CN118035848A - Intelligent aging test system for electronic element - Google Patents
Intelligent aging test system for electronic element Download PDFInfo
- Publication number
- CN118035848A CN118035848A CN202410425973.3A CN202410425973A CN118035848A CN 118035848 A CN118035848 A CN 118035848A CN 202410425973 A CN202410425973 A CN 202410425973A CN 118035848 A CN118035848 A CN 118035848A
- Authority
- CN
- China
- Prior art keywords
- aging
- performance
- prediction
- failure
- attenuation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 230000032683 aging Effects 0.000 title claims abstract description 181
- 238000012360 testing method Methods 0.000 title claims abstract description 23
- 238000004458 analytical method Methods 0.000 claims abstract description 92
- 238000005457 optimization Methods 0.000 claims abstract description 35
- 230000006399 behavior Effects 0.000 claims abstract description 32
- 230000035882 stress Effects 0.000 claims abstract description 31
- 230000000694 effects Effects 0.000 claims abstract description 24
- 238000004088 simulation Methods 0.000 claims abstract description 19
- 230000002159 abnormal effect Effects 0.000 claims abstract description 13
- 238000010801 machine learning Methods 0.000 claims abstract description 13
- 230000008859 change Effects 0.000 claims description 67
- 238000000034 method Methods 0.000 claims description 24
- 238000011156 evaluation Methods 0.000 claims description 19
- 230000007613 environmental effect Effects 0.000 claims description 13
- 238000012417 linear regression Methods 0.000 claims description 12
- 230000010485 coping Effects 0.000 claims description 11
- 238000013461 design Methods 0.000 claims description 11
- 230000004044 response Effects 0.000 claims description 11
- 230000008901 benefit Effects 0.000 claims description 10
- 238000004422 calculation algorithm Methods 0.000 claims description 10
- 230000009467 reduction Effects 0.000 claims description 9
- 238000000342 Monte Carlo simulation Methods 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 8
- 239000000284 extract Substances 0.000 claims description 7
- 239000010410 layer Substances 0.000 claims description 7
- 230000003449 preventive effect Effects 0.000 claims description 7
- 230000001133 acceleration Effects 0.000 claims description 6
- 230000015556 catabolic process Effects 0.000 claims description 6
- 238000006731 degradation reaction Methods 0.000 claims description 6
- 230000010354 integration Effects 0.000 claims description 6
- 238000000611 regression analysis Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 238000012300 Sequence Analysis Methods 0.000 claims description 5
- 238000012795 verification Methods 0.000 claims description 5
- 230000008451 emotion Effects 0.000 claims description 4
- 238000003062 neural network model Methods 0.000 claims description 4
- 238000003909 pattern recognition Methods 0.000 claims description 4
- 238000012502 risk assessment Methods 0.000 claims description 4
- 238000012216 screening Methods 0.000 claims description 4
- 239000011229 interlayer Substances 0.000 claims description 3
- 238000011158 quantitative evaluation Methods 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 230000004807 localization Effects 0.000 claims 1
- 230000006872 improvement Effects 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 33
- 238000010586 diagram Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 6
- 238000000556 factor analysis Methods 0.000 description 5
- 230000007774 longterm Effects 0.000 description 4
- 230000003938 response to stress Effects 0.000 description 4
- 208000025174 PANDAS Diseases 0.000 description 3
- 208000021155 Paediatric autoimmune neuropsychiatric disorders associated with streptococcal infection Diseases 0.000 description 3
- 240000004718 Panda Species 0.000 description 3
- 235000016496 Panda oleosa Nutrition 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000012706 support-vector machine Methods 0.000 description 3
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 2
- 238000003066 decision tree Methods 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- YHXISWVBGDMDLQ-UHFFFAOYSA-N moclobemide Chemical compound C1=CC(Cl)=CC=C1C(=O)NCCN1CCOCC1 YHXISWVBGDMDLQ-UHFFFAOYSA-N 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 230000004083 survival effect Effects 0.000 description 2
- 241000677647 Proba Species 0.000 description 1
- 238000005299 abrasion Methods 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000005094 computer simulation Methods 0.000 description 1
- 230000001351 cycling effect Effects 0.000 description 1
- 238000013016 damping Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of electronic element testing, in particular to an intelligent aging testing system for electronic elements, which comprises: the system comprises a nonlinear attenuation behavior identification module, a stress simulation and effect analysis module, an attenuation mode dynamic modeling module, a performance prediction and failure analysis module, a machine learning enhancement prediction module, an aging path optimization module and a failure point positioning module. According to the invention, the adopted nonlinear attenuation behavior recognition module can accurately capture the abnormal attenuation mode in the aging process, so that the accuracy and efficiency of the aging test are remarkably improved, the machine learning enhancement prediction module maps the aging behavior and failure risk of the electronic element by analyzing the performance and failure prediction data, optimizes the prediction model, further improves the prediction accuracy, and combines the technical means, so that the system not only can accurately predict the performance attenuation and failure point of the electronic element, but also can provide a targeted improvement strategy for manufacturers, remarkably improve the reliability of the electronic element and prolong the service life.
Description
Technical Field
The invention relates to the technical field of electronic element testing, in particular to an intelligent aging test system for electronic elements.
Background
The field of electronic component testing technology is focused on evaluating and verifying the performance, reliability, and durability of electronic components. This field encompasses a wide range of testing methods including, but not limited to, electrical testing, thermal testing, mechanical testing, and environmental testing to ensure that the electronic components meet the intended standards and requirements during design, production, and use. Through these tests, manufacturers and designers may find potential design defects, material problems, or production defects, thereby optimizing product design and improving product reliability and performance.
The intelligent aging test system for the electronic components is an important component in the technical field of electronic component testing, and aims to simulate long-term stress and abrasion experienced by the electronic components in an actual use environment so as to predict the service life and reliability of the electronic components. The performance of the electronic component is observed over time by applying continuous or periodic stresses (e.g., temperature cycling, voltage stress, mechanical stress, etc.). The goal is to identify weaknesses that lead to early failure, thereby helping manufacturers improve product design, reduce failure rates during or after warranty, and improve end-user satisfaction and trust.
