Mobile phone data line performance test and fault prediction method and system

Abstract

The invention discloses a mobile phone data line performance test and fault prediction method and system, and belongs to the technical field of electronic product detection and fault prediction. The power-on resistance, the voltage stability, the charging current response and the data transmission rate of a target data line are collected, and under the conditions of different bending angles, insertion and extraction cycles and environment temperatures, the data transmission rate of the target data line is calculated; acquiring stress response data, and extracting multi-dimensional stress principal component characteristics; the electrical property and stress data are fused, a performance degradation correlation model is established, and coupling modeling of data line internal structure damage and electrical property degradation is achieved; historical use data are input, analogue simulation is carried out through the model, and residual performance indexes are generated; judging whether a fault risk exists or not according to a set threshold value, and outputting early warning information if the fault risk is lower than the threshold value; meanwhile, a fault position and a mode type are inferred in combination with an internal characteristic spectrum of the model; the method has high prediction precision, strong structural adaptability and good expansibility, and is suitable for quality control and intelligent diagnosis of consumer electronic cable products.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electronic product detection and fault prediction, and in particular relates to a mobile phone data line performance test and fault prediction method and system. BACKGROUND

[0002] With the large-scale popularity of smart phones, mobile phone data lines, as one of the daily electronic accessories, play a crucial role in the charging and data transmission process. Due to frequent bending, plugging, pulling and other operations by users, the data line is prone to internal wire breakage, poor plug contact, and hidden faults such as shielding layer peeling off during use. Such faults often manifest as sporadic failures in the early stage, such as intermittent inability to charge, slow transmission rate, and system prompt "non-original accessory", which seriously affects user experience.

[0003] Currently, the detection of mobile phone data lines on the market relies mainly on manual visual inspection or simple resistance and voltage detection, which cannot effectively identify and predict the internal performance degradation trend of the data line, and cannot identify potential fault risks in advance, resulting in a large number of potential hidden danger products flowing into the market or entering the user end, increasing maintenance costs and reducing brand trust.

[0004] Therefore, there is an urgent need for a method that can comprehensively consider key physical indicators during use, combine test data in multiple scenarios, and achieve performance state evaluation and fault risk prediction of mobile phone data lines through modeling and analysis, in order to improve detection accuracy and prolong the service life. SUMMARY

[0005] The purpose of the present application is to provide a mobile phone data line performance test and fault prediction method and system to solve the problems in the background art.

[0006] In order to achieve the above purpose, the present application provides the following technical solution: a mobile phone data line performance test and fault prediction method, comprising:

[0007] S100, collecting multiple sets of electrical performance data of the target mobile phone data line under different test scenarios, including power-on resistance, voltage stability, charging current response and data transmission rate, and constructing an electrical performance parameter set E;

[0008] S200, obtaining a stress response data set S of the mobile phone data line under different bending angles, plug-in times and environmental temperature use conditions;

[0009] S300, establishing a corresponding performance degradation correlation model W according to the electrical performance parameter set E and the stress response data set S, for describing the relationship between the internal structure change of the data line and the electrical performance degradation;

[0010] S400, inputting historical usage data of the target mobile phone data line, performing simulation through the W model, and obtaining a data line remaining performance index P in the current state;

[0011] S500: Based on the set performance threshold Ph, compare the indicator P to determine whether the data line has a fault risk. If P <Ph,则输出预警信息并标记为高风险;

[0012] S600: Output the fault location and fault mode of the data line according to the simulation result, so as to implement performance testing and fault prediction of the target mobile phone data line.

[0013] Preferably, the S100 includes:

[0014] S101, periodically scanning the target mobile phone's data line for electrical characteristics under four typical usage scenarios: constant voltage charging, charging while using, data synchronization, and fast charging with high load, to obtain the transient resistance, voltage fluctuation amplitude, dynamic current response, and data throughput rate under each scenario;

[0015] S102, constructing a four-dimensional electrical performance vector based on each set of original test data, uniformly mapping the electrical performance characteristics under each test scenario into a normalized parameter space, and forming a feature matrix Em that can be distinguished in the time series dimension;

[0016] S103. Perform feature compression and nonlinear decoupling on Em through an embedded variational autoencoder to extract a low-dimensional potential electrical expression factor E.

[0017] Preferably, the S200 includes:

[0018] S201. Construct a three-dimensional coupled loading platform, and use a programmable flexible mechanism to apply different bending angles, plug-in cycles, and constant / variable temperature environment control to the target data cable to achieve simulation of typical stress field conditions.

[0019] S202. Synchronously deploy strain fiber array sensors and micro-thermocouples under each loading combination condition to record the strain distribution, microcrack induction points, and abnormal heat conduction paths of the data cable outer sheath and conductor core in real time, forming a high-resolution stress response matrix Si.

