Mobile phone motherboard detection system and method

Abstract

The present invention relates to the technical field of computer models, and particularly to a mobile phone motherboard detection system and method. Specifically, it includes: First, obtaining the edge data and non-edge data of the mobile phone motherboard, screening, preprocessing, and feature extraction of the edge data to obtain the edge data analysis result, preprocessing the non-edge data, and analyzing and feature selection of the preprocessed non-edge data to obtain the non-edge data analysis result; Then, constructing an intelligent recognition model, and introducing a dynamic change factor to optimize the intelligent recognition model, and using the optimized intelligent recognition model to analyze the edge data analysis result and the non-edge data analysis result to identify the aging mode and trend of the mobile phone motherboard. It solves the technical problems that the existing technology has inaccurate processing of mobile phone motherboard detection data and inaccurate detection results.

Description

Technical Field

[0001] The present invention relates to the technical field of computer models, and particularly to a mobile phone motherboard detection system and method. Background Art

[0002] With the rapid development of smartphone technology, the performance and functions of mobile phones have become increasingly powerful, which poses higher requirements for the hardware of mobile phones, especially the stability and reliability of the motherboard. As the most important component of a mobile phone, the motherboard carries core circuits such as processors, memory, and power management, and its performance directly affects the working efficiency of the entire mobile phone and the user experience. However, due to the compact design and complex circuit layout of mobile phones, the motherboard is vulnerable to various factors such as overheating, electromagnetic interference, and physical damage, which may lead to a decline in device performance, data loss, or even complete damage. Although there are some mobile phone hardware detection tools and software on the market currently, they mainly focus on the detection of parameters such as CPU usage, memory usage, and battery health status, and have weak early warning capabilities for the comprehensive performance and potential faults of mobile phone motherboards. In addition, existing detection methods often require high hardware resource consumption or have certain interference with user operations, which to a certain extent limits their application scope and user experience.

[0003] There are many methods for detecting mobile phone motherboards. The Chinese invention patent "A Method and System for Testing the Performance of a Mobile Phone Motherboard", application number: "CN202311798611.0", mainly includes: obtaining at least one key boundary electronic component among the various electronic components of the motherboard topology in the mobile phone motherboard; determining multiple reference electronic components reached by the first key boundary electronic component and the second key boundary electronic component according to a preset specified number of hops, and determining the shortest distances from the first key boundary electronic component and the second key boundary electronic component to each reference electronic component respectively; determining the closeness centrality of the first key boundary electronic component and the second key boundary electronic component respectively according to the shortest distances from the first key boundary electronic component and the second key boundary electronic component to each reference electronic component and the number of each reference electronic component. It improves the calculation of the importance of boundary electronic components in the motherboard topology and improves the performance of the mobile phone.

[0004] However, the above technology has at least the following technical problems: the technical problems of inaccurate processing of mobile phone motherboard detection data and inaccurate detection results. Summary of the Invention

[0005] The present invention provides a mobile phone motherboard detection system and method, which solves the technical problems of inaccurate processing of mobile phone motherboard detection data and inaccurate detection results in the prior art, and achieves the technical effect of accurately processing and accurately detecting mobile phone motherboard detection data.

[0006] The mobile phone motherboard detection system and method of the present invention specifically include the following technical solutions:

[0007] The mobile phone motherboard detection method includes the following steps:

[0008] S1. Obtain the edge data and non-edge data of the mobile phone motherboard, screen, preprocess, and extract features from the edge data to obtain the edge data analysis result, preprocess the non-edge data, and analyze and select features from the preprocessed non-edge data to obtain the non-edge data analysis result;

[0009] S2. Construct an intelligent recognition model, introduce a dynamic change factor to optimize the intelligent recognition model, and use the optimized intelligent recognition model to analyze the edge data analysis result and the non-edge data analysis result to identify the aging mode and trend of the mobile phone motherboard.

[0010] Preferably, the S1 specifically includes:

[0011] When extracting features from the edge data, introduce the dynamic weighted trend analysis method to perform dynamic weighted analysis on the trend of time series data.