Methods of conventional electronic component burn-in systems have limited accuracy in handling complex burn-in behavior, particularly in identifying non-linearities and non-routine burn-in processes. Such limitations make it difficult to effectively predict early failure points or provide targeted optimization strategies. The lack of advanced technology applications such as machine learning makes it difficult for conventional models to adapt to complex varying aging behaviors, affecting accurate assessment of electronic component reliability. These deficiencies lead to manufacturers facing higher failure rates and maintenance costs, reducing user satisfaction and trust, affecting brand reputation and market competitiveness.
Disclosure of Invention
The invention aims to solve the defects in the prior art, and provides an intelligent aging test system for electronic elements.
In order to achieve the above purpose, the present invention adopts the following technical scheme: an intelligent burn-in system for electronic components includes:
The nonlinear attenuation behavior recognition module is used for analyzing performance fluctuation and variation trend by collecting electronic element operation data and recognizing an abnormal attenuation mode in an aging process to obtain a nonlinear attenuation characteristic set;
the stress simulation and effect analysis module simulates the response of the electronic element based on the nonlinear attenuation characteristic set, analyzes the influence of stress conditions on the element performance, and obtains performance change trend data;
the attenuation mode dynamic modeling module constructs a dynamic model of the electronic element aging process according to the performance change trend data, acquires a performance attenuation path of the element along with time, and forms a dynamic attenuation path model;
the performance prediction and failure analysis module predicts the performance and failure point of the electronic element based on the dynamic attenuation path model, analyzes the performance of the element in the differential aging stage, and outputs performance and failure prediction data;
The machine learning enhancement prediction module optimizes the prediction model through the analysis performance and failure prediction data, maps the aging behavior and failure risk of the electronic element and outputs a refined prediction model;
the aging path optimization module analyzes the aging process of the electronic element by utilizing the refined prediction model, and proposes a coping strategy to obtain an aging optimization strategy set;
and the failure point positioning module analyzes key links in the element aging process based on the aging optimization strategy set, sets failure points and obtains failure point positioning data.
As a further scheme of the invention, the nonlinear attenuation characteristic set comprises characteristics of response time extension, power consumption increase proportion and performance attenuation rate of an element at ambient temperature, the performance change trend data comprises performance decrease speed under temperature stress, electrical characteristic change under humidity condition and life prediction data under mechanical stress influence, the dynamic attenuation path model comprises time sequence analysis results, performance decrease key nodes and ageing acceleration factor influence evaluation contents, the performance and failure prediction data comprises performance stability prediction in a short term, failure time prediction in a medium term and a long term and performance decrease map in a differential ageing stage, the refinement prediction model specifically comprises a neural network model trained based on historical data, regression analysis results of ageing behaviors and quantitative evaluation indexes of failure risk, the ageing optimization strategy set comprises temperature control optimization, humidity management scheme and current load adjustment strategy, and the failure point positioning data comprises time marks of ageing key points, rate threshold values of performance decrease and safe working ranges of key electrical parameters.
As a further aspect of the present invention, the nonlinear decay behavior identification module includes:
The fluctuation analysis submodule carries out fluctuation condition analysis based on the operation data of the electronic element, measures the fluctuation amplitude and the frequency of multiple parameters, determines the performance stability and generates a fluctuation analysis result;
The trend prediction submodule adopts the fluctuation analysis result, performs change trend analysis by using a linear regression algorithm, identifies the problem trend of performance reduction or stability, predicts future performance change and generates a change trend prediction result;
the abnormal pattern recognition submodule captures the change trend according to the change trend prediction result, predicts the performance change according to the change trend, extracts a performance attenuation signal from the prediction result, recognizes a nonlinear attenuation pattern, sets abnormal attenuation characteristics according to nonlinear characteristics, and generates a nonlinear attenuation characteristic set.
As a further aspect of the present invention, the linear regression algorithm uses the formula:
；
calculating a fluctuation amplitude variation based on the time-series data, wherein, Is time series data,/>Is a square term of time series data to increase the nonlinear capturing capability of the model to time trends,/>Is natural logarithmic transformation of time series data to reduce bias of data,/>Is an external influencing factor, including market emotion index,/>For intercept,/>Is a linear coefficient of time series data,/>Coefficients that are square terms of time series data,/>Coefficient of natural logarithm of time series data,/>Is the coefficient of external influence factor,/>Is an error term.
As a further aspect of the present invention, the attenuation mode dynamic modeling module includes:
the trend analysis submodule adopts the performance change trend data of the electronic element to identify characteristics of early, middle and late attenuation, analyzes performance fluctuation in a time sequence, extracts attenuation rate, compares performance change among differential elements, and generates attenuation rate and trend comparison analysis results;
The attenuation path modeling submodule calibrates key nodes in the attenuation process based on the attenuation rate and the trend comparison analysis result, maps the relation between performance indexes and time, constructs an attenuation path frame, draws the change track of element performance along with the time, and obtains a preliminary attenuation path model;
The model verification submodule compares actual application data with model prediction data by using the preliminary attenuation path model, determines a parameter adjustment direction through differential analysis, adjusts model parameters and establishes a dynamic attenuation path model.
As a further aspect of the present invention, the machine learning enhancement prediction module includes:
The aging characteristic analysis submodule carries out key index screening based on the performance and failure prediction data, identifies factors influencing the performance and service life of the electronic element, eliminates indexes with small fluctuation through variation coefficients, strengthens aging sensitive characteristics, and generates a key aging characteristic data set;
the model refinement training sub-module carries out model structural design based on the key aging characteristic data set, adjusts a data input layer to match characteristic dimensions, enhances characteristic expression through interlayer connection, calibrates and outputs the prediction requirement of matching aging behaviors, and obtains an optimized aging behavior prediction model;
and the failure risk prediction submodule is used for carrying out failure risk analysis by utilizing a Monte Carlo simulation method based on the optimized aging behavior prediction model, evaluating the aging speed and the failure probability of the element by simulating a prediction scene under the differentiated use condition, comprehensively obtaining the failure risk level of the element, and producing a refined prediction model.