[0020] S203. Use the tensor decomposition method to reduce the dimensionality of all Si, extract the key stress principal component features, and construct a multidimensional stress response data set S.

[0021] Preferably, the S300 includes:

[0022] S301, performing timestamp alignment and feature synchronization on the constructed electrical performance parameter set E and the extracted stress response data set S, and realizing high-dimensional correlation extraction of stress and electrical characteristics through a multi-channel residual fusion network;

[0023] S302, introducing a structural embedding tensor to construct an implicit variable map of the internal structural changes of the data line, and embedding it into the performance degradation correlation model W as a model structure constraint;

[0024] S303. Dynamically train the model W using a time series prediction module based on an LSTM-GRU hybrid unit, so that it can simulate the nonlinear degradation process of electrical performance parameters over time or number of cycles driven by different stress paths.

[0025] Preferably, the S400 includes:

[0026] S401. Collect historical usage data of the target mobile phone data cable, including cumulative plug-in and unplug times, typical bending trajectory, total charging time, ambient temperature distribution, and usage scenario labels, and construct a usage behavior vector sequence Ut.

[0027] S402: Input the usage behavior vector Ut into the trained performance degradation association model W, use the time recursion capability of the model to calculate the electrical performance evolution path within the prediction window, and output a simulation state curve Pt;

[0028] S403: Based on the dynamic change rates of the resistance, current and rate parameters in the simulation results, a weighted fusion index function is constructed, Pt is aggregated and reduced, and the data line residual performance index Po in the current state is generated.

[0029] The present invention also provides a mobile phone data line performance testing and fault prediction system, comprising:

[0030] The data acquisition module collects multiple sets of electrical performance data of the target mobile phone data cable under different test scenarios, including on-resistance, voltage stability, charging current response and data transmission rate, and constructs the electrical performance parameter set E;

[0031] The stress analysis module obtains a stress response data set S of the mobile phone data cable under different bending angles, plug-in and plug-out times, and ambient temperature conditions;

[0032] A performance degradation modeling module, based on the electrical performance parameter set E and the stress response data set S, establishes a corresponding performance degradation correlation model W for describing the relationship between the internal structure change of the data line and the electrical performance degradation;

[0033] An indicator generation module inputs the historical usage data of the target mobile phone data line, performs simulation through the W model, and obtains the data line remaining performance indicator P in the current state;

[0034] The risk warning module compares the index P with the set performance threshold Ph to determine whether the data line has a risk of failure, and outputs a warning information and marks it as high risk if P < Ph.

[0035] The mode recognition module outputs the failure position and failure mode of the data line according to the simulation result, and realizes performance testing and failure prediction of the target mobile phone data line.

[0036] In the above technical solution, the present application provides technical effects and advantages:

[0037] 1. The present application establishes a performance degradation correlation model W for mobile phone data lines by fusing multi-scene electrical performance collection, three-dimensional coupled stress loading, high-dimensional feature extraction and structure prior modeling, realizes a complete mapping path from physical response to electrical performance evolution, and has stronger robustness, predictability and engineering applicability compared with traditional methods based on single index or manual detection.

[0038] 2. The present application combines time recursion simulation and weighted fusion evaluation mechanism to construct a quantifiable residual performance index P, and further cooperates with threshold judgment and deep recognition network to realize intelligent failure warning and mode positioning, which significantly improves the accuracy and automation level of failure diagnosis. This method is not only suitable for quality control and operation and maintenance of consumer electronics products, but also has technical universality and expansion potential for other flexible cable structures. BRIEF DESCRIPTION OF DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can also be obtained by those skilled in the art according to these drawings.

[0040] Figure 1 The method mind map of the present application.

[0041] Figure 2 The system module mind map of the present application. DETAILED DESCRIPTION

[0042] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0043] Embodiment 1, please refer to Figure 1 The performance test and fault prediction method for a mobile phone data line described in the embodiment includes the following steps:

[0044] S100, a plurality of sets of electrical performance data of a target mobile phone data line under different test scenarios are collected, including turn-on resistance, voltage stability, charging current response and data transmission rate, and an electrical performance parameter set E is constructed;

[0045] S200, a stress response data set S of the mobile phone data line under different bending angles, plug-in times and environmental temperature use conditions is obtained;

[0046] S300, a corresponding performance degradation correlation model W is established according to the electrical performance parameter set E and the stress response data set S, for describing the relationship between the internal structure change and the electrical performance degradation of the data line;

[0047] S400, historical use data of the target mobile phone data line is input, and a simulation is performed through the W model to obtain a residual performance index P of the data line under a current state;

[0048] S500, according to a set performance threshold Ph, the index P is compared to determine whether the data line has a fault risk, if P < Ph, an early warning information is output and marked as high risk;

[0049] S600, according to the simulation result, a fault position and a fault mode of the data line are output, and the performance test and fault prediction of the target mobile phone data line are realized.

