearly judge the adaptability level of the current strategy. By setting the threshold interval of the compression impact coefficient, the strategy impact level is divided into three levels: low, medium, and high, and corresponding compression adjustment actions are triggered, realizing the differential and hierarchical control of the fatigue data compression strategy, thereby effectively improving the credibility of the model output and the robustness of the compression strategy.

[0040] 3. The present invention further introduces a strategy adaptive optimization module. Based on the error comparison result between the fatigue life prediction result and the experimental verification feedback, combined with the execution effect tracking and backtracking of the historical compression strategy in the mutation segment, it realizes the automatic optimization and adjustment of compression parameters and algorithms. The platform continuously monitors the change trend of the fatigue performance of cable materials during the service process, dynamically senses the change of material behavior patterns, and accordingly self - updates the compression strategy, effectively avoiding the problem of adaptation failure caused by the long - term static state of the strategy. This module establishes a closed - loop mechanism of "evaluation - feedback - update", enabling the system to have the ability of long - term self - evolution and strategy adaptation, greatly enhancing the intelligent level, data fidelity ability, and engineering application expansibility of the lightweight power cable material property database and analysis platform. Brief Description of the Drawings

[0041] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0042] Figure 1 It is a schematic diagram of the modules of the lightweight power cable material characteristic database and analysis platform of the present invention. Specific embodiments

[0043] Now, the exemplary embodiments will be described more comprehensively with reference to the drawings. However, the exemplary embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these exemplary embodiments are provided so that the description of the present disclosure will be more complete and thorough, and will fully convey the concept of the exemplary embodiments to those skilled in the art.

[0044] The present invention provides a Figure 1 lightweight power cable material characteristic database and analysis platform as shown, including a compression configuration initialization module, a mutation time section identification module, a compression impact assessment module, a compression strategy adjustment module, and a strategy self-adaptive optimization module;

[0045] The compression configuration initialization module analyzes the fatigue behavior characteristics of the lightweight power cable, and based on the material type, service conditions, and historical fatigue data, establishes a corresponding fatigue data storage compression configuration strategy to initialize the database compression processing conditions of the fatigue data matching the fatigue characteristics of this type of material;

[0046] In order to analyze the fatigue behavior characteristics of the lightweight power cable and establish a matching compression configuration strategy, the platform can extract multi-dimensional characteristic data including material type, mechanical properties, service environment parameters (such as temperature, humidity, load type, stress cycle frequency), and past fatigue curves by calling the material database and historical fatigue experiment data interface, and construct a material fatigue behavior classifier based on a rule model and a machine learning model. This classifier can classify the fatigue curves of different materials into types such as slow decay type, mutation expansion type, multi-segment turning type, etc., and accordingly preset data compression configuration parameters corresponding to each type of fatigue behavior, such as the upper limit of the compression ratio, the mutation detection threshold, the sampling frequency range, etc. According to the analysis results, the platform generates a corresponding compression configuration file for the target material in the database and automatically binds it to the subsequent data writing process to guide the compression processing behavior of the fatigue data during storage, thereby realizing personalized compression initialization driven by fatigue characteristics.

[0047] Due to the complex coupling effect of the fatigue behavior of lightweight power cables on material properties and operating conditions, there are significant differences in the performance decay modes of different types of materials during fatigue. If a unified compression configuration strategy is adopted, it will be difficult to balance storage efficiency and fatigue life prediction accuracy simultaneously. Especially when there are abrupt behaviors such as crack propagation, the fixed compression algorithm is prone to cause key data loss, thereby affecting subsequent life modeling and material evaluation. By actively analyzing the fatigue behavior characteristics based on materials and historical data during the system initialization phase and establishing compression configuration parameters that match them, a "compression adaptability model" can be constructed in advance, enabling the database to have differential processing capabilities before receiving new data. This configuration strategy not only improves the system's adaptability to various materials but also provides a basic framework for subsequent mutation point identification, compression impact assessment, and dynamic strategy optimization, ensuring that the intelligent compression logic of the entire system has prior pertinence and scalability.

[0048] The mutation time segment identification module identifies the mutation points existing in the stored fatigue data through a behavior identification method, determines the corresponding time range centered on the mutation points, and calibrates it as the mutation time segment.

[0049] In this embodiment, in the mutation time segment identification module, based on the change trend of the stored fatigue data of the lightweight power cable, a behavior identification method including first derivative mutation detection, curvature change analysis, and local fitting residual anomaly identification is used to detect the non-linear change points existing in the fatigue curve to determine the position of the mutation points. Centered on the mutation points, combined with the data sampling period and performance change density, a time window with a preset range is extended forward and backward, and the covered time range is calibrated as the mutation time segment for subsequent differential processing of the compression strategy and key data retention control.