As a further scheme of the present invention, the monte carlo simulation method adopts the formula:
；
The probability of an event occurring is calculated, wherein, For the number of times event X occurs in the simulation,/>For the total simulation times,/>Adjusting the coefficient for environmental factors, wherein S is an environmental stability index,/>For using conditional complexity coefficients, T is the total number of test scenarios,/>And U is a diversity index of user operation.
As a further aspect of the present invention, the aging path optimization module includes:
the aging process analysis submodule identifies key time points and condition changes in the aging process based on the refined prediction model, analyzes the condition changes affecting the element performance, selects aging acceleration and slowing factors and generates an aging key point analysis result;
The strategy making submodule makes preventive measures oriented to a key aging stage based on the aging key point analysis result, designs an intervention strategy to adjust the use condition or the physical structure of the element, and obtains a preliminary aging coping strategy set;
The strategy effect prediction submodule simulates aging path change after strategy implementation based on the preliminary aging coping strategy set, evaluates influence of preventive measures on the service life of the extension element, screens an effect optimal strategy and obtains a strategy effect evaluation result;
And the optimization strategy integration submodule optimizes and integrates strategies based on the strategy effect evaluation result, identifies the advantages and limitations of each strategy by analyzing benefit and cost data of multiple strategies, classifies and distributes weights of the strategies, combines the aging characteristics of the electronic components and the expected service life targets, and establishes a comprehensive aging optimization scheme to produce an aging optimization strategy set.
As a further aspect of the present invention, the failure point positioning module includes:
The aging process tracking submodule is used for recording performance parameter changes and environmental conditions of the element in a plurality of aging stages based on the aging optimization strategy set, identifying and recording key links causing performance degradation, and generating an aging track record;
the failure point presetting submodule analyzes key time of performance reduction based on the aging track record, establishes a reason behind the failure point, presets a predicted failure point and acquires preset failure point analysis;
And the data positioning and integrating sub-module carries out logic judgment and rule matching on key factors and preset failure points in the aging process based on the analysis of the preset failure points, evaluates the correlation between the key factors and the preset failure points, integrates analysis and matching results, and constructs a connection map between the failure points and the aging factors to obtain failure point positioning data.
Compared with the prior art, the invention has the advantages and positive effects that:
According to the invention, the adopted nonlinear attenuation behavior recognition module can accurately capture the abnormal attenuation mode in the aging process, and the technical innovation obviously improves the precision and efficiency of the aging test. The machine learning enhancement prediction module maps the aging behavior and failure risk of the electronic element by analyzing the performance and failure prediction data, optimizes the prediction model and further improves the prediction accuracy. The combination of the technical means ensures that the system not only can accurately predict the performance attenuation and failure point of the electronic element, but also can provide a targeted improvement strategy for manufacturers, obviously improve the reliability of the electronic element, prolong the service life and further improve the satisfaction degree and the trust degree of the end user.
Drawings
FIG. 1 is a system flow diagram of the present invention;
FIG. 2 is a schematic diagram of a system framework of the present invention;
FIG. 3 is a flow chart of a nonlinear damping behavior recognition module of the present invention;
FIG. 4 is a flow chart of a stress simulation and effect analysis module according to the present invention;
FIG. 5 is a flow chart of the attenuation mode dynamic modeling module of the present invention;
FIG. 6 is a flow chart of a performance prediction and failure analysis module according to the present invention;
FIG. 7 is a flow chart of a machine learning enhancement prediction module of the present invention;
FIG. 8 is a flowchart of an aging path optimization module according to the present invention;
FIG. 9 is a flow chart of a failure point location module of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
In the description of the present invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," and the like indicate orientations or positional relationships based on the orientation or positional relationships shown in the drawings, merely to facilitate describing the present invention and simplify the description, and do not indicate or imply that the devices or elements referred to must have a specific orientation, be configured and operated in a specific orientation, and therefore should not be construed as limiting the present invention. Furthermore, in the description of the present invention, the meaning of "a plurality" is two or more, unless explicitly defined otherwise.
Embodiment one: referring to fig. 1 to 2, an intelligent burn-in system for electronic devices includes:
The nonlinear attenuation behavior recognition module is used for analyzing performance fluctuation and variation trend by collecting electronic element operation data and recognizing an abnormal attenuation mode in an aging process to obtain a nonlinear attenuation characteristic set;
the stress simulation and effect analysis module simulates response of the electronic element based on the nonlinear attenuation characteristic set, analyzes influence of stress conditions on element performance, and obtains performance change trend data;
the attenuation mode dynamic modeling module constructs a dynamic model of the electronic element aging process according to the performance change trend data, acquires a performance attenuation path of the element along with time, and forms a dynamic attenuation path model;
the performance prediction and failure analysis module predicts the performance and failure point of the electronic element based on the dynamic attenuation path model, analyzes the performance of the element in the differential aging stage, and outputs performance and failure prediction data;
the machine learning enhancement prediction module optimizes the prediction model by analyzing the performance and failure prediction data, maps the aging behavior and failure risk of the electronic element and outputs a refined prediction model;
The aging path optimization module analyzes the aging process of the electronic element by using the refined prediction model, and proposes a coping strategy to obtain an aging optimization strategy set;
the failure point positioning module analyzes key links in the aging process of the element based on the aging optimization strategy set, sets failure points and obtains failure point positioning data.
The nonlinear attenuation characteristic set comprises characteristics of response time extension, power consumption increasing proportion and performance attenuation rate of an element at ambient temperature, performance change trend data comprise performance reduction speed under temperature stress, electrical characteristic change under humidity condition and life prediction data under mechanical stress influence, the dynamic attenuation path model comprises time sequence analysis results, performance reduction key nodes and content of influence evaluation of aging acceleration factors, the performance and failure prediction data comprise performance stability prediction in a short term, middle-long term failure time prediction and performance reduction maps of differential aging stages, the refinement prediction model is specifically a neural network model trained based on historical data, regression analysis results of aging behaviors and quantitative evaluation indexes of failure risks, the aging optimization strategy set comprises temperature control optimization, humidity management scheme and current load adjustment strategy, the failure point positioning data comprise time marks of aging key points, performance reduction speed thresholds and safe working ranges of key electrical parameters.