[0050] In the performance test and fault prediction method for a mobile phone data line provided in the present application, step S100 is the collection and preprocessing stage of electrical performance data, which is a key basic link for subsequent performance degradation modeling and fault prediction. This step realizes high-fidelity data modeling of the running state of the mobile phone data line by constructing a multi-condition and multi-index electrical performance test framework and cooperating with high-dimensional feature expression and compression mechanism. This step mainly includes three sub-steps S101 to S103, which are as follows:

[0051] S101: First, to comprehensively characterize the electrical performance of the target mobile phone data cable under real-world usage conditions, it is proposed to conduct periodic electrical scans under four typical operating conditions:

[0052] (1) Constant voltage charging scenario, that is, using a regulated power supply to perform a standard charging process on the data line;

[0053] (2) Charging and using scenario, simulating the common usage state of users operating their mobile phones while charging;

[0054] (3) Data synchronization scenario, i.e., data transmission process via USB connection to a PC or other devices;

[0055] (4) Fast charging high load scenario, that is, using a fast charging power supply that supports the PD / QC protocol to quickly charge the mobile phone.

[0056] In each scenario, a precision electrical parameter test module is used to periodically scan the data line. It is recommended that each scan period be set between 100ms and 500ms to balance data accuracy and response efficiency. During the scan process, the following core indicator data is collected:

[0057] Transient resistance: reflects the dynamic impedance change of the data line conductor during the power-on process;

[0058] Voltage fluctuation amplitude: measures end-to-end voltage stability and reflects contact quality;

[0059] Dynamic current response: records transient current characteristics under changing charging load;

[0060] Data throughput rate: reflects the data transmission capability of the USB channel.

[0061] The above indicators constitute the original electrical performance test data set, covering electrical response information in multiple dimensions and time nodes.

[0062] S102: To achieve a unified expression of electrical parameters across different scenarios, a method for constructing a four-dimensional electrical performance vector is proposed. Specifically, for each set of raw electrical performance data collected for each scenario, a unified vector representation is adopted, in the form of: ;in, represents the transient resistance, Indicates the voltage fluctuation amplitude, represents the dynamic current response, Represents the data throughput rate. The vectors obtained from all scenarios are labeled, and a normalization mapping function is used to uniformly scale each indicator to the range of [0, 1] to eliminate the interference of indicator dimension differences on subsequent model training.

[0063] Furthermore, the normalized electrical performance vectors collected in each scenario are arranged in chronological order to form a feature matrix Em with time series resolution capability, the structure of which is shown below:

[0064] ; Where m is the total number of scenes (4 in this embodiment), n is the number of time points in each scene, is the four-dimensional electrical energy vector at time n.

[0065] S103: Considering the high dimensionality, high redundancy, and strong correlation of the original feature matrix Em, traditional feature extraction methods (such as PCA or conventional autoencoders) struggle to achieve effective compression while preserving nonlinear relationships. To address this, the present invention introduces an Embedded Variational Autoencoder (E-VAE) structure to perform feature compression and nonlinear decoupling on the feature matrix.

[0066] The E-VAE structure consists of an encoder, a latent space, and a decoder, and has the following characteristics:

[0067] The encoder uses a multi-layer convolutional network to automatically extract the spatiotemporal coupling features in the electrical performance vector at each time step;

[0068] The latent space is approximated by Gaussian distribution to generate statistically significant latent expression vectors;

[0069] The decoder is used to reconstruct the original features to verify the integrity of the compressed information. The training goal is to maximize the reconstruction accuracy and minimize the KL divergence.

[0070] The final output is a set of low-dimensional latent electrical expression factors E. This expression factor retains the dominant variation trend of electrical performance across scenarios, significantly reduces the complexity of model training, and enhances the ability to identify subtle performance degradation features.

[0071] The S100's comprehensive electrical performance acquisition and expression mechanism not only enables systematic modeling of the multi-dimensional electrical performance of mobile phone data cables under typical usage scenarios, but also integrates advanced algorithms such as normalized mapping, high-dimensional feature matrix construction, and variational compression at the model structure level, effectively enhancing the sensitivity and generalization capabilities of subsequent performance degradation models to early anomalies. Compared to existing solutions that rely on single-metric detection or static testing, this solution is significantly more practical, advanced, and scalable.