[0012] Preferably, in the S1, it further includes:

[0013] The specific implementation of the dynamic weighted trend analysis method is as follows:

[0014] The first step is to calculate the local trend;

[0015] The second step is to evaluate the stability of the local trend;

[0016] The third step is to calculate the dynamic weight;

[0017] The fourth step is feature extraction.

[0018] Preferably, in the S1, it further includes:

[0019] When selecting features from the non-edge data, introduce the feature intelligent optimization algorithm to identify and select features.

[0020] Preferably, in the S1, it further includes:

[0021] The specific implementation process of the feature intelligent optimization algorithm is as follows:

[0022] The first step is to calculate the feature correlation degree; construct a feature matrix and a target vector, and calculate the correlation degree between each feature and the target vector;

[0023] The second step is feature optimization; based on the obtained correlation degree, select the preferred feature set.

[0024] Preferably, the S2 specifically includes:

[0025] The specific implementation process of the intelligent recognition model is as follows: The long short-term memory network is used as the basic model according to the time characteristics of edge data and non-edge data; when there are spatial features in the input data, a convolutional neural network layer is integrated in front of the long short-term memory network to extract spatial features.

[0026] Preferably, in S2, it further includes:

[0027] Construct a dynamic change layer to integrate the information of dynamic change factors, integrate the dynamic change layer into the intelligent recognition model, and perform dynamic optimization on the intelligent recognition model.

[0028] The mobile phone motherboard detection system includes the following parts:

[0029] A data acquisition module, an edge processing module, a central data processing module, an intelligent recognition module, and a dynamic optimization module;

[0030] The data acquisition module collects the real-time operation data of the mobile phone motherboard through the sensors of the mobile phone motherboard as edge data; at the same time, it obtains non-edge data through system logs, performance monitoring software, and user feedback; and sends the edge data to the edge processing module and the non-edge data to the central data processing module;

[0031] The edge processing module screens, preprocesses, extracts features from, and analyzes the edge data to obtain the edge data analysis result; provides input for the intelligent recognition module;

[0032] The central data processing module summarizes and preprocesses the non-edge data to obtain the preprocessed non-edge data, and performs data analysis and pattern recognition on the preprocessed non-edge data to obtain the non-edge data analysis result, providing data support for the intelligent recognition module;

[0033] The intelligent recognition module constructs an intelligent recognition model to analyze the analysis results of the edge processing module and the central data processing module, and identify the aging mode and trend of the mobile phone motherboard;

[0034] The dynamic optimization module introduces dynamic change factors to optimize the intelligent recognition model to obtain an optimized intelligent recognition model.

[0035] The beneficial effects of the technical solution of the present invention are:

[0036] 1. The present invention stabilizes the feature extraction process by introducing a dynamic weighted trend analysis method for edge data, reduces the influence of environmental noise and equipment fluctuations, and improves the accuracy and reliability of edge data analysis; for non-edge data, a feature intelligent optimization algorithm is introduced to evaluate the correlation and information gain between features and target aging indicators, so as to select the most valuable features and improve the accuracy and efficiency of non-edge data analysis; by extracting in-depth insights and predicting aging trends from mobile phone motherboard data, it helps to improve the performance monitoring and maintenance efficiency of mobile phone motherboards, and reduce the risk of potential failures and damages.

[0037] 2. The present invention selects the long short-term memory network as the basic model, which is suitable for processing time series data of mobile phone motherboards. By integrating convolutional neural network layers, spatial features are also considered, improving the model's ability to represent data; a dynamic change factor is introduced, and a dynamic change layer is introduced based on the dynamic change factor to integrate the information of all dynamic change factors, and weights are applied after the input layer of the model to achieve dynamic optimization of the model, which helps the model better adapt to different environments and usage conditions; by training and validating the model, the intelligent recognition model is applied to the results of edge data analysis and non-edge data analysis to achieve intelligent recognition and analysis of the aging patterns and trends of mobile phone motherboards, which helps to timely discover potential problems and take corresponding measures to maintain the performance and reliability of mobile phone motherboards. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a module diagram of a mobile phone motherboard detection system provided by an embodiment of the present invention;

[0039] Figure 2 It is a flowchart of a mobile phone motherboard detection method provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0040] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0041] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0042] The specific solutions of the mobile phone motherboard detection system and method provided by the present invention will be specifically described below with reference to the accompanying drawings.