Referring to fig. 3, the nonlinear decay behavior recognition module includes:
The fluctuation analysis submodule carries out fluctuation condition analysis based on the operation data of the electronic element, measures the fluctuation amplitude and the frequency of multiple parameters, determines the performance stability and generates a fluctuation analysis result;
The fluctuation analysis submodule performs signal frequency analysis by using a NumPy library to execute an fft function based on electronic element operation data, sets variable data as input electronic element operation data, calls np.fft.fft (data) to execute fast fourier transform, acquires frequency components through np.fft.fftfreq (len (data)), calculates the amplitude of each frequency component, calculates an amplitude value by using abs (fft result), records the amplitude and the frequency, and generates a fluctuation analysis result.
The trend prediction submodule adopts a fluctuation analysis result, performs change trend analysis by using a linear regression algorithm, identifies a performance reduction or stability problem trend, predicts future performance change and generates a change trend prediction result;
the trend prediction submodule adopts a fluctuation analysis result, a linear regression algorithm is utilized, a LinearRegression class of scikit-learn library is used for creating a linear regression model, X is set as time series data, y is the amplitude in the fluctuation analysis result, fit (X.reshape (-1, 1) is called, y) is used for training the model, predict (X.reshape (-1, 1)) is used for predicting the future amplitude change based on the existing data, the predicted amplitude change trend is recorded, and a change trend prediction result is generated.
The linear regression algorithm uses the formula:
；
calculating a fluctuation amplitude variation based on the time-series data, wherein, Is time series data,/>Is a square term of time series data to increase the nonlinear capturing capability of the model to time trends,/>Is natural logarithmic transformation of time series data to reduce bias of data,/>Is an external influencing factor, including market emotion index,/>For intercept,/>Is a linear coefficient of time series data,/>Coefficients that are square terms of time series data,/>Coefficient of natural logarithm of time series data,/>Is the coefficient of external influence factor,/>Is an error term.
First, time-series data is obtainedSubstituted into the original formula for capturing linear trend, and then introduced/>The nonlinear capture capacity of the model over time trend is increased, followed by/>Converting the skewness of the processed data to make the model more accurate, and then adding/>Parameters representing external influencing factors, such as market emotion, to take into account other influencing factors besides time series, each coefficient/>、/>、/>、/>、/>Obtained by data fitting, ensures that the model can accurately reflect the influence of time series data and external factors on the fluctuation amplitude, and finally, the error term/>Representing random fluctuations that the model cannot interpret.
The abnormal pattern recognition submodule captures a change trend through a change trend prediction result, predicts performance change through the change trend, extracts a performance attenuation signal from the prediction result, recognizes a nonlinear attenuation pattern, sets abnormal attenuation characteristics according to nonlinear characteristics, and generates a nonlinear attenuation characteristic set.
The abnormal pattern recognition submodule classifies performance change through a change trend prediction result by utilizing a Support Vector Machine (SVM) algorithm, SVC class of scikit-learn library is used for classifying performance change, variable X is set to be time and prediction amplitude in the change trend prediction result, y is set to be a mark of a performance attenuation signal, fit (X, y) is called for model training, the performance attenuation pattern is classified by utilizing a prediction (X) based on the prediction result, the identified nonlinear attenuation pattern is analyzed according to decision boundaries of the SVM model, abnormal attenuation characteristics are determined, and a nonlinear attenuation characteristic set is generated.
Referring to fig. 4, the stress simulation and effect analysis module includes:
The stress simulation submodule simulates the response of the electronic element under the differential stress condition based on the nonlinear attenuation characteristic set, wherein the response comprises the change of temperature, voltage and vibration factors, records the performance response data of the element under various stresses, and generates a stress response data set;
The stress simulation submodule adopts a machine learning algorithm based on a nonlinear attenuation characteristic set, utilizes Python language programming, calls a random forest regressor (RandomForestRegressor) through a scikit-learn library, sets the number of trees as 100, and the maximum depth as 10, simulates the response of electronic elements under the condition of simulating differential stress, including the changes of temperature, voltage and vibration factors, simulates each electronic element, records the performance response data of the element under various stresses by setting different temperature, voltage and vibration parameters as input, and generates a stress response data set.
The effect analysis submodule adopts a stress response data set to analyze the influence of differential stress conditions on the performance of the electronic element, compares the change of the element performance under differential stress, identifies key influence factors and generates an influence factor analysis result;
The effect analysis sub-module adopts a statistical analysis method, utilizes Python language programming, performs data processing through Pandas libraries, obtains a statistical abstract of a stress response data set through describe () function, analyzes the influence of differential stress conditions on the performance of an electronic element, draws a scatter diagram and a line diagram through matplotlib libraries, compares the change of the performance of the element under differential stress, and identifies key influence factors through calculating the change rate and fluctuation range to generate an influence factor analysis result.
And the performance prediction submodule evaluates performance attenuation trend according to the relation model of the long-term stress data and the performance change through the influence factor analysis result, draws a future performance attenuation path diagram and generates performance change trend data.
The performance prediction submodule utilizes a time sequence prediction model, adopts Python language programming, constructs a long-term and short-term memory network (LSTM) through TensorFlow and Keras libraries, sets a layer to be 3 layers, processes an influence factor analysis result, takes the influence factor analysis result as input, evaluates a performance attenuation trend through a relation model of long-term stress data and performance change, uses a common () function to set an optimizer as adam, selects a mean_squared_error through a loss function, draws a future performance attenuation path diagram, and generates performance change trend data.