[0072] In the mobile phone data cable performance testing and fault prediction method proposed in this invention, step S200 primarily constructs a data cable mechanical stress response mechanism under real-world operating conditions. By introducing multi-field coupled loading, advanced sensing technology, and multidimensional data extraction methods, the system acquires key stress information that affects the data cable's lifespan and performance stability, providing accurate structural response data support for subsequent degradation models. This step includes three substeps, S201 to S203, as follows:

[0073] S201: In order to effectively simulate the complex mechanical loads and environmental influences that mobile phone data cables are subjected to during actual use, the present invention proposes to construct a three-dimensional coupled loading platform. The platform consists of the following modules:

[0074] Programmable flexible loading mechanism: Using a multi-axis servo actuator, combined with a flexible robotic arm system, it achieves multi-angle continuous loading control of the target data line in the horizontal, vertical, and spatial bending directions. The bending angle range is designed to be 0°~180°, with a control accuracy better than ±1°;

[0075] Plug-in cycle drive unit: Through the electronically controlled propulsion module and the standard USB interface, it simulates the frequent plug-in and pull-out actions of actual users. The plug-in and pull-out frequency can be set between 0.5Hz and 2Hz, supporting more than 10,000 cycles.

[0076] Environmental temperature control module: This module uses a thermoelectric cooler and PID temperature control system to achieve constant temperature (±0.5°C control accuracy) and variable temperature (rise / fall rate controlled within 2°C / min) environmental simulation. The temperature range is set to -20°C to +60°C, covering everyday use environments.

[0077] This 3D loading platform orchestrates load paths through a central controller, enabling load application in a predetermined sequence and time step, enabling the construction of highly repeatable and controllable typical stress field conditions. This platform can simulate stress-induced degradation processes such as material fatigue, conductor fracture, and sheath damage in data cables caused by frequent bending, excessive plugging and unplugging, and extreme temperature fluctuations.

[0078] S202: To accurately capture the mechanical response of the data line under various loading combinations, the present invention simultaneously deploys a multimodal sensing system during the loading process, specifically including:

[0079] Strain fiber array sensor: A microfiber Bragg grating (FBG) array is evenly distributed on the outer layer of the data cable, with sensing points spaced 5 mm apart, for real-time measurement of local strain field distribution. FBG sensors offer high sensitivity, immunity to electromagnetic interference, and the ability to achieve long-distance synchronous data acquisition, making them suitable for monitoring flexible structures.

[0080] Miniature thermocouple probe: A T-type thermocouple is placed at the junction of the data cable core and the sheath to monitor key thermal response characteristics of the conductor material, such as the heat conduction path and local temperature rise abnormality areas under different ambient temperatures.

[0081] The above-mentioned sensing system uses a synchronous acquisition unit and a clock synchronization protocol to ensure that the strain and temperature signals are fully aligned at each loading moment. Under each set of loading conditions, the system will collect and form a set of high-resolution stress response matrices Si with spatial dimensions, temporal dimensions, and multi-physics coupling properties. Its structure can be expressed as: ; Among them, ε(x,y,t) represents the strain response of each sensor node at a certain moment, and T(x,y,t) represents the corresponding temperature distribution.

[0082] This matrix can effectively reflect the local damage evolution trend, potential microcrack induction points and abnormal heat diffusion paths of the internal structure of the material under a composite stress field, providing a rich source of information for constructing damage perception mechanisms and degradation prediction models.

[0083] S203: Given that the original stress response matrix Si is large in data volume, high in dimensionality, and has information redundancy, using it directly for modeling will result in decreased computational efficiency and the potential for overfitting. Therefore, the present invention proposes dimensionality reduction of all Si data based on a tensor decomposition method.

[0084] Specifically, the higher-order singular value decomposition (HOSVD) algorithm is used to perform principal component extraction on the three-dimensional tensor Si, compressing the multidimensional stress response information into several representative tensor principal component vectors to form a representative stress main feature space.

[0085] After HOSVD processing, a set of low-dimensional key stress principal element features Si′ is obtained, which contains the stress field distribution information that has the greatest impact on the structural performance of the data line under this working condition. The Si′ under all loading combinations is integrated and constructed to form a multidimensional stress response data set S: ;

[0086] Where z is the number of loading condition combinations, and each Si′ corresponds to the main stress characteristic under a loading combination scenario.

[0087] Through the implementation of S200, the present invention can simulate typical stress loading conditions experienced during the use of mobile phone data cables with high precision under nondestructive conditions, and acquire stress response data at the microstructural level through a multimodal sensing mechanism. Combined with high-dimensional tensor dimensionality reduction technology, this ultimately forms a compact expression feature set S that can be used for modeling and analysis. Compared to traditional single-field loading or purely empirical modeling approaches, this method offers greater structural adaptability, anomaly sensitivity, and modeling versatility, and is a key foundation for achieving accurate performance degradation modeling and fault prediction.