[0043] Refer to the appendixFigure 1 , which shows a module diagram of the mobile phone motherboard detection system provided by an embodiment of the present invention. The system includes the following parts:

[0044] A data acquisition module, an edge processing module, a central data processing module, an intelligent recognition module, and a dynamic optimization module;

[0045] The data acquisition module collects real-time operation data of the mobile phone motherboard as edge data through various sensors on the mobile phone motherboard. The edge data includes physical parameters such as temperature, voltage, and current. At the same time, non-real-time non-edge data is obtained through system logs, performance monitoring software, and user feedback. The non-real-time non-edge data includes long-term performance logs and user usage pattern data. The edge data is sent to the edge processing module, and the non-edge data is sent to the central data processing module;

[0046] The edge processing module screens, preprocesses, extracts features, and performs preliminary analysis on the edge data to obtain edge data analysis results. The edge data analysis results include key features and warning indicators, which provide inputs for the intelligent recognition module;

[0047] The central data processing module summarizes and preprocesses the non-edge data to obtain preprocessed non-edge data, and performs data analysis and pattern recognition on the preprocessed non-edge data to extract in-depth insights from multi-dimensional data to obtain non-edge data analysis results, providing data support for the intelligent recognition module;

[0048] The intelligent recognition module constructs an intelligent recognition model to analyze the analysis results of the edge processing module and the central data processing module, identifies the aging mode and trend of the mobile phone motherboard, and realizes the detection of the mobile phone motherboard;

[0049] The dynamic optimization module introduces a dynamic change factor to optimize the intelligent recognition model to obtain an optimized intelligent recognition model, improving the accuracy of mobile phone motherboard detection.

[0050] Referring to the appendix Figure 2 , which shows a flowchart of the mobile phone motherboard detection method provided by an embodiment of the present invention. The method includes the following steps:

[0051] S1. Obtain the edge data and non-edge data of the mobile phone motherboard, screen, preprocess, and extract features from the edge data to obtain edge data analysis results, preprocess the non-edge data, and analyze and select features from the preprocessed non-edge data to obtain non-edge data analysis results;

[0052] Collect the real-time operation data of the mobile phone motherboard through sensors on the motherboard, and use the collected real-time operation data as edge data; at the same time, obtain non-real-time non-edge data of the mobile phone motherboard through system logs, performance monitoring software, and user feedback;

[0053] For the edge data of the mobile phone motherboard:

[0054] First, use a noise filtering algorithm to filter the edge data of the mobile phone motherboard to smooth the data and reduce the impact of random fluctuations, achieving noise filtering and obtaining the edge data of the mobile phone motherboard after noise reduction; screen the edge data of the mobile phone motherboard after noise reduction according to predefined thresholds and standards based on the expert experience method, removing data points that clearly do not meet logical and physical limits, such as temperature or current readings exceeding the maximum operating range of the device; obtain the screened edge data of the mobile phone motherboard; perform format conversion on the screened edge data of the mobile phone motherboard to facilitate subsequent processing and analysis, and perform normalization processing on the edge data of the mobile phone motherboard after format conversion to eliminate the impact of different dimensions; obtain the preprocessed edge data of the mobile phone motherboard;