[0050] In the mutation time segment recognition module, the platform first preprocesses the stored lightweight power cable fatigue data, including outlier removal, sampling frequency unification, and trend smoothing operations, to ensure that the data quality meets the analysis requirements. Subsequently, the platform constructs a fatigue curve with time or the number of cycles as the horizontal axis and performance indicators (such as stress, strain, stiffness, or conductivity) as the vertical axis, and sequentially traverses the entire curve region in a sliding window manner. In each sliding window, the software system sequentially executes multi-dimensional behavior recognition algorithms, including local derivative calculation, curvature estimation, and fitting error analysis of the data, to extract the change characteristic indicators at each position. The platform compares the characteristic indicators at each position by setting a group of adaptive thresholds. When there is a sudden jump in the first derivative, a sharp increase in curvature change, or a significant increase in the fitting residual at a certain position, the system marks this position as a potential mutation point. Subsequently, the platform further conducts stability verification to eliminate false mutations caused by local noise interference, and finally determines the accurate position of the mutation point in the fatigue data, providing a data basis for the subsequent construction of the mutation time segment and the evaluation of the compression strategy.

[0051] "First derivative mutation detection" refers to continuously calculating the first derivative (i.e., the performance change rate) of the fatigue curve in the time or cycle dimension and observing whether there are sudden jumps, reversals, or sharp slope changes in its local range. Such mutations usually indicate that the material performance begins to degrade rapidly or enters the non-linear crack growth stage. "Curvature change analysis" uses the second derivative of the curve (or based on the geometric curvature formula) to capture the bending degree of the data shape. By calculating the local curvature value at each point and comparing it with the mean value of the surrounding segments, it is judged whether there is a sharp deformation inflection point. The appearance of high curvature points in the fatigue curve usually represents material instability or a mutation in the damage mechanism. "Local fitting residual anomaly recognition" refers to using methods such as polynomial fitting and spline interpolation to perform trend modeling on local data segments and calculating the residual between the original data and the fitting curve. When the residual is significantly greater than the normal fluctuation range in a certain window, it indicates that there is an abnormal deviation from the trend in this area, which is usually related to sudden damage or a sudden change in material performance. The three complement each other, which can effectively improve the accuracy and anti-interference ability of mutation point detection, especially suitable for dealing with the non-linear degradation characteristics of lightweight power cable materials in complex fatigue environments.

[0052] After the mutation point is identified, the platform uses this mutation point as the center of the section and automatically constructs a mutation time section in software by combining the sampling period of the data and the density of performance changes. When specifically implemented, the platform first calculates the number of data points per unit time based on the data sampling frequency recorded in the database to determine the minimum scalable granularity in the time dimension. Then, the platform dynamically evaluates the change intensity of this area based on the performance change rates before and after this mutation point (such as strain increment, modulus decline rate, residual stress fluctuation amplitude), and determines the data lengths that need to be extended outward before and after the mutation according to the set change threshold. The platform forms an extended time section that not only includes the mutation point itself but also completely covers the data characteristics before and after the mutation by setting an initial time window (such as a fixed time value or the number of cycles) and then adjusting the window by superimposing the results of the change density analysis. Finally, the platform marks the data range within this time window as the "mutation time section" and stores it in the metadata table together with the position of the mutation point as the key object for subsequent compression strategy evaluation and storage control.

[0053] During the service process of lightweight power cables, the degradation of their fatigue performance often does not occur isolatedly at a certain point, but forms a continuous change process before and after the mutation point. For example, the transition of the crack initiation - propagation - stable stage, and these stages often show a continuous but accelerating change trend. If only the mutation point is identified without expanding its influence section, it is easy to simplify or discard the key data before and after the mutation point during the compression process, thus damaging the overall understanding of the mutation stage behavior by the fatigue life prediction model. By intelligently expanding the mutation section by combining the sampling period and the density of performance changes, not only can the key behavior areas of fatigue degradation be more accurately enveloped, but also the accuracy of subsequent compression evaluation and strategy allocation can be improved, ensuring that the database can specifically protect these "high-value data areas" during the compression process. This method enables the platform to have the ability to "shift from point identification to section modeling", enhances the prediction robustness of the system and the controllability of the compression strategy, and is an essential part of realizing intelligent compression processing.

[0054] The compression impact assessment module analyzes the fatigue data of the lightweight power cable within the mutation time section and assesses the impact degree of the current storage compression strategy of the fatigue data on the accuracy of fatigue life prediction;

[0055] In this embodiment, in the compression impact assessment module, a fatigue life prediction model is constructed based on the fatigue data of the lightweight power cable within the mutation time section. The fatigue life predictions are respectively made for the original fatigue data before compression and the decompressed fatigue data, and life assessment analysis information is generated. After generation, preprocessing is performed. Life prediction result information and model fitting modulus information are extracted from the preprocessed life assessment analysis information, and after extraction, analysis is carried out to respectively generate a prediction offset sensitivity index and a decompression reduction error index. A compression impact assessment model is constructed for the generated prediction offset sensitivity index and decompression reduction error index, and a compression impact coefficient is generated through weighted summation. A preset compression impact coefficient threshold interval is determined, and after determination, it is compared with the generated compression impact coefficient. According to the comparison result, the influence degree of the current fatigue data storage compression strategy on the accuracy of fatigue life prediction is evaluated.