[0088] In the proposed method for performance testing and fault prediction of mobile phone data cables, step S300 constructs a performance degradation correlation model W, which serves as the core bridge connecting the nonlinear relationship between electrical performance parameters and structural stress responses. By integrating high-dimensional feature alignment, multimodal data fusion, structural knowledge embedding, and a time series prediction mechanism, the temporal evolution of the electrical performance of data cables under complex stress paths is modeled and predicted. This step includes three substeps, S301 to S303, as follows:

[0089] S301: Because the electrical performance parameter set E and the stress response data set S originate from different subsystems, their sampling frequencies, time dimensions, and structural dimensions are inconsistent. Directly inputting these into the modeling module will result in information coupling failure and training non-convergence. Therefore, the timestamp alignment and feature synchronization of the two data types must be completed first.

[0090] In this embodiment, a sliding window-based time series registration algorithm is used to resample and interpolate the original data of E and S, unify the time axis dimension, and construct a joint input sequence with a frame-to-frame correspondence: Xt=[et∥st], where et represents the electrical performance characteristic vector under the t-th frame, st is the corresponding stress principal component characteristic vector, and ∥ represents the vector splicing operation.

[0091] In order to extract high-dimensional correlation features, the present invention constructs a multi-channel residual fusion network with the following structural features:

[0092] Dual-input branch structure: constructs independent convolution channels for electrical performance vectors and stress vectors respectively to extract local features of each type of data;

[0093] Feature Cross Residual Block: Introduces cross-modal interaction paths through residual connections, enabling information fusion between different feature dimensions and preventing gradient vanishing.

[0094] Multi-scale aggregation module: Integrates multi-granularity information extracted by different convolution kernels to improve the model's sensitivity to changes in temporal patterns.

[0095] The final output is a unified correlation feature tensor Ft∈Rk, which serves as the input for subsequent structural graph embedding and temporal modeling.

[0096] S302: Considering the typical structural damage evolution paths during data cable degradation, such as core fatigue fractures extending outward from the center and sheath aging often starting at the bending stress concentration point, data-driven models alone tend to ignore these a priori patterns. To enhance the model's physical rationality and structural explanatory power, this paper introduces a structural embedding tensor as a constraint mechanism.

[0097] The construction method is as follows:

[0098] According to the spatial distribution of stress response in S200, the response significance indicators at different positions inside the data line are extracted, and a spatial node set V = {v1, v2, ..., vm} is constructed; m is the total number of nodes.

[0099] Construct an adjacency matrix A of the stress propagation path and material coupling relationship between nodes, and construct a spatial graph using graph convolution methods;

[0100] The structural nodes and adjacency relationships are encoded into tensor form SET and fused with the feature tensor Ft as the structural prior constraint input model W.

[0101] The structural embedding tensor-guided model prioritizes characteristic changes in key structural regions when learning degradation trends, making the W model more physically interpretable and predictively reasonable.

[0102] S303: Performance degradation is a complex evolutionary process characterized by nonlinearity, temporal order, and path dependency. To accurately simulate the dynamic attenuation of electrical performance over the lifespan, the present invention proposes training the W model using a temporal prediction module based on a LSTM-GRU hybrid unit.

[0103] The hybrid unit structure has the following advantages:

[0104] LSTM (Long Short-Term Memory) is good at modeling long-term dependencies and is suitable for capturing the degradation trend of devices over hundreds of cycles;

[0105] GRU (Gated Recurrent Unit) has high computational efficiency and fewer parameters, and is more suitable for processing mutation features within a short period;

[0106] By connecting the LSTM and GRU structures in series, the model can take into account both global trend modeling and local anomaly detection capabilities.

[0107] The training input is the fused feature tensor {F1, F2, ..., FT} of the time series, and the output is the predicted electrical performance index {e1, e2, ..., eT} at the corresponding moment. The model loss function uses a joint optimization of the mean square error with a weighted coefficient and the structural consistency loss: ; where Dstruct represents the difference measure function between predicted features and structure priors, which can be cosine similarity distance or KL divergence; represents the mean square error, which is used to measure the deviation between the model prediction value and the true value, and is the main supervision target in training; and α, β represent weight coefficients (positive real numbers) for adjusting the relative importance of the two parts of the loss function: α represents the degree of attention of the model to prediction accuracy; and β represents the learning intensity of the model to structural consistency. Typical settings are: α = 1.0, β = 0.2: the prediction accuracy is dominant, and the structure is auxiliary; or can be dynamically optimized through cross-validation.

[0108] In the mobile phone data line performance test and fault prediction method provided in the application, step S400 aims to simulate and predict the current performance state of the target data line based on the user's real use behavior data, using the trained performance degradation correlation model W, and generate a quantitative indicator reflecting the health degree. This step is based on the W model, and through the time recursion and parameter aggregation mechanism, the residual performance index Po with determinability is output, which provides data support for subsequent risk warning. This step includes three sub-steps S401 to S403, as follows:

[0109] S401: Considering that the performance degradation of the data line is closely related to its actual use, it is difficult to accurately evaluate its life state by static test alone, therefore, a set of use behavior data collection and vector construction mechanism is designed. The collected historical behavior data includes but is not limited to the following key dimensions:

[0110] Cumulative plug-in times: the connection period is recorded by the embedded interface counting module, reflecting the fatigue degree of the contact end;

[0111] Typical bending trajectory: the bending angle and frequency in daily use are recorded by the embedded IMU (inertial measurement unit), and the deformation mode is extracted;

[0112] Total charging time and cycle distribution: the daily charging time and charging behavior law are counted by the port power supply current monitoring module;

[0113] Environmental temperature distribution: the external temperature change is recorded in combination with the environmental sensor to evaluate the influence of thermal cycle loading on performance;

[0114] Use scene label: based on the historical charging environment (such as fast charging, vehicle-mounted, and edge use), a use background classification variable is constructed.

[0115] The above original behavior data is uniformly encoded and time-sequenced, and a multi-dimensional time sequence vector set is constructed, denoted as: ; represents the data line use state vector at time step t, and s is the number of use behavior dimensions.

[0116] To adapt to the model input format, Perform normalization and window sliding processing to form a historical behavior sequence within the time window ; T is the window length, which serves as the input data basis of model W.

[0117] S402: The above constructed usage behavior vector sequence The input is fed into the trained performance degradation association model W, and its time recursion and state transfer capabilities are utilized to complete the simulation of the performance state within the future short-term window.

[0118] The model W is internally combined with the LSTM-GRU hybrid prediction unit (see S303 for details). Under the constraints of the structural prior tensor SET, it recursively generates the predicted values ​​of the key electrical performance of the data line from the current to several moments in the future, forming a simulation state curve Pt. The specific form is: ;in: represents the predicted transient resistance; represents the predicted dynamic current response; The entire simulation path can be considered as the electrical performance evolution trajectory of the target data line over a period of time under the current usage trend.

[0119] To ensure the stability and physical consistency of the simulation output, model W also introduces an internal regularization mechanism and a historical trajectory alignment module during the recursive process to ensure that the output results fluctuate reasonably within the boundary conditions and avoid prediction offset or oscillation.

[0120] S403: After obtaining the simulation state curve Pt, in order to more intuitively reflect the health of the current data line and the potential failure risk, the present invention proposes to construct a weighted fusion performance evaluation function to compress multiple predicted electrical performance parameters into a single numerical indicator: the residual performance indicator Po.

[0121] The construction method is as follows:

[0122] right Perform a first-order difference operation and perform weighted average fusion of the calculation results to obtain the residual performance index Po: the final generated Po value range is [0,1], where a value close to 1 indicates good performance, and a value close to 0 indicates significant performance degradation.

[0123] In the mobile phone data line performance testing and fault prediction method proposed in the present invention, steps S500 and S600 respectively complete the intelligent judgment of the target data line performance status and the fault identification and positioning functions, which are the key technical links for realizing the transformation from prediction results to practical output.

[0124] By combining performance threshold judgment logic with a multi-source information fusion reasoning model, it is not only possible to accurately determine whether the target data line is currently on the verge of potential failure, but also to output the specific fault location and possible failure mode type, providing direct technical support for product early warning, quality traceability, and structural improvement. The following details the specific implementation of S500 and S600:

[0125] S500: After obtaining the remaining performance index P of the current data line (generated in S403), a comparison is performed based on a preset performance safety threshold Ph to determine whether the index is in a critical failure state, thereby triggering a risk warning mechanism.

[0126] In this embodiment, the performance threshold Ph is set based on a large amount of experimental data and engineering tolerance requirements, and is defined as follows:

[0127] Ph∈(0,1), its value reflects the lower limit of the allowable level of electrical performance degradation. If it is lower than this value, fault manifestations such as intermittent charging, data transmission failure or thermal failure may occur.

[0128] The system's early warning logic is as follows:

[0129] If the current remaining performance index P≥Ph, the system considers that the data line is in the normal working range and the status is marked as "safe";

[0130] If P <Ph,则系统识别为存在潜在失效风险,状态标记为“高风险”,并自动输出预警信息W,其内容包括:

[0131] Data line identification information;

[0132] Current predicted remaining performance value P;

[0133] The risk level of the trigger;

[0134] Recommended treatment measures (such as reminders for replacement, temporary load restrictions, inspections at testing stations, etc.).

[0135] In addition, to improve the practicality of the system, the warning information W can be synchronously uploaded to the host system or mobile APP through the local interface (such as screen prompts or LED indicators) or wireless communication modules (such as Bluetooth, WiFi), so that users or maintenance personnel can understand the operating status of the data line in real time.