Referring to fig. 5, the attenuation mode dynamic modeling module includes:
the trend analysis submodule adopts the performance change trend data of the electronic element to identify the characteristics of the early stage, the middle stage and the late stage of attenuation, analyzes the performance fluctuation in the time sequence, extracts the attenuation rate, compares the performance change among the differential elements, and generates the attenuation rate and a trend comparison analysis result;
The trend analysis submodule adopts a time sequence analysis method, utilizes Python language programming, invokes an ARIMA model (AutoRegressive Integrated Moving Average) through a Statsmodels library, sets an autoregressive item to be 2, has a differential order of 1 and a moving average item of 2, analyzes the data based on the performance change trend data of the electronic element, carries out time sequence modeling on the data of each electronic element, identifies the characteristics of initial stage, middle stage and late stage of attenuation through model fitting, draws a model diagnosis chart through a plot_diagnostics () function, analyzes performance fluctuation in the time sequence, extracts attenuation rate, compares the performance change among the differential elements through a comparison analysis function, and generates attenuation rate and trend comparison analysis results.
The attenuation path modeling submodule calibrates key nodes in the attenuation process based on attenuation rate and trend comparison analysis results, maps the relation between performance indexes and time, constructs an attenuation path frame, draws a change track of element performance along with time, and obtains a preliminary attenuation path model;
the attenuation path modeling submodule is based on a regression analysis technology of machine learning, programming is conducted through a Python language, a polynomial regression function (PolynomialFeatures) is called through a scikit-learn library, the degree of a polynomial is set to be 3, attenuation rate and trend comparison analysis results are processed, data are converted into polynomial characteristics through a fit_transform () function, key nodes in an attenuation process are calibrated, the relation between performance indexes and time is mapped, a regression model is built through the LinearRegression () function, the change track of element performance along with time is drawn, a preliminary attenuation path model is obtained, and the preliminary attenuation path model is generated.
The model verification submodule utilizes the preliminary attenuation path model to compare actual application data with model prediction data, determines a parameter adjustment direction through differential analysis, adjusts model parameters and establishes a dynamic attenuation path model.
The model verification sub-module uses a cross verification method, adopts Python language programming, calls a cross_val_score function through a scikit-learn library, sets the fold number to be 5, verifies the preliminary attenuation path model, calculates an average mean square error by comparing actual application data with model prediction data and using a mean_squared_error () function, determines a parameter adjustment direction through differential analysis, adjusts model parameters such as the degree of adjustment polynomial regression or parameters of an ARIMA model, and establishes a dynamic attenuation path model.
Referring to fig. 6, the performance prediction and failure analysis module includes:
The failure trend analysis submodule identifies key failure indexes of the electronic element based on the dynamic attenuation path model, tracks the key failure indexes, analyzes the performance attenuation trend and generates a failure trend analysis result;
The failure trend analysis submodule adopts a survival analysis method, utilizes Python language programming, calls a Kaplan-Meier fitting function through a LIFELINES library, tracks key failure indexes of the electronic element based on a dynamic attenuation path model, fits data through a fit () function, analyzes performance attenuation trend, and draws a survival curve through a plot () function, so that key failure time of the electronic element is identified, and a failure trend analysis result is generated.
The aging stage performance evaluation submodule adopts failure trend analysis results to compare performance performances of the electronic components in the differentiated aging stage, identifies key stages of performance change and generates aging stage performance evaluation results;
The aging stage performance evaluation submodule adopts a variance analysis method, utilizes Python language programming, calls f_ oneway () functions through SciPy libraries, processes failure trend analysis results, compares performance performances of electronic elements in a differential aging stage, and identifies key stages of performance change by calculating F values and p values, wherein the step helps to determine significant difference stages of performance performances in an aging process and generate aging stage performance evaluation results.
And the comprehensive failure prediction sub-module comprehensively analyzes the performance and potential failure points of the electronic element in a future time period through the performance evaluation result in the aging stage, evaluates the performance difference in the aging process and generates performance and failure prediction data.
The comprehensive failure prediction submodule utilizes a multiple regression analysis technology, adopts Python language programming, calls LinearRegression () functions through scikit-learn libraries, builds a model to comprehensively analyze performance and potential failure points of the electronic component in a future time period through performance evaluation results of an aging stage, uses fit () functions to fit the performance evaluation results and time data of the aging stage, evaluates performance differences in the aging process, predicts the future performance through predict () functions, and generates performance and failure prediction data.
Referring to fig. 7, the machine learning enhancement prediction module includes:
The aging characteristic analysis submodule carries out key index screening based on the performance and failure prediction data, identifies factors influencing the performance and service life of the electronic element, eliminates indexes with small fluctuation through variation coefficients, strengthens aging sensitive characteristics, and generates a key aging characteristic data set;
The aging characteristic analysis submodule adopts a statistical analysis method, utilizes Python language programming, performs data processing through Pandas libraries, calculates standard deviation of each index based on performance and failure prediction data by using std () function, calculates average value by mean () function, performs key index screening by calculating variation coefficient (CV=standard deviation/average value), eliminates index with small variation coefficient, strengthens aging sensitive characteristics, and generates a key aging characteristic data set.
The model refinement training sub-module carries out model structural design based on the key aging characteristic data set, adjusts a data input layer to match characteristic dimensions, enhances characteristic expression through interlayer connection, calibrates and outputs the prediction requirement of matching aging behaviors, and obtains an optimized aging behavior prediction model;
The model refinement training submodule adopts a deep learning algorithm, utilizes Python language programming, constructs a neural network model through TensorFlow and Keras libraries, adjusts a data input layer to match feature dimensions based on a key aging feature data set, uses a Sequential () function to define a model structure, adds a Dense layer, activates function enhancement feature expression through activation= 'relu', uses a rule () function to calibrate and output the prediction requirement of matching aging behaviors, sets an optimizer to 'adam', and selects 'mean_squared_error' for a loss function to generate an optimized aging behavior prediction model.
The failure risk prediction submodule is used for carrying out failure risk analysis by utilizing a Monte Carlo simulation method based on the optimized aging behavior prediction model, evaluating the aging speed and the failure probability of the element by simulating a prediction scene under the differentiated use condition, comprehensively obtaining the failure risk level of the element, and producing a refined prediction model.