[0055] For the preprocessed edge data of the mobile phone motherboard, first, define and extract features representing the operating state and potential aging signs of the mobile phone motherboard; the features include statistical features, time series features, and anomaly detection features; for statistical features, use statistical analysis methods to obtain statistical features for each key parameter. Statistical analysis methods are well-known technical means in the art, and statistical features include maximum value, minimum value, average value, median, standard deviation, etc., to capture the basic distribution characteristics of the data; for time series features, introduce time series analysis methods to extract the preprocessed edge data of the mobile phone motherboard, such as trend slope, periodic components (extracted through Fourier transform), and autocorrelation, to identify and quantify the trends and periodic components of the data; for anomaly detection features, based on set thresholds, analyze abnormal fluctuations or over-standard events in the data, record the frequency and duration of the events, and anomaly detection features include the time length exceeding the threshold and the number of abnormal fluctuations; then integrate the extracted features into a preliminary comprehensive feature set, and use feature selection techniques to perform feature selection and dimensionality reduction on the preliminary comprehensive feature set to reduce the number of features while retaining the most informative features; obtain a comprehensive data set representing the edge data analysis results;

[0056] When extracting features for edge data, due to environmental noise and device fluctuations, the directly extracted features may have significant instability, affecting the accuracy of the final feature set. The present invention introduces a dynamic weighted trend analysis method to stabilize feature extraction; the specific implementation of the dynamic weighted trend analysis method is as follows:

[0057] Dynamic weighted trend analysis method, through dynamic weighted analysis of the trend of time series data, to stabilize the feature extraction process and reduce the influence of noise and fluctuations;

[0058] The first step is to calculate the local trend; for the time series data x(t) of the edge data of the mobile phone motherboard, calculate the local trend L′ at the time point t within the time window W t :

[0059]

[0060] where α is a decay coefficient used to strengthen the influence of the most recent data points; x(t - i) represents the time series data of the edge data of the mobile phone motherboard at the time point t - i; x(t - i - 1) represents the time series data of the edge data of the mobile phone motherboard at the time point t - i - 1;

[0061] The second step is to evaluate the stability of the local trend; introduce a trend consistency metric for evaluation:

[0062]

[0063] where ∈ is a very small positive number used to prevent the denominator from being zero; L′ T-i , L′ t-i-1 respectively represent the local trends at the time points t - i and t - i - 1; C t is the trend consistency metric that evaluates the consistency of a series of local trends;

[0064] The third step is to calculate the dynamic weight; according to the stability and consistency of the local trend, calculate the dynamic weight W t ′:

[0065]

[0066] where σ L′t is the standard deviation of the local trend stability, used to measure the dispersion degree of a series of local trend values;

[0067] The fourth step is feature extraction; based on the time series data x(t) of the edge data of the mobile phone motherboard, finally obtain the dynamic weight W t ′, use the dynamic weight to weight the local trend, realize the stable feature extraction process, and reduce the influence of noise and fluctuations; specifically:

[0068]

[0069] where F′ is the stably extracted feature, which comprehensively considers the indicators of the trend changes of all local time windows in the entire time series; T is the total length of the time series.

[0070] Non-edge data of the mobile phone motherboard:

[0071] First, perform preprocessing on the non-edge data of the mobile phone motherboard, such as data cleaning, format transformation, segmentation, and classification; obtain the preprocessed non-edge data of the mobile phone motherboard; process the preprocessed non-edge data of the mobile phone motherboard according to the analysis process of the edge data of the mobile phone motherboard to obtain a comprehensive feature set representing the analysis results of the non-edge data of the mobile phone motherboard;

[0072] When performing feature selection on non-edge data, it is difficult to automatically identify and select the most informative features from a large number of features, which affects the accuracy of the feature set; the present invention introduces a feature intelligent optimization algorithm to intelligently select the most valuable features by evaluating the correlation degree and information gain between features and the target aging index; the specific implementation process is as follows:

[0073] The first step is to calculate the feature correlation degree; calculate the correlation degree between each feature and the target aging index to evaluate the contribution degree of each feature to predicting the aging trend:

[0074] First, construct a feature matrix and a target vector; construct a feature matrix F and a target vector Y, where each column of F represents a feature; Y represents the aging index; ensure that each row of F and Y corresponds to the same data point (i.e., the non-edge data at the same observation time point);

[0075] Then traverse each column of the feature matrix F and calculate the mutual information; for each column (feature) F k in the feature matrix F, calculate the mutual information I(F k ; Y) between F k and the target vector Y:

[0076]

[0077] where p(f,y) is the joint probability that the feature F k takes the value f and the aging index Y takes the value y; p(f) represents the marginal probability that the feature F k takes a specific value f; p(y) is the marginal probability that the aging index Y takes a specific value y;

[0078] Finally, calculate I(F k ; Y) according to the mutual information formula to obtain the correlation degree between each feature and the aging index.

[0079] The second step is feature optimization; based on the calculated feature correlation degree, select the features with the highest correlation degree to form an optimized feature set for subsequent aging pattern recognition;

[0080] Set the threshold θ according to the expert experience method, which is used to screen features with a correlation degree higher than the threshold θ. Traverse all features, and select the features with the correlation degree I(F k ; Y) higher than the threshold θ into the preferred feature set F selected ;

[0081] S2. Construct an intelligent recognition model, and at the same time introduce a dynamic change factor to optimize the intelligent recognition model. Use the optimized intelligent recognition model to analyze the edge data analysis results and non-edge data analysis results to identify the aging mode and trend of the mobile phone motherboard.

[0082] Construct an intelligent recognition model to identify the aging model and trend of the mobile phone motherboard. The specific implementation process is as follows:

[0083] First, select the long short-term memory network as the basic model according to the time characteristics of the edge data and non-edge data of the mobile phone motherboard. The basic units of the long short-term memory network are:

[0084] f t =σ(w f ·[h t-1 ,Fea t +b f )

[0085] i t =σ(W i ·[h t-1 ,Fea t +b i )

[0086]

[0087]

[0088] o t =σ(W o ·[h t-1 ,Fea t +b o )

[0089] h t =o t .×tanh(C t )

[0090] Among them, f t ,i t ,o t are the activation values of the forget gate, input gate and output gate respectively; is the candidate cell state; C t is the cell state; h t is the hidden state; W f ,W i ,WC , W o and b f , b i , b C , b o are the weights and biases of each cell in the LSTM model, σ is the sigmoid function, and.× represents element-wise multiplication; Fea t represents the input data at time step t; [h t-1 , Fea t is a combined vector that concatenates the hidden state h t-1 at time step t - 1 and the input data Fea t at the current time step t. The combined vector provides historical information for the LSTM cell (through h t-1 and the current information Fea t ), enabling the LSTM model to consider the historical dependence of time series data);

[0091] When there are spatial features in the input data, a convolutional neural network (CNN) layer is integrated before the LSTM to extract spatial features. The convolutional neural network layer can be expressed as:

[0092] C out = ReLU(W cnn * C in + b cnn )

[0093] where C out is the output of the convolutional layer, C in is the input of the convolutional layer, Wc cnn is the convolutional kernel, b cnn is the bias term, * represents the convolution operation, and ReLU is the activation function.

[0094] In the intelligent recognition model, the mean squared error combined with L2 regularization is used as the loss function to measure the difference between the model prediction and the actual value. Finally, the cross-validation method is used to validate the intelligent recognition model to obtain the final intelligent recognition model; the mean squared error and L2 regularization are well-known technical means to those skilled in the art;

[0095] A dynamic change layer is constructed to integrate the dynamic factors that affect the mobile phone motherboard, such as environmental temperature, device usage frequency, and battery power. The dynamic factors are used as dynamic change factors {D 1 , D 2 , …, D n}, where n represents the number of dynamic change factors, and D n represents the nth dynamic change factor; for each dynamic change factor D i , a mapping function f i (Di ) Map the dynamic change factors to an appropriate range to better combine with the intelligent recognition model; further, the dynamic change layer is used to integrate the information of all dynamic change factors, and the implementation of the dynamic change layer is expressed as:

[0096]

[0097] where D out is the output of the dynamic change layer, and w i is the weight related to the dynamic change factor D i .