[0136] S600: After determining that there is a fault risk, step S600 is further implemented to analyze the spatial distribution characteristics and pattern types of the fault, that is, to locate and classify the structural fault of the target data line.

[0137] The technical key of this step is to combine the intermediate variables predicted by model W (such as stress characteristic mapping, thermal response distribution, and electrical parameter mutation points) with the structural prior information SET to construct a positioning mechanism based on rule reasoning and deep feature fusion.

[0138] The specific sub-processes include the following:

[0139] S601: Combine the stress response data collected in S200 with the structural embedding tensor SET in S302 to construct a multidimensional position mapping function: L = floc(SET, ∇Pt); this function represents the gradient change rate of the electrical parameter along time or along the structural node, reflecting the area where local performance degradation is exacerbated. The floc function can be implemented as follows:

[0140] Analyze the local change rate of the resistance and current response curves to identify the "mutation point";

[0141] Bind the time node corresponding to the mutation point to the stress sensor position, and infer the fault location in the physical space;

[0142] Output the position coordinates L, or express it as a structure section number (such as "connector A to cable section B").

[0143] In this way, accurate marking of locations such as "fatigue fracture in the middle of the conductor", "local damage to the sheath" or "abnormal plug contact" can be achieved.

[0144] S602: Based on the differences in characteristic patterns corresponding to different fault manifestations, a fault mode determination mechanism is proposed. The system performs pattern recognition based on the multi-parameter evolution path and SET map output by model W. Common fault modes include but are not limited to:

[0145] Resistance fault (such as partial wire break): manifested as a sudden increase in Rt and a decrease in It;

[0146] Poor contact faults (such as loose plugs): manifest as increased voltage fluctuations and intermittent decreases in data rates;

[0147] Thermal degradation faults (such as insulation degradation caused by overheating): manifest as long-term abnormal thermal response and slow parameter drift;

[0148] Frequent bending fatigue failure: manifested as continuous oscillation of electrical parameters corresponding to stress concentration points.

[0149] In this embodiment, a lightweight fault classification network based on attention mechanism (Fault Attention Network, FAN-Net) is used to compare the known fault template library with the current predicted characteristics, and to complete the rapid matching output of multiple fault modes. The output result is in the form of: M=Mode(Pt,SET)→“contact failure” / “fracture trend” / …; The cooperation of S500 and S600 makes the present application not only have performance prediction ability, but also have practical decision value and engineering adaptability. Unlike existing methods that only provide numerical detection or rough failure judgment, the method described in the present application can achieve the following innovative effects:

[0150] The numerical prediction result is converted into an explicit “high risk” / “safe” state, which facilitates rapid response; Through high-resolution stress mapping and model-embedded structural knowledge, physical-level positioning is achieved; Combined with deep learning and prior rule library, different degradation mechanisms can be accurately distinguished; It can be linked with the back-end maintenance platform or terminal equipment to support prompt, recall or restriction strategies.

[0151] In summary, S500 and S600 are the key steps for the present application to transition from a data-driven model to intelligent decision-making and application deployment, significantly improving the engineering feasibility and technical advancement of the present application.

[0152] Embodiment 2, please refer to Figure 2 As shown in the figure, the mobile phone data line performance test and fault prediction system described in this embodiment comprises:

[0153] The data acquisition module acquires multiple sets of electrical performance data of the target mobile phone data line under different test scenarios, including power-on resistance, voltage stability, charging current response, and data transmission rate, to construct an electrical performance parameter set E;

[0154] The stress analysis module obtains a stress response data set S of the mobile phone data line under different bending angles, plug-in times, and environmental temperature use conditions;

[0155] The performance degradation modeling module establishes a corresponding performance degradation correlation model W according to the electrical performance parameter set E and the stress response data set S, which is used to describe the relationship between the internal structure change and the electrical performance degradation of the data line;

[0156] The index generation module inputs the historical use data of the target mobile phone data line, simulates through the W model, and obtains the remaining performance index P of the data line under the current state;

[0157] The risk warning module compares the index P with the set performance threshold Ph to determine whether the data line has a fault risk. If P<Ph, the warning information is output and marked as high risk;

[0158] The mode recognition module outputs the fault position and fault mode of the data line according to the simulation result, and realizes performance testing and fault prediction of the target mobile phone data line.

[0159] The above merely provides a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A method for performance testing and fault prediction of a mobile phone data cable, characterized by: Including: S100. Collect multiple groups of electrical performance data of the target mobile phone data cable in different test scenarios, including power-on resistance, voltage stability, charging current response, and data transmission rate, and construct an electrical performance parameter set E; S200. Obtain a stress response data set S of the mobile phone data cable under different bending angles, plugging and unplugging times, and environmental temperature usage conditions; S300. According to the electrical performance parameter set E and the stress response data set S, establish a corresponding performance degradation correlation model W to describe the relationship between the internal structure change of the data cable and the electrical performance decline; S400. Input the historical usage data of the target mobile phone data cable, and perform simulation through the W model to obtain the remaining performance index P of the data cable in the current state; S500. According to the set performance threshold Ph, compare the index P to judge whether there is a fault risk for the data cable. If P < Ph, output a warning message and mark it as a high risk; S600. Based on the simulation results, output the fault location and fault mode of the data cable to achieve performance testing and fault prediction of the target mobile phone data cable.

2. The method for performance testing and fault prediction of a mobile phone data cable according to claim 1, wherein: The S100 includes: S101. Perform periodic electrical scans on the target mobile phone data cable in four typical usage scenarios of constant voltage charging, charging while using, data synchronization, and fast charging with high load, and obtain the transient resistance, voltage fluctuation amplitude, dynamic current response, and data throughput rate in each scenario; S102. Construct a four-dimensional electrical performance vector based on each group of original test data, and uniformly map the electrical performance characteristics in each test scenario to the normalized parameter space to form a feature matrix Em that can be distinguished in the time series dimension; S103. Perform feature compression and non-linear decoupling on Em through an embedded variational autoencoder to extract a low-dimensional latent electrical expression factor E.

3. The method for performance testing and fault prediction of a mobile phone data cable according to claim 1, wherein: The S200 includes: S201. Construct a three-dimensional coupled loading platform, and realize the simulation of typical stress field conditions through a programmable flexible mechanism to apply different bending angles, plugging and unplugging cycles, and constant temperature / variable temperature environment joint control to the target data cable; S202. Synchronously deploy a strain fiber array sensor and a micro-thermocouple under each loading combination condition, and record the strain distribution, micro-crack induction points, and abnormal heat conduction paths of the outer sheath and conductor core wire of the data cable in real time to form a high-resolution stress response matrix Si; S203. Use the tensor decomposition method to perform dimensionality reduction processing on all Si, extract the key stress principal element features, and construct a multi-dimensional stress response data set S.

4. The method for performance testing and fault prediction of a mobile phone data cable according to claim 1, wherein: The S300 includes: S301. Align the time stamps and synchronize the features of the constructed electrical performance parameter set E and the extracted stress response data set S, and realize the high-dimensional correlation extraction of stress and electrical characteristics through a multi-channel residual fusion network; S302. Introduce a structure embedding tensor to construct an implicit variable map of the internal structure change of the data cable, and embed it into the performance degradation correlation model W as a model structure constraint condition; S303. Adopt a time series prediction module based on a hybrid unit of LSTM-GRU to dynamically train the model W so that it can simulate the non-linear degradation process of electrical performance parameters over time or the number of cycles under different stress paths.

5. The method for performance testing and fault prediction of a mobile phone data cable according to claim 1, wherein: The S400 includes: S401. Collect the historical usage data of the target mobile phone data cable, including the cumulative number of pluggings and unpluggings, typical bending trajectories, total charging duration, environmental temperature distribution, and usage scenario tags, and construct a usage behavior vector sequence Ut; S402. Input the usage behavior vector Ut into the trained performance degradation correlation model W, and use the time recurrence ability of the model to calculate the electrical performance evolution path within the prediction window, and output the simulated state curve Pt; S403. Based on the dynamic change rates of the resistance, current, and rate parameters in the simulation results, construct a weighted fusion index function, aggregate and reduce Pt, and generate the remaining performance index Po of the data cable in the current state.

6. A mobile phone data line performance testing and fault prediction system, used to implement the mobile phone data line performance testing and fault prediction method according to any one of claims 1 to 5, characterized in that: It includes: A data acquisition module that acquires multiple groups of electrical performance data of the target mobile phone data cable in different test scenarios, including the on-resistance, voltage stability, charging current response, and data transmission rate, and constructs an electrical performance parameter set E; A stress analysis module that obtains the stress response data set S of the mobile phone data cable under different bending angles, number of pluggings and unpluggings, and environmental temperature usage conditions; A performance degradation modeling module that establishes a corresponding performance degradation correlation model W according to the electrical performance parameter set E and the stress response data set S, and is used to describe the relationship between the internal structure change of the data cable and the electrical performance decline; An index generation module that inputs the historical usage data of the target mobile phone data cable, and through simulation by the W model, obtains the remaining performance index P of the data cable in the current state; A risk warning module that compares the index P with the set performance threshold Ph to judge whether there is a fault risk for the data cable. If P < Ph, it outputs a warning message and marks it as a high risk; A pattern recognition module that outputs the fault location and fault mode of the data cable according to the simulation results, and realizes the performance test and fault prediction of the target mobile phone data cable.

Images