The failure risk prediction submodule utilizes a Monte Carlo simulation method, adopts Python language programming, generates a prediction scene under the random number simulation differentiation use condition through a NumPy library, performs failure risk analysis on each simulation scene by using a predict () function based on the optimized aging behavior prediction model, evaluates the aging speed and the failure probability of the element through a statistical simulation result, comprehensively obtains the failure risk level of the element, and generates a refined prediction model.
The Monte Carlo simulation method adopts the formula:
；
The probability of an event occurring is calculated, wherein, For the number of times event X occurs in the simulation,/>For the total simulation times,/>Adjusting the coefficient for environmental factors, wherein S is an environmental stability index,/>For using conditional complexity coefficients, T is the total number of test scenarios,/>And U is a diversity index of user operation.
Firstly, generating random numbers to simulate prediction scenes under different use conditions by using a Monte Carlo simulation method, and calculatingAnd/>To estimate the probability of event occurrence, and then introduce an environmental factor adjustment coefficient/>Adjusting success times according to the environmental stability index S, and using the condition complexity coefficient/>And testing the total number of scenes T to consider the complexity of the use condition, the user interaction frequency coefficient/>And the diversity index U of the user operation is used for adjusting the total simulation times so as to more accurately reflect the probability under the actual condition.
Referring to fig. 8, the aging path optimization module includes:
The aging process analysis submodule identifies key time points and condition changes in the aging process based on the refined prediction model, analyzes the condition changes affecting the element performance, selects aging acceleration and slowing factors and generates an aging key point analysis result;
The aging process analysis submodule adopts a differential analysis method, utilizes Python language programming, carries out numerical calculation through Numpy libraries, identifies key time points and condition changes in the aging process by calculating the difference of performance indexes at different time points based on a refined prediction model, selects factors for accelerating and slowing aging through a variation coefficient (standard deviation divided by average value), and generates an aging key point analysis result.
The strategy making submodule makes preventive measures oriented to a key aging stage based on an aging key point analysis result, designs an intervention strategy to adjust element use conditions or physical structures, and obtains a preliminary aging coping strategy set;
the strategy making submodule adopts a decision tree analysis method, utilizes Python language programming, invokes DecisionTreeClassifier functions through a Scikit-learn library, builds a decision tree model based on the analysis result of the aging key points to make precautions for the key aging stage, designs intervention strategies to adjust the use conditions or physical structures of elements, and generates a preliminary aging coping strategy set.
The strategy effect prediction submodule simulates aging path change after strategy implementation based on the preliminary aging coping strategy set, evaluates influence of preventive measures on the service life of the extension element, screens an effect optimal strategy and obtains a strategy effect evaluation result;
The strategy effect prediction submodule utilizes a simulation modeling technology, adopts Python language programming, performs process simulation through SimPy libraries, and based on a preliminary aging coping strategy set, simulates aging path change after strategy implementation, evaluates influence of preventive measures on service life of an extension element, screens an effect optimal strategy and generates a strategy effect evaluation result.
The optimizing strategy integration sub-module optimizes and integrates strategies based on strategy effect evaluation results, identifies advantages and limitations of each strategy by analyzing benefit and cost data of multiple strategies, classifies and distributes weights of the strategies, combines ageing characteristics of electronic elements and expected service life targets, and establishes a comprehensive ageing optimization scheme to produce an ageing optimization strategy set.
The optimizing strategy integration submodule adopts a multi-criterion decision analysis technology, utilizes Python language programming, carries out strategy evaluation and selection through a PyMCDA library, based on a strategy effect evaluation result, identifies advantages and limitations of each strategy by analyzing benefit and cost data of multiple strategies, classifies and weights the strategies, combines ageing characteristics of electronic elements and expected service life targets, formulates a comprehensive ageing optimization scheme and generates an ageing optimization strategy set.
Referring to fig. 9, the failure point positioning module includes:
The aging process tracking submodule is used for recording performance parameter changes and environmental conditions of the element in a plurality of aging stages based on an aging optimization strategy set, identifying and recording key links causing performance degradation, and generating an aging track record;
The aging process tracking submodule adopts a time sequence analysis method, utilizes Python language programming, performs data processing through Pandas libraries, records performance parameter changes and environmental conditions of elements in a plurality of aging stages based on an aging optimization strategy set, calculates moving averages by using rolling () functions to smooth short-term fluctuation, and diff () functions identify the change trend of the performance parameters, identify and record key links causing performance degradation, and generate aging track records.
The failure point presetting submodule analyzes key time of performance degradation based on the aging track record, establishes a reason behind the failure point, presets the predicted failure point and acquires preset failure point analysis;
The failure point presetting submodule adopts a regression analysis method, utilizes Python language programming, calls LinearRegression functions through a Scikit-learn library, analyzes key moments of performance degradation based on aging track records, establishes a linear relation between performance parameters and time through a fit () function, predicts a time point when a future performance index reaches a critical value, establishes a back reason, presets the predicted failure point, and obtains preset failure point analysis.
The data positioning integration submodule carries out logic judgment and rule matching on key factors and preset failure points in the aging process based on preset failure point analysis, evaluates the correlation between the key factors and the preset failure points, integrates analysis and matching results, and constructs a connection map between the failure points and the aging factors to obtain failure point positioning data.
The data positioning integration sub-module adopts a logistic regression analysis method, utilizes Python language programming, invokes LogisticRegression functions through a Scikit-learn library, carries out logic judgment and rule matching on key factors and preset failure points in the aging process based on preset failure point analysis, establishes a logic relation between the aging factors and the failure points by a fit () function, evaluates the correlation between the key factors and the preset failure points by a predict _ proba () function, integrates analysis and matching results, and constructs a connection map between the failure points and the aging factors to obtain failure point positioning data.
The present invention is not limited to the above embodiments, and any equivalent embodiments which can be changed or modified by the technical disclosure described above can be applied to other fields, but any simple modification, equivalent changes and modification made to the above embodiments according to the technical matter of the present invention will still fall within the scope of the technical disclosure.