[0098] Integrate the dynamic change layer into the intelligent recognition model as the first layer after the input layer to achieve dynamic optimization of the intelligent recognition model;

[0099] Finally, use the optimized intelligent recognition model to perform intelligent recognition and analysis on the edge data analysis results and non-edge data analysis results, identify the aging mode and trend of the mobile phone motherboard, and realize the detection of the mobile phone motherboard.

[0100] In summary, the mobile phone motherboard detection system and method are completed.

[0101] The order of the invention embodiments is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0102] Each embodiment in this specification is described in a progressive manner. For the same or similar parts between each embodiment, reference can be made to each other. The key point of each embodiment is to illustrate the differences from other embodiments.

[0103] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for detecting a mobile phone motherboard, characterized in that: The following steps are involved: S1. The real-time operation data of the mobile phone motherboard is collected through sensors on the mobile phone motherboard, and the collected real-time operation data is used as edge data; at the same time, the non-real-time non-edge data of the mobile phone motherboard is obtained through system logs, performance monitoring software and user feedback; The edge data is screened, preprocessed and feature extracted to obtain the edge data analysis results; when extracting features from edge data, the dynamic weighted trend analysis method is introduced to perform dynamic weighted analysis on the trend of time series data; the specific implementation of the dynamic weighted trend analysis method is as follows: The first step is to calculate the local trend; The second step is to assess the stability of local trends; The third step is to calculate the dynamic weight; The fourth step is feature extraction; Preprocess the non-edge data of the mobile phone motherboard by data cleaning, format conversion, segmentation, and classification; Obtain preprocessed non-edge data of the mobile phone mainboard; The preprocessed non-edge data of the mobile phone mainboard is processed according to the analysis process of the edge data of the mobile phone mainboard to obtain a comprehensive feature set representing the analysis results of the non-edge data of the mobile phone mainboard; S2. Build an intelligent recognition model to identify the aging patterns and trends of mobile phone motherboards. The specific implementation process is as follows: The long short-term memory network is used as the basic model according to the temporal characteristics of edge data and non-edge data. When there are spatial features in the input data, a convolutional neural network layer is integrated before the long short-term memory network to extract the spatial features. Construct a dynamic change layer, integrate the information of dynamic change factors, integrate the dynamic change layer into the intelligent recognition model, and dynamically optimize the intelligent recognition model.

2. The mobile phone motherboard detection method according to claim 1, characterized in that: In said S1, it also includes: When performing feature selection on non-edge data, an intelligent feature optimization algorithm is introduced to identify and select features.

3. The mobile phone motherboard detection method according to claim 2, characterized in that: In said S1, it also includes: The specific implementation process of the feature intelligent optimization algorithm is as follows: The first step is to calculate the feature correlation; construct the feature matrix and the target vector, and calculate the correlation between each feature and the target vector; The second step is feature optimization; based on the obtained correlation, the optimal feature set is selected.

4. A mobile phone motherboard detection system, applied to the mobile phone motherboard detection method as claimed in claim 1, characterized in that: Includes the following sections: Data acquisition module, edge processing module, central data processing module, intelligent identification module, dynamic optimization module; The data acquisition module collects the real-time operation data of the mobile phone motherboard as edge data through the sensor of the mobile phone motherboard; at the same time, it obtains non-edge data through system logs, performance monitoring software and user feedback; and sends the edge data to the edge processing module, and sends the non-edge data to the central data processing module; The edge processing module screens, preprocesses, extracts and analyzes edge data to obtain edge data analysis results; Provide input for intelligent recognition module; The central data processing module aggregates and preprocesses the non-edge data to obtain preprocessed non-edge data, and performs data analysis and pattern recognition on the preprocessed non-edge data to obtain non-edge data analysis results to provide data support for the intelligent recognition module; The intelligent recognition module constructs an intelligent recognition model to analyze the analysis results of the edge processing module and the central data processing module to identify the aging mode and trend of the mobile phone motherboard; The dynamic optimization module introduces a dynamic change factor to optimize the intelligent recognition model to obtain an optimized intelligent recognition model.

Images