[0056] In the compression impact assessment module, the platform first constructs a fatigue life prediction model based on the fatigue data of the lightweight power cable within the mutation time section. This process can call the historical fatigue experiment database through software, extract structural features and performance indicators including stress amplitude, strain amplitude, number of cycles, equivalent modulus, etc., and combine the actual life results collected in the experiment as labels. A regression algorithm (such as polynomial regression, SN curve fitting, Coffin-Manson model extension algorithm, or curve fitting based on neural network) is used to establish a fatigue life prediction model. After the model is constructed, the platform takes the original fatigue data before compression and the decompressed data as inputs respectively. The input dimensions include the same type of index features used for model training mentioned above. Through the model prediction process, the corresponding fatigue life prediction values or life decay trends are respectively output for subsequent difference evaluation. This process can be automatically executed by the modeling engine module of the platform, and the batch output of prediction results is realized in combination with the model parameter visualization interface.

[0057] Since the performance degradation behavior of the lightweight power cable within the mutation time section has a significant impact on its fatigue life, if the compressed data loses key information, it will directly interfere with the accuracy of life prediction. To determine whether the current compression strategy weakens the life prediction ability, it is necessary to independently predict the data before and after compression through the prediction model and compare the deviation or stability between the prediction results. Through this process, the "potential prediction error caused by compression" can be transformed into a "quantifiable evaluation index", providing a direct basis for subsequent compression impact assessment. This method integrates the life prediction model into the compression assessment logic, realizing the evaluation upgrade from "data accuracy" to "functional impact", and has higher technical value and practical significance.

[0058] The purpose of preprocessing after generating life assessment analysis information is to eliminate data noise, standardize data dimensions, enhance the stability and comparability of parameter calculations, and ensure that the two exponential parameters in subsequent calculations have a consistent data basis under different compression states. The preprocessing process can be automatically completed by software and mainly includes the following steps: First, perform normalization processing to uniformly map all indicators (such as modulus, stress, life value, etc.) to a standard scale (such as 0 to 1) to avoid affecting the calculation results due to differences in numerical dimensions; Second, perform outlier detection and correction to eliminate local sampling outliers based on statistical outlier detection or the moving window variance method; Third, perform interpolation or resampling processing to ensure that the time points of the data before and after compression are aligned and avoid error accumulation caused by inconsistent time dimensions; Fourth, smooth the model fitting residuals to reduce the impact of jump points caused by data compression in the prediction curve. These processes can be completed through the built-in data cleaning module of the platform and are automatically triggered before model evaluation.

[0059] After the life assessment analysis information has been preprocessed, the platform can structurally extract various contents in this information through the built-in data parsing and label separation module. Specifically, the platform classifies and identifies the life assessment analysis information by field or data label, distinguishing the life prediction value sequence and the fitting modulus sequence generated by inputting the data before compression and the decompressed data into the fatigue life prediction model respectively. For the life prediction result information, the platform recognizes it as the direct output result of the model and extracts the fatigue life prediction value corresponding to each moment (such as remaining life, total life, or life decay rate, etc.) according to the time node; for the model fitting modulus information, the platform calls the internal state data or fitting parameters of the model to output and extracts the elastic modulus value fitted to the corresponding fatigue state, usually from the intermediate prediction variables or residual regression terms in the model fitting process. The extraction process uses a field mapping and data path identification mechanism to support the unified parsing of structured data and semi-structured model outputs, ensuring that the life prediction values and modulus values are in one-to-one correspondence and are automatically associated with the subsequent exponential calculation logic. The entire extraction process can be achieved through the automatic data label matching and field shunting rules of the software without manual intervention.

[0060] To determine the pre-set threshold interval of the compression impact coefficient, the platform can construct multiple groups of compression impact coefficient samples based on the historical fatigue experiment data set and the life prediction error performance under various compression strategies, and perform interval division in combination with the supervised classification method. The specific implementation method is that the platform first performs simulated compression and life prediction on the existing data under different material types, load conditions, and compression algorithms through the software system, calculates the corresponding prediction offset sensitivity index and compression reduction error index, and generates the compression impact coefficient of the historical samples through the weighted model. Subsequently, the platform combines the actual life error level of each group of samples and labels them as slightly, moderately, or severely affected according to the actual engineering impact degree. Based on this labeled data set, the platform uses algorithms such as clustering analysis, decision tree segmentation, or quantile division to automatically learn the distribution law of the compression impact coefficient and determine the coefficient value interval ranges corresponding to slight impact, moderate impact, and severe impact. This process can periodically perform model self-training in the background, enabling the threshold interval to have the ability of adaptive update to ensure effective discrimination ability under the evolution of different materials and compression strategies. The entire threshold interval setting logic can be implemented through data-driven software algorithms without manual setting.

[0061] In this embodiment, the acquisition logic of the prediction offset sensitivity index is as follows:

[0062] Extract the life prediction result information from the pre-processed life assessment and analysis information, specifically including the fatigue life prediction values generated by using the original fatigue data before compression at different times within the mutation time section as the input of the fatigue life prediction model and the fatigue life prediction values generated by using the fatigue data after compression reduction as the input of the fatigue life prediction model, and label them respectively as and , represents the fatigue life prediction value generated by using the original fatigue data before compression at the th moment within the mutation time section as the input of the fatigue life prediction model, represents the fatigue life prediction value generated by using the fatigue data after compression reduction at the th moment within the mutation time section as the input of the fatigue life prediction model, , is a positive integer;

[0063] To achieve real-time acquisition of the fatigue life prediction values generated at different time points for the original fatigue data before compression and the fatigue data after decompression within the mutation time segment, the platform can, through its built-in data tracking and model invocation mechanism, synchronously transmit the fatigue data to the life prediction model module when it flows into the database. The specific implementation method is as follows: When the platform detects that the fatigue data in the mutation time segment is calibrated and archived, the system will automatically call the original fatigue data before compression and extract the corresponding data segments at each moment at a set time interval as model inputs to generate the life prediction value at the m-th moment in real time; meanwhile, the platform will retrieve the data that has been compressed and decompressed at the same time point as the second set of model inputs to generate the corresponding decompression prediction value. These two types of life prediction values are uniformly output through the life prediction model module, saved in a structured manner, and automatically associated according to the time stamp. Among them, the life prediction value refers to the remaining life, total life, or fatigue damage degree of the material calculated by the model after inputting the fatigue performance parameters (such as equivalent modulus, stress amplitude, strain response, etc.) at the current moment, and can be represented in the standard unit of "number of cycles" or "service time". Through the data listener, model API scheduler, and prediction cache mechanism, the platform can ensure that and the real-time acquisition and recording are consistent and responsive, thus providing a continuously updated data basis for the subsequent calculation of the prediction offset sensitivity index.

[0064] Calculate the prediction offset sensitivity index, and the specific calculation formula is as follows:

[0065]

[0066] In the formula, is the prediction offset sensitivity index.

[0067] Adopting the calculation formula of this prediction offset sensitivity index is aimed at finely quantifying the overall deviation degree between the fatigue life prediction results before and after compression, so as to evaluate the direct impact of the compression strategy on the accuracy of life prediction. Among them, the numerator part represents the absolute difference between the life prediction values generated by the data before compression and the decompressed data at the -th moment, reflecting the perturbation intensity of the compression behavior on the single-point life prediction result; the denominator plays a role in normalization, avoiding the relative influence of errors being masked when the life value is large, and adding 1 to prevent the denominator from being zero; the overall square of this ratio can amplify the influence of points with larger deviation values on the overall evaluation result and improve the sensitivity to high-risk compression errors; taking the average of all moments can measure the extensiveness and persistence of the compression influence and reflect its overall interference level on the entire mutation time segment; the outermost natural logarithm operation is used to compress the exponential fluctuation range, making The value is numerically smoother and converges while retaining the ability to respond to drastic changes. Through this calculation method, the platform can obtain a quantitative index of compression impact that combines local sensitivity and overall stability.

[0068] Prediction offset sensitivity index The numerical size directly reflects the overall consistency of the fatigue life prediction results before and after compression in the mutation time section, and there is a positive correlation between it and "evaluating the impact degree of the current storage compression strategy of fatigue data on the accuracy of fatigue life prediction". Specifically, when the value is small, it indicates that the prediction results generated after the compressed and restored data and the original data are input into the fatigue life prediction model are highly similar, and the compression process has minimal interference on the accuracy of life prediction, indicating that the current compression strategy has a light impact on the prediction performance; conversely, when the value is large, it indicates that there is an obvious deviation between the prediction results before and after compression, indicating that the compression behavior has lost or distorted key fatigue information in the mutation section, resulting in a significant deviation in life prediction, and it is evaluated that the compression strategy has a heavy impact on the accuracy of life prediction. Therefore, can be used as an important metric for judging the quality and risk level of the compression strategy.

[0069] In this embodiment, the acquisition logic of the compression reduction error index is as follows:

[0070] Extract the model fitting modulus information from the preprocessed life assessment analysis information, specifically including the equivalent modulus values recorded in the original fatigue data before compression at different times in the mutation time section, the equivalent modulus values recorded in the compressed and restored fatigue data, and the theoretical modulus values obtained by fitting through the life prediction model, and respectively calibrate them as 、 and , represents the equivalent modulus value recorded in the original fatigue data before compression at the time in the mutation time section, represents the equivalent modulus value recorded in the compressed and restored fatigue data at the time in the mutation time section, represents the theoretical modulus value obtained by fitting through the life prediction model at the time in the mutation time section, , is a positive integer;

[0071] During the mutation time period, the platform can process the original fatigue data before compression, the decompressed data, and the fitting results of the life prediction model in real time through the fatigue data monitoring and analysis system to obtain two types of key parameters: the equivalent modulus value and the theoretical modulus value. Specifically, the equivalent modulus value refers to the material stiffness index obtained by recording the stress response and strain response at a specified sampling position (such as the stress concentration area of the conductor layer material) at each moment during the fatigue test and calculating the numerical ratio using the stress-strain relationship in the elastic stage. Its unit is usually GPa, which can dynamically reflect the mechanical properties of the material in the current state. The platform can obtain the equivalent modulus value corresponding to the original data before compression by extracting these two original variables in real time through the stress-strain data channel and performing a ratio operation. The equivalent modulus value after decompression is obtained by repeating the above stress-strain calculation logic after processing the compressed data. The platform inputs the decompressed stress and strain sequences into the modulus calculation module to obtain the modulus value corresponding to the original data structure for comparative analysis of its fidelity. The theoretical modulus value is the theoretical calculation result obtained by fitting the material fatigue behavior through the life prediction model built into the platform. During the model training stage, the platform established a life-modulus evolution relationship model (such as the modulus decay function fitted based on the number of cycles, stress amplitude, etc.) using historical fatigue test data. After inputting the relevant working condition parameters (such as the current loading frequency, temperature, load amplitude) within the mutation time period, this model can output the equivalent modulus value that the material should theoretically have at each moment, serving as a reference value for evaluating whether the decompressed data deviates from the physical evolution trend. Through the multi-channel data synchronization mechanism and model inference engine of the platform, the above three types of data can be dynamically obtained and structured archived at each moment within the mutation time period for subsequent calculation of the error index and strategy evaluation.

[0072] Calculate the decompression reduction error index, and the specific calculation formula is as follows:

[0073]

[0074] In the formula, is the decompression reduction error index, represents the maximum value between at the moment and 1, represents the maximum value between at the moment and 1.

[0075] This decompression reduction error index 's calculation formula combines two types of key modulus difference calculation structures, aiming to comprehensively evaluate the dual impacts of the decompression reduction process on the physical consistency of fatigue data and the credibility of model fitting within the mutation time period. The first term Represents the compression reduction modulus The modulus fitted to the life prediction model The relative square error between them, which is used to measure whether the restored data deviates from the mechanical response trend expected by the material fatigue behavior model. The square term amplifies large deviation values, making them occupy a higher weight in the overall average, while the denominator uses the maximum value function To ensure that the error is not abnormally amplified when the modulus value is small, enhancing the stability of the model fitness evaluation; the second term of the formula Is used to evaluate the logarithmic relative deviation between the compression reduction modulus and the original physical response modulus Among them, the logarithmic function is used to smooth high-frequency errors, avoiding sudden changes in the overall error index caused by local fluctuations. At the same time, the absolute value operation ensures that all deviations are positive inputs, reflecting the degree of consistency damage of the compression reduction in the original data fidelity. Overall, through the combination of "square difference + logarithmic difference", this formula not only maintains high sensitivity to serious data distortion but also takes into account the stable response ability to continuous errors, enabling the compression reduction error index to comprehensively reflect the comprehensive impact of the compression strategy on the quality of fatigue data.

[0076] Compression reduction error index The numerical value of has a positive correlation with "evaluating the impact degree of the current fatigue data storage compression strategy on the accuracy of fatigue life prediction". When The value is small, indicating that the difference between the equivalent modulus after compression reduction and the original modulus is small, and it is highly consistent with the theoretical modulus fitted to the life prediction model, indicating that the compression process has little impact on the physical integrity of the fatigue data and the model fitness, and the life prediction result maintains a high accuracy, and the compression strategy can be regarded as a low-impact strategy; while when The value is large, it indicates that significant deviations have occurred in the mechanical response level of the compression reduction data, which not only fails to accurately restore the original modulus characteristics but also significantly deviates from the model prediction trend, easily leading to distortion of the life prediction result and a decrease in accuracy, indicating that the current compression strategy has a large interference on the fatigue life prediction and belongs to a high-impact strategy. Therefore,[[]] Can be used as an important basis for quantitatively judging the impact degree of the compression strategy on the reliability of life prediction, providing support for subsequent strategy adjustment and optimization.

[0077] In this embodiment, for the generated prediction offset sensitivity index And the compression reduction error index Construct a compression impact evaluation model, and generate a compression impact coefficient through weighted summation. The specific calculation formula is as follows:

[0078]

[0079] In the formula, is the compression influence coefficient, and are respectively the prediction offset sensitivity index and the compression restoration error index of non - zero weight coefficients, and .

[0080] This compression influence evaluation model is implemented through the parameter fusion module built into the platform. After the system obtains the prediction offset sensitivity index and the compression restoration error index , it performs a linear combination of the two indices according to the preset weighting strategy to generate the compression influence coefficient for comprehensively determining the influence degree of the compression strategy. Among them, the two non - zero weight coefficients and respectively correspond to and 's evaluation contribution degrees, reflecting the degree of emphasis of the system on the two dimensions of "prediction accuracy offset" and "physical and fitting distortion". The weight values are obtained by the platform through regression analysis or cross - validation optimization of a large number of historical samples during the model training stage to ensure that under different material types and working conditions, can accurately reflect the actual influence level of the compression strategy on fatigue life prediction. The platform supports the adaptive adjustment strategy of these two weights, enabling them to be flexibly set according to the sensitive points of the current application scenario (such as the focus on prediction stability or physical consistency), and at the same time ensuring , avoiding distortion or bias from affecting the credibility of the comprehensive index. The entire process is automatically completed by the software system without manual intervention.

[0081] In this embodiment, a preset compression influence coefficient threshold interval is determined and compared with the generated compression influence coefficient after determination. According to the comparison result, the influence degree of the current fatigue data storage compression strategy on the fatigue life prediction accuracy is evaluated. The specific comparison and analysis are as follows:

[0082] If , the influence degree of the current fatigue data storage compression strategy on the fatigue life prediction accuracy is at a low influence level;

[0083] This situation indicates that the fatigue data after compression and restoration highly matches the original data in terms of both physical consistency and model prediction stability. The compression process does not cause the loss or distortion of key information, and the life prediction model can still maintain accurate output when processing the restored data. In this case, it shows that the current compression strategy has good fidelity performance and prediction compatibility, can be regarded as a low-impact strategy, and will not pose an obvious interference to subsequent data analysis, model training, or material evaluation. The platform can continue to use this compression parameter or algorithm for long-term data compression, with high availability and stability.

[0084] If , the influence degree of the current storage compression strategy of fatigue data on the fatigue life prediction accuracy is at a medium level.

[0085] This situation indicates that there are certain differences between the compressed and restored data and the original data in terms of physical response or model fitting, but it has not reached a destructive or unacceptable level. The life prediction results may have a slight deviation, but the overall trend is still of reference significance. This shows that the current compression strategy has achieved a balance between accuracy and compression efficiency, and is suitable for scenarios where the requirement for prediction accuracy is moderate but the storage resources are limited. The platform can flexibly decide whether to locally adjust the compression strategy according to different task objectives, or adopt a differential compression processing scheme to reduce the cumulative impact of medium deviations.

[0086] If , the influence degree of the current storage compression strategy of fatigue data on the fatigue life prediction accuracy is at a high level.

[0087] This situation indicates that significant deviations have occurred in the compressed and restored data in two dimensions: physical modulus restoration and life prediction fitting, resulting in a decrease in model prediction accuracy, which may misjudge the fatigue life trend and bring the risk of distorted evaluation results or even wrong decisions. This high influence degree reflects that the current compression strategy fails to effectively retain key data features in the mutation section and is not suitable as the standard compression method in the material database. The platform should immediately trigger the adjustment mechanism of the compression strategy at this time, trace back the data quality, and reconfigure the compression parameters or switch to a high-precision compression scheme to ensure the reliability of fatigue data and the effectiveness of subsequent analysis.

[0088] The compression strategy adjustment module adjusts the data compression strategy for the mutation time section according to the evaluation results and different influence degrees;

[0089] In this embodiment, in the compression strategy adjustment module, according to the evaluation results, the data compression strategy for the mutation time section is adjusted according to different influence degrees, specifically including:

[0090] If the evaluation result is of low impact level, the specific adjustment to the data compression strategy for the mutated time segment is as follows: Keep the current compression strategy unchanged and maintain the original compression parameter configuration;

[0091] When the platform evaluation result is of low impact level, the system determines that the current compression and restoration strategy has achieved a high degree of consistency in retaining the key features of fatigue data, and the prediction result is accurate and reliable. Therefore, there is no need to modify the existing compression strategy. In this case, the software platform, through the decision-making logic in the compression strategy control module, calls the processing path with "impact level = low", automatically skips the compression parameter modification process, keeps the current compression algorithm unchanged, and locks the current compression parameter configuration file. During the data scheduling process, this module can also label this segment with "compression stable", which serves as the basis for automatic inheritance of future compression strategy templates. This setting helps save computing resources, avoid a decrease in compression efficiency caused by excessive adjustment, and at the same time ensure the compression stability and consistency of the area where the evaluation accuracy has reached the standard.

[0092] If the evaluation result is of medium impact level, the specific adjustment to the data compression strategy for the mutated time segment is as follows: Reduce the compression ratio in the current compression parameters to within the preset reduction range of the original value and enable the redundant data verification mechanism to enhance the restoration accuracy;

[0093] When the compression impact level is determined to be medium, the platform will automatically lower the compression ratio parameter corresponding to this mutated time segment to within the preset safety threshold (such as 80% - 90% of the original value) through the parameter scheduling interface. This range can be pre-configured as a dynamic table in the system. Before updating the strategy, the software platform first calls the predefined medium impact processing template in the "compression strategy template management module" to lower the current compression ratio numerically and synchronously update the strategy mapping table; subsequently, the system enables the "redundant data verification module" to enhance the robustness of data restoration by adding mirror points or periodic error check bits to specific sampling points during the compression process, ensuring that the key segment still has controllable restoration under high compression conditions. The core of this strategy is to improve the prediction accuracy without significantly sacrificing the compression efficiency, which is applicable to typical scenarios where data accuracy and storage pressure need to be balanced.

[0094] If the evaluation result is of high impact level, the specific adjustment to the data compression strategy for the mutated time segment is as follows: Replace it with a precision - priority compression algorithm, enable the lossless compression mode, and turn off the automatic inheritance function of compression parameters to ensure the complete retention of key fatigue information.

[0095] In the case of evaluating a high impact level, the system will determine that the current compression strategy has seriously affected the restoration quality of fatigue data and the prediction reliability, and a strategy replacement at the strong intervention level must be executed. The software platform automatically switches the compression engine to the "accuracy - first mode" in this situation, calls high - precision or lossless compression algorithms (such as point - by - point difference compression, data mirror compression, etc.), and switches the compression method of this mutation section from the standard mode to the lossless compression channel. In the system compression parameter scheduling logic, at the same time, by disabling the "parameter automatic inheritance switch", it prevents the current high - impact compression parameters from being used for adjacent or subsequent data sections to avoid mistransmission and misjudgment. This strategy can maximize the retention of the physical response curve near the key mutation points, ensuring the reliability of the database in high - precision demand scenarios such as fatigue life prediction and crack propagation modeling, and is a protection measure that must be implemented for high - risk data sections.

[0096] The strategy adaptive optimization module, combining the fatigue life prediction results with the experimental verification feedback, tracks and retraces the storage strategy for the mutation time section, adjusts and updates the compression processing logic according to the prediction error, and continuously monitors the change trend of the fatigue performance of the lightweight power cable, enabling the storage strategy to have the ability of adaptive optimization.

[0097] The implementation of the strategy adaptive optimization module relies on the multi - source data fusion and strategy dynamic scheduling mechanism of the platform, mainly including three key steps: First, the life prediction module continuously generates life prediction results for the data in the mutation time section and compares them with the actual experimental verification data to automatically calculate the historical prediction error; this error can be archived as the "strategy execution effect score" for a specific section and bound to the specific compression parameter configuration through the strategy tracking interface. Second, the platform calls the strategy backtracking mechanism according to this score to quickly retrieve the historical sections using the same compression strategy. Once it is found that the prediction errors of multiple sections exceed the set threshold under this strategy, the re - configuration process of the compression processing logic is triggered, and the system will update the parameters in the original compression template in real - time or replace the compression strategy with a version with stronger error - control ability. Third, the fatigue behavior trend analysis module built into the platform is responsible for continuously monitoring the change trend of the fatigue performance of the cable during long - term operation. If it detects mutations in parameters such as the modulus decline rate and the abnormal strain frequency, the system will automatically determine that the service state of the material has changed, and re - plan the storage strategy configuration in combination with the historical strategy and the current performance characteristics to ensure its synchronous update with the dynamic evolution of the material state.

[0098] The settings of the policy adaptive optimization module are designed to address the fundamental limitation of traditional compression policies, namely "static configuration and inability to self-correct". In existing platforms, even if the compression policy is initially set reasonably, if the compression effect and model prediction error are not continuously monitored during actual operation, it may lead to continuous deterioration of local data without the system's awareness, thereby affecting the stability of the life prediction of the entire database and the credibility of data quality. By introducing a policy adaptive optimization mechanism, the platform can establish a closed-loop logic between "prediction accuracy feedback" and "compression policy evolution", not only achieving posterior correction and self-evolution of the compression scheme, but also adapting to the changes in the behavior pattern of lightweight power cables due to fatigue accumulation in the service environment. In addition, incorporating the change trend of fatigue performance into the policy judgment basis can further enhance the system's response ability to sudden material aging behavior, enabling intelligent balance among data quality control, compression efficiency, and prediction accuracy through the coordinated linkage of data compression and model prediction.

[0099] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0100] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired or wireless (such as infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium accessible by the computer or a data storage device such as a server or data center containing one or more collections of available media. The available media can be magnetic media (such as floppy disks, hard disks, magnetic tapes), optical media (such as DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0101] It should be understood that in various embodiments of the present application, the magnitudes of the serial numbers of the above processes do not imply the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0102] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of this application.

[0103] In several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the above-described embodiments are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of devices or units can be in electrical, mechanical or other forms.

[0104] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0105] In addition, the functional units in each embodiment of this application can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0106] As described above, this is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all should be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. Lightweight power cable material characteristic database and analysis platform, characterized in that It includes a compression configuration initialization module, a mutation time segment identification module, a compression impact assessment module, a compression strategy adjustment module, and a strategy adaptive optimization module; The compression configuration initialization module analyzes the fatigue behavior characteristics of the lightweight power cable, and based on the material type, service conditions, and historical fatigue data, establishes a corresponding fatigue data storage compression configuration strategy; The mutation time segment identification module identifies the mutation points existing in the stored fatigue data through a behavior recognition method, determines the corresponding time range centered on the mutation points, and labels it as the mutation time segment; The compression impact assessment module analyzes the fatigue data of the lightweight power cable in the mutation time segment, and evaluates the impact degree of the current fatigue data storage compression strategy on the fatigue life prediction accuracy; In the compression impact assessment module, a fatigue life prediction model is constructed based on the fatigue data of the lightweight power cable in the mutation time segment, and the life predictions are respectively performed on the original fatigue data before compression and the decompressed fatigue data, generating life assessment analysis information, and performing preprocessing after generation; extracting the life prediction result information and the model fitting modulus information from the preprocessed life assessment analysis information, and performing analysis after extraction, respectively generating a prediction offset sensitivity index and a decompression reduction error index; constructing a compression impact assessment model for the generated prediction offset sensitivity index and decompression reduction error index, generating a compression impact coefficient through weighted summation; determining a preset compression impact coefficient threshold interval, and comparing it with the generated compression impact coefficient after determination, and evaluating the impact degree of the current fatigue data storage compression strategy on the fatigue life prediction accuracy according to the comparison result; The acquisition logic of the prediction offset sensitivity index is as follows: Extract the life prediction result information from the pre - processed life assessment analysis information, specifically including the fatigue life prediction values generated by using the original fatigue data before compression at different times within the mutation time section as the input of the fatigue life prediction model and the fatigue life prediction values generated by using the decompressed fatigue data as the input of the fatigue life prediction model, and label them respectively as and , indicating the fatigue life prediction value generated by using the original fatigue data before compression at the time of within the mutation time section as the input of the fatigue life prediction model, indicating the fatigue life prediction value generated by using the decompressed fatigue data at the time of within the mutation time section as the input of the fatigue life prediction model, , is a positive integer; Calculate the prediction offset sensitivity index, and the specific calculation formula is as follows: In the formula, is the prediction offset sensitivity index; The acquisition logic of the decompression reduction error index is as follows: Extract the model fitting modulus information from the preprocessed life assessment analysis information, specifically including the equivalent modulus values recorded in the original fatigue data before compression at different moments within the mutation time segment, the equivalent modulus values recorded in the fatigue data after compression reduction, and the theoretical modulus values obtained by fitting the life prediction model, and calibrate them respectively as , and , represents the equivalent modulus value recorded in the original fatigue data before compression at the moment within the mutation time segment, represents the equivalent modulus value recorded in the fatigue data after compression reduction at the moment within the mutation time segment, represents the theoretical modulus value obtained by fitting the life prediction model at the moment within the mutation time segment, , is a positive integer; Calculate the decompression reduction error index, and the specific calculation formula is as follows: wherein, is the compression restoration error index, represents the maximum value between at the th moment and 1, represents the maximum value between at the th moment and 1; The compression strategy adjustment module adjusts the data compression strategy of the mutation time segment according to the evaluation result for different impact degrees; Specifically including: If the evaluation result is a low impact degree, the adjustment of the data compression strategy for the mutation time segment is specifically: keeping the current compression strategy unchanged and maintaining the original compression parameter configuration; If the evaluation result is a medium impact degree, the adjustment of the data compression strategy for the mutation time segment is specifically: reducing the compression ratio in the current compression parameters to within a preset reduction range of the original value, and enabling a redundant data verification mechanism to enhance the reduction accuracy; If the evaluation result is a high impact degree, the adjustment of the data compression strategy for the mutation time segment is specifically: replacing it with a precision - priority compression algorithm, enabling a lossless compression mode, and turning off the automatic inheritance function of the compression parameters; The strategy adaptive optimization module combines the fatigue life prediction results and the experimental verification feedback, tracks and retraces the storage strategy of the mutation time segment, adjusts and updates the compression processing logic according to the prediction error, and continuously monitors the change trend of the fatigue performance of the lightweight power cable.

2. The lightweight power cable material property database and analysis platform according to claim 1, characterized in that In the mutation time segment recognition module, based on the change trend of the stored lightweight power cable fatigue data, a behavior recognition method including first derivative mutation detection, curvature change analysis, and local fitting residual anomaly recognition is adopted to detect the non-linear change points in the fatigue curve to determine the position of the mutation point. Taking this mutation point as the center, combined with the data sampling period and the performance change density, a time window with a preset range is extended forward and backward, and the covered time range is calibrated as the mutation time segment.

3. The lightweight power cable material characteristic database and analysis platform according to claim 2, wherein Sensitivity index of generated prediction offset and compression reduction error index Build a compression impact evaluation model, and generate a compression impact coefficient through weighted summation. The specific calculation formula is as follows: In the formula, is the compression influence coefficient, and are the non-zero weight coefficients of the prediction offset sensitivity index and the compression restoration error index respectively, and .

4. The lightweight power cable material property database and analysis platform according to claim 3, characterized in that Determine the pre-set threshold range of the compression influence coefficient , is the minimum value of the pre-set threshold range of the compression influence coefficient, is the maximum value of the pre-set threshold range of the compression influence coefficient, and after determination, it is compared with the generated compression influence coefficient to evaluate the influence degree of the current storage compression strategy of fatigue data on the prediction accuracy of fatigue life according to the comparison result. The specific comparison and analysis are as follows: If , the influence degree of the current storage compression strategy of fatigue data on the prediction accuracy of fatigue life is at a low influence level; If , the influence degree of the current storage compression strategy of fatigue data on the prediction accuracy of fatigue life is at a medium level; If , the influence degree of the current storage compression strategy of fatigue data on the prediction accuracy of fatigue life is a high influence degree.

Images