Claims (9)
1. An intelligent burn-in system for electronic components, the system comprising:
The nonlinear attenuation behavior recognition module is used for analyzing performance fluctuation and variation trend by collecting electronic element operation data and recognizing an abnormal attenuation mode in an aging process to obtain a nonlinear attenuation characteristic set;
the stress simulation and effect analysis module simulates the response of the electronic element based on the nonlinear attenuation characteristic set, analyzes the influence of stress conditions on the element performance, and obtains performance change trend data;
the attenuation mode dynamic modeling module constructs a dynamic model of the electronic element aging process according to the performance change trend data, acquires a performance attenuation path of the element along with time, and forms a dynamic attenuation path model;
the performance prediction and failure analysis module predicts the performance and failure point of the electronic element based on the dynamic attenuation path model, analyzes the performance of the element in the differential aging stage, and outputs performance and failure prediction data;
The machine learning enhancement prediction module optimizes the prediction model through the analysis performance and failure prediction data, maps the aging behavior and failure risk of the electronic element and outputs a refined prediction model;
the aging path optimization module analyzes the aging process of the electronic element by utilizing the refined prediction model, and proposes a coping strategy to obtain an aging optimization strategy set;
and the failure point positioning module analyzes key links in the element aging process based on the aging optimization strategy set, sets failure points and obtains failure point positioning data.
2. The electronic component intelligent aging test system according to claim 1, wherein the nonlinear attenuation characteristic set includes characteristics of response time extension, power consumption increase proportion and performance attenuation rate of the component at ambient temperature, the performance change trend data includes data of performance decrease speed under temperature stress, electrical characteristic change under humidity condition and life prediction under mechanical stress influence, the dynamic attenuation path model includes contents of time sequence analysis results, performance decrease key nodes and aging acceleration factor influence evaluation, the performance and failure prediction data includes performance stability prediction in short term, middle-long term failure time prediction and performance decrease map of differential aging stage, the refinement prediction model is specifically a neural network model trained based on historical data, regression analysis results of aging behaviors and quantitative evaluation indexes of failure risk, the aging optimization policy set includes temperature control optimization, humidity management scheme and current load adjustment policy, and the failure point positioning data includes time marks of aging key points, performance decrease speed threshold and safety working range of key parameters.
3. The intelligent burn-in system of claim 1 wherein said nonlinear decay behavior identification module comprises:
The fluctuation analysis submodule carries out fluctuation condition analysis based on the operation data of the electronic element, measures the fluctuation amplitude and the frequency of multiple parameters, determines the performance stability and generates a fluctuation analysis result;
The trend prediction submodule adopts the fluctuation analysis result, performs change trend analysis by using a linear regression algorithm, identifies the problem trend of performance reduction or stability, predicts future performance change and generates a change trend prediction result;
the abnormal pattern recognition submodule captures the change trend according to the change trend prediction result, predicts the performance change according to the change trend, extracts a performance attenuation signal from the prediction result, recognizes a nonlinear attenuation pattern, sets abnormal attenuation characteristics according to nonlinear characteristics, and generates a nonlinear attenuation characteristic set.
4. The intelligent burn-in system of claim 3 wherein said linear regression algorithm employs the formula:
；
calculating a fluctuation amplitude variation based on the time-series data, wherein, Is time series data,/>Is a square term of time series data to increase the nonlinear capturing capability of the model to time trends,/>Is natural logarithmic transformation of time series data to reduce bias of data,/>Is an external influencing factor, including market emotion index,/>For intercept,/>Is a linear coefficient of time series data,/>Coefficients that are square terms of time series data,/>Coefficient of natural logarithm of time series data,/>Is the coefficient of external influence factor,/>Is an error term.
5. The electronic component intelligent burn-in system of claim 1, wherein said decay pattern dynamic modeling module comprises:
the trend analysis submodule adopts the performance change trend data of the electronic element to identify characteristics of early, middle and late attenuation, analyzes performance fluctuation in a time sequence, extracts attenuation rate, compares performance change among differential elements, and generates attenuation rate and trend comparison analysis results;
The attenuation path modeling submodule calibrates key nodes in the attenuation process based on the attenuation rate and the trend comparison analysis result, maps the relation between performance indexes and time, constructs an attenuation path frame, draws the change track of element performance along with the time, and obtains a preliminary attenuation path model;
The model verification submodule compares actual application data with model prediction data by using the preliminary attenuation path model, determines a parameter adjustment direction through differential analysis, adjusts model parameters and establishes a dynamic attenuation path model.
6. The electronic component intelligent burn-in system of claim 1 wherein said machine learning enhancement prediction module comprises:
The aging characteristic analysis submodule carries out key index screening based on the performance and failure prediction data, identifies factors influencing the performance and service life of the electronic element, eliminates indexes with small fluctuation through variation coefficients, strengthens aging sensitive characteristics, and generates a key aging characteristic data set;
the model refinement training sub-module carries out model structural design based on the key aging characteristic data set, adjusts a data input layer to match characteristic dimensions, enhances characteristic expression through interlayer connection, calibrates and outputs the prediction requirement of matching aging behaviors, and obtains an optimized aging behavior prediction model;
and the failure risk prediction submodule is used for carrying out failure risk analysis by utilizing a Monte Carlo simulation method based on the optimized aging behavior prediction model, evaluating the aging speed and the failure probability of the element by simulating a prediction scene under the differentiated use condition, comprehensively obtaining the failure risk level of the element, and producing a refined prediction model.
7. The intelligent burn-in system of claim 6 wherein said monte carlo simulation method uses the formula:
；
The probability of an event occurring is calculated, wherein, For the number of times event X occurs in the simulation,/>As a result of the total number of simulations,Adjusting the coefficient for environmental factors, wherein S is an environmental stability index,/>For using conditional complexity coefficients, T is the total number of test scenarios,/>And U is a diversity index of user operation.
8. The electronic component intelligent burn-in system of claim 1, wherein said burn-in path optimization module comprises:
the aging process analysis submodule identifies key time points and condition changes in the aging process based on the refined prediction model, analyzes the condition changes affecting the element performance, selects aging acceleration and slowing factors and generates an aging key point analysis result;
The strategy making submodule makes preventive measures oriented to a key aging stage based on the aging key point analysis result, designs an intervention strategy to adjust the use condition or the physical structure of the element, and obtains a preliminary aging coping strategy set;
The strategy effect prediction submodule simulates aging path change after strategy implementation based on the preliminary aging coping strategy set, evaluates influence of preventive measures on the service life of the extension element, screens an effect optimal strategy and obtains a strategy effect evaluation result;
And the optimization strategy integration submodule optimizes and integrates strategies based on the strategy effect evaluation result, identifies the advantages and limitations of each strategy by analyzing benefit and cost data of multiple strategies, classifies and distributes weights of the strategies, combines the aging characteristics of the electronic components and the expected service life targets, and establishes a comprehensive aging optimization scheme to produce an aging optimization strategy set.
9. The electronic component intelligent burn-in system of claim 1, wherein said point of failure localization module comprises:
The aging process tracking submodule is used for recording performance parameter changes and environmental conditions of the element in a plurality of aging stages based on the aging optimization strategy set, identifying and recording key links causing performance degradation, and generating an aging track record;
the failure point presetting submodule analyzes key time of performance reduction based on the aging track record, establishes a reason behind the failure point, presets a predicted failure point and acquires preset failure point analysis;
And the data positioning and integrating sub-module carries out logic judgment and rule matching on key factors and preset failure points in the aging process based on the analysis of the preset failure points, evaluates the correlation between the key factors and the preset failure points, integrates analysis and matching results, and constructs a connection map between the failure points and the aging factors to obtain failure point positioning data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410425973.3A CN118035848B (en) | 2024-04-10 | Intelligent aging test system for electronic element |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410425973.3A CN118035848B (en) | 2024-04-10 | Intelligent aging test system for electronic element |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118035848A true CN118035848A (en) | 2024-05-14 |
| CN118035848B CN118035848B (en) | 2024-07-16 |
Family
ID=
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109388830A (en) * | 2017-08-10 | 2019-02-26 | 湖南中车时代电动汽车股份有限公司 | A kind of circuit board life-span prediction method |
| US20210165941A1 (en) * | 2018-06-19 | 2021-06-03 | Proteantecs Ltd. | Efficient integrated circuit simulation and testing |
| CN115906573A (en) * | 2022-11-29 | 2023-04-04 | 中车大连电力牵引研发中心有限公司 | PCB service life analysis method based on reliability analysis |
| US11790127B1 (en) * | 2019-05-11 | 2023-10-17 | Synopsys, Inc. | Full correlation aging analysis over combined process voltage temperature variation |
| WO2023227071A1 (en) * | 2022-05-25 | 2023-11-30 | 中国电子科技集团公司第十研究所 | Multi-model fused avionic product health assessment method |
| CN117318255A (en) * | 2023-11-30 | 2023-12-29 | 北京中铁建电气化设计研究院有限公司 | Battery state analysis system and method based on big data visualization |
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109388830A (en) * | 2017-08-10 | 2019-02-26 | 湖南中车时代电动汽车股份有限公司 | A kind of circuit board life-span prediction method |
| US20210165941A1 (en) * | 2018-06-19 | 2021-06-03 | Proteantecs Ltd. | Efficient integrated circuit simulation and testing |
| US11790127B1 (en) * | 2019-05-11 | 2023-10-17 | Synopsys, Inc. | Full correlation aging analysis over combined process voltage temperature variation |
| WO2023227071A1 (en) * | 2022-05-25 | 2023-11-30 | 中国电子科技集团公司第十研究所 | Multi-model fused avionic product health assessment method |
| CN115906573A (en) * | 2022-11-29 | 2023-04-04 | 中车大连电力牵引研发中心有限公司 | PCB service life analysis method based on reliability analysis |
| CN117318255A (en) * | 2023-11-30 | 2023-12-29 | 北京中铁建电气化设计研究院有限公司 | Battery state analysis system and method based on big data visualization |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112202736B (en) | Communication network anomaly classification method based on statistical learning and deep learning | |
| CN111222549B (en) | Unmanned aerial vehicle fault prediction method based on deep neural network | |
| CN110084610B (en) | Network transaction fraud detection system based on twin neural network | |
| CN107725283A (en) | A kind of fan trouble detection method based on depth belief network model | |
| CN109800875A (en) | Chemical industry fault detection method based on particle group optimizing and noise reduction sparse coding machine | |
| CN106656357B (en) | Power frequency communication channel state evaluation system and method | |
| CN117633722B (en) | Detection control method and system based on intelligent detection robot | |
| CN110083065B (en) | Self-adaptive soft measurement method based on flow type variational Bayesian supervised factor analysis | |
| CN117421684B (en) | Abnormal data monitoring and analyzing method based on data mining and neural network | |
| CN117611015B (en) | Real-time monitoring system for quality of building engineering | |
| CN117113729A (en) | Digital twinning-based power equipment online state monitoring system | |
| CN118035848B (en) | Intelligent aging test system for electronic element | |
| CN117574264A (en) | Transformer fault diagnosis method and system based on knowledge constraint neural network | |
| CN117669384A (en) | Intelligent monitoring method and system for temperature sensor production based on Internet of things | |
| CN118035848A (en) | Intelligent aging test system for electronic element | |
| CN110956112B (en) | Novel high-reliability slewing bearing service life assessment method | |
| CN113126489A (en) | CNN-GRU-BINN-based intelligent BIT design method for heavy-duty gas turbine control system | |
| CN115033893B (en) | Information vulnerability data analysis method of improved clustering algorithm | |
| CN114357864A (en) | Phase modulator state evaluation method and evaluation system based on fuzzy reasoning | |
| CN115423370A (en) | Relay protection equipment health state assessment method and device | |
| CN113051809A (en) | Virtual health factor construction method based on improved restricted Boltzmann machine | |
| CN112862180A (en) | Denitration system inlet NOx concentration prediction method | |
| CN118014373B (en) | Risk identification model based on data quality monitoring and construction method thereof | |
| CN113159131B (en) | Hierarchical prediction method and hierarchical prediction system for running conditions of bioreactor | |
| CN117724348B (en) | Accurate pressure regulation and control system based on explosion testing machine |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |