CN118330447B - Semiconductor integrated circuit test system - Google Patents
Semiconductor integrated circuit test system Download PDFInfo
- Publication number
- CN118330447B CN118330447B CN202410760811.5A CN202410760811A CN118330447B CN 118330447 B CN118330447 B CN 118330447B CN 202410760811 A CN202410760811 A CN 202410760811A CN 118330447 B CN118330447 B CN 118330447B
- Authority
- CN
- China
- Prior art keywords
- test
- data
- represented
- signal
- parameter
- 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.)
- Active
Links
- 238000012360 testing method Methods 0.000 title claims abstract description 258
- 239000004065 semiconductor Substances 0.000 title claims abstract description 48
- 238000000034 method Methods 0.000 claims abstract description 62
- 230000008569 process Effects 0.000 claims abstract description 59
- 238000001914 filtration Methods 0.000 claims abstract description 35
- 238000013528 artificial neural network Methods 0.000 claims abstract description 34
- 238000012545 processing Methods 0.000 claims abstract description 23
- 238000012937 correction Methods 0.000 claims abstract description 15
- 230000003044 adaptive effect Effects 0.000 claims abstract description 12
- 238000001514 detection method Methods 0.000 claims abstract description 12
- 230000001105 regulatory effect Effects 0.000 claims abstract description 6
- 238000005516 engineering process Methods 0.000 claims abstract description 5
- 230000007613 environmental effect Effects 0.000 claims description 73
- 238000004364 calculation method Methods 0.000 claims description 59
- 238000004422 calculation algorithm Methods 0.000 claims description 39
- 230000006870 function Effects 0.000 claims description 33
- 238000003745 diagnosis Methods 0.000 claims description 32
- 230000004044 response Effects 0.000 claims description 21
- 238000010801 machine learning Methods 0.000 claims description 18
- 238000004458 analytical method Methods 0.000 claims description 17
- 230000000875 corresponding effect Effects 0.000 claims description 17
- 239000011159 matrix material Substances 0.000 claims description 13
- 238000010606 normalization Methods 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 6
- 238000013101 initial test Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 125000004122 cyclic group Chemical group 0.000 claims description 5
- 230000009471 action Effects 0.000 claims description 4
- 230000000306 recurrent effect Effects 0.000 claims description 4
- 101100001674 Emericella variicolor andI gene Proteins 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims description 3
- 238000013459 approach Methods 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000012706 support-vector machine Methods 0.000 claims description 3
- 230000001276 controlling effect Effects 0.000 abstract description 3
- 238000005259 measurement Methods 0.000 abstract description 3
- 230000009467 reduction Effects 0.000 abstract description 2
- 230000036962 time dependent Effects 0.000 description 6
- 238000004590 computer program Methods 0.000 description 4
- 230000007547 defect Effects 0.000 description 2
- 230000005672 electromagnetic field Effects 0.000 description 2
- 238000012905 input function Methods 0.000 description 2
- 238000004806 packaging method and process Methods 0.000 description 2
- 239000003990 capacitor Substances 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Landscapes
- Tests Of Electronic Circuits (AREA)
- Testing Or Measuring Of Semiconductors Or The Like (AREA)
Abstract
The invention relates to the technical field of semiconductors, in particular to a semiconductor integrated circuit testing system, which comprises an EMI and RFI filtering compensation module: using an adaptive filtering technology of an active noise control ANC or an adaptive noise cancellation ANE to carry out filtering processing on test data containing EMI and RFI signals which are actually received so as to reduce interference, outputting the filtered test data and transmitting the test data to a neural network prediction calibration module; according to the invention, through arranging the EMI and RFI filtering compensation module, the environment parameter regulation and control module and the to-be-tested position detection error correction module, the problems of electromagnetic interference or radio frequency interference influence, unstable influence of a test environment and test position error influence in the test process are comprehensively avoided, the measurement result of interference reduction by effective filtering is realized, the accuracy of measurement data is improved, the function of regulating and controlling the environment parameters of temperature and humidity in real time is realized, and the stability of the test environment is ensured.
Description
Technical Field
The invention relates to the technical field of semiconductors, in particular to a semiconductor integrated circuit testing system.
Background
The semiconductor integrated circuit is an integrated circuit integrating a plurality of electronic devices such as transistors, resistors, capacitors and the like on a single semiconductor material, can realize complex functions, is an important component in modern electronic equipment, and a semiconductor integrated circuit test system is used for testing various functionalities and reliability of the semiconductor integrated circuit.
For example, in the patent application of CN116184177a with publication date of 2023, 05 and 30, and entitled "test system for semiconductor integrated circuit package", the signal input module is used to input test signals to the test board, the data acquisition module is used to acquire information data of the circuit board, each pin of the test board and the connection socket, the pin alignment detection module is used to detect and acquire the alignment connection data of the pin and the connection socket, the pin data extraction module is used to extract and acquire the received data of the pin of the package circuit board, the control module is used to control the operation of each module in the test system, analyze the detected data of the circuit board, determine whether the package of the circuit board is qualified, the threat signal output acquisition module is used for acquiring and outputting threat signal data in a test process, the data calculation module is used for substituting threat signal data into a threat calculation strategy to calculate threat values of threat signals, the packaging quality of the circuit board is rapidly tested, the accuracy and the testing efficiency of the packaging quality test of the circuit board are effectively improved, the primary total threat data and the secondary total threat data are respectively calculated, the primary total threat data and the secondary total threat data are added to obtain real threat data, then the real threat data are compared with a set real threat data threshold value, the testing accuracy is further improved, but external electromagnetic interference or radio frequency interference signal interference exists, and the defect that the test result is unstable is caused by temperature change in a test environment.
The prior art has the following defects: the existing semiconductor integrated circuit test system may be interfered by external signals of electromagnetic interference or radio frequency interference, so that the test result is unstable; temperature variation in the test environment may affect the performance of the test equipment, resulting in inaccurate test results, and the sensors in the test system are used for detecting parameters of temperature, voltage and current, but if the sensor calibration is inaccurate before detection, the test results will deviate, and the test process may be affected by mechanical vibration, resulting in inaccurate test positions, thereby affecting the accuracy of the test results.
The above information disclosed in the background section is only for enhancement of understanding of the background of the disclosure and therefore it may include information that does not form the prior art that is already known to a person of ordinary skill in the art.
Disclosure of Invention
The invention aims to provide a semiconductor integrated circuit testing system, which increases the functions of filtering electromagnetic interference and radio frequency interference, testing the stability of the environment and correcting the accuracy of the measured position by arranging an EMI and RFI filtering compensation module, an environment parameter regulation module, a position detection error correction module to be tested, a neural network prediction calibration module and an analysis and diagnosis fault response module so as to solve the problems in the background technology.
In order to achieve the above object, the present invention provides the following technical solutions: a semiconductor integrated circuit testing system comprising an EMI and RFI filter compensation module: using an adaptive filtering technology of an active noise control ANC or an adaptive noise cancellation ANE to carry out filtering processing on test data containing EMI and RFI signals which are actually received so as to reduce interference, outputting the filtered test data and transmitting the test data to a neural network prediction calibration module;
An environmental parameter regulation and control module: receiving environmental parameter data of temperature and humidity from an intelligent sensor, adjusting the temperature and the humidity by using a PID control algorithm according to the environmental parameter data so as to adjust a test environment, obtaining constant environmental state data, and transmitting the constant environmental state data to a neural network prediction calibration module;
The position detection error correction module to be detected: collecting dynamic mechanical position data point signals of which the test points are affected by vibration from an intelligent sensor, correcting errors of the test positions by using a gradient descent optimization algorithm to obtain corrected accurate test position signals, and transmitting the accurate test position signals to a neural network prediction calibration module;
Test parameter acquisition processing module: receiving test data after the EMI and RFI filtering compensation module carries out filtering treatment, carrying out constant environmental state data after the environmental parameter regulation and control module carries out the environmental parameter regulation and control, carrying out accurate test position signals after the position error correction by the position detection error correction module to be tested, carrying out preprocessing by using a characteristic extraction and data standardization algorithm, obtaining preprocessed test parameter signals, and transmitting the preprocessed test parameter signals to the neural network prediction calibration module;
The neural network prediction calibration module: receiving the test parameter signals processed by the test parameter acquisition processing module, predicting and calibrating the test parameter signals by using a cyclic neural network (RNN) model to obtain fault hidden danger signals, and transmitting the fault hidden danger signals to the analysis and diagnosis fault response module;
An analysis and diagnosis fault response module: and receiving the predicted fault hidden danger signals in the neural network prediction calibration module, establishing a machine learning diagnosis model, analyzing and diagnosing according to the fault hidden danger signals, generating a diagnosis report and corresponding response measures, and feeding back to an operator for early warning response.
Optionally, the filtering processing steps of the ANC are as follows:
Capturing electromagnetic interference EMI and radiofrequency interference RFI noise in the background using electromagnetic field detectors during semiconductor integrated circuit testing, calibrated as At the same time, obtain the original test data, calibrated as；
Generating a reference signal according to the detected noise signal, and generating a signal with the same amplitude as the reference signal but opposite in phase by using an adaptive filtering algorithm to obtain an inverted phase signal, wherein the calculation formula of the adaptive filtering algorithm is as followsAnd (2) andIn which, in the process,Represented as at time stepsThe filter coefficients of the time-dependent filter coefficients,Expressed as time stepsThe filter coefficients of the time-dependent filter coefficients,Represented by the rate of learning,Represented as an error signal and is provided by a reference signal,Represented as at time stepsNoise signal in time input signalAnd raw test dataIs used as an input function of the (c),Represented by the time step(s) indicated,Represented as a transpose,Represented as a standard matrix of the type described above,Represented as at time stepsA one-dimensional column vector matrix of the input signal;
adding the inverted signal to the original test data, i.e Signals for eliminating EMI and RFI noise portions to obtain filtered test data。
Optionally, the step of adjusting the environmental parameter data by the PID control algorithm is as follows:
The real-time acquisition of the environmental parameter data of the actual temperature and humidity of the semiconductor integrated circuit in the test process from the intelligent sensor is calibrated as And set target values of temperature and humidity in a suitable test environment for the semiconductor integrated circuit, calibrated as；
Calculating the deviation between the actual environment parameter data and the set target value, wherein the calculation formula of the deviation is as followsIn which, in the process,Represented as a result of a deviation between the actual environmental parameter data and the set target value,Expressed as the temperature in the target valueTemperature of the actual environmental parameterThe absolute value of the difference between them,Expressed as humidity in target valueHumidity with actual environmental parametersThe absolute value of the difference between them;
According to the deviation result Calculating PID control action comprising three parts of proportion, integral and derivative, respectively calibrated asThe calculation formula of the PID control quantity isIn which, in the process,Represented as a PID control quantity,Gain values expressed as proportional, integral, differential,Represented by the time step(s) indicated,Represented as at time stepsDeviation results at time；
According to PID control quantityThe output power control element of the heater or cooler, humidifier or dryer is adjusted to dynamically regulate temperature and humidity data, so that the regulated environmental parameters approach the target values.
Optionally, the logic for obtaining the constant environmental state data is as follows:
by adjusting PID control quantity, the deviation result is obtained Approaching zero, when the environment state dataWhen reaching a steady-state point, continuously acquiring environmental parameter data through an intelligent sensor to acquire constant environmental state data;
The constant environmental state data is calculated as In which, in the process,Represented as constant environmental state data,Represented as atEnvironmental parameter data acquired by the intelligent sensor in real time,Represented as being infinitely close to the symbol.
Optionally, the error correction step of the gradient descent optimization algorithm on the test position is as follows:
Semiconductor integrated circuit currently using intelligent sensor acquisition test instrument The test point position signal at the moment is calibrated asAnd sets a position signal of an ideal test as；
According to the currentTest point position signal at timeAnd position signal for ideal testBetween, set an error function, calibrate asThe calculation formula of the error function isIn which, in the process,Expressed as presentTest point position signal at timeAnd position signal for ideal testThe value of the error between them,Represented as in a position signalThe difference in the coordinate points of the axes,Represented as in a position signalDifference of the axis coordinate points;
Selection of Position signal of initial test point at momentIn combination with the currentTest point position signal at timeCalculating the gradient of the error function, and calibrating asThe calculation formula of the gradient isAnd (2) andIn which, in the process,Expressed as presentTest point position signal at timeTo the point ofPosition signal of initial test point at momentIs used to determine the error value of (a),Represented as initialFrom moment to momentA time period of the moment;
Updating the position signal of the test point by using a gradient descent algorithm to reduce errors, and updating the calculation formula of the position signal of the test point to be In which, in the process,Is shown as the nextUpdating the calculated position signal at the moment in time,Expressed as a learning rate;
repeating the steps of calculating the gradient value and updating the position signal of the test point until the maximum iteration number is reached;
the acquisition logic for the accurate test position signal is as follows:
continuously updating the test point position through a gradient descent algorithm To reduce the error function value；
According to the error valueThe iteration calculation is reduced to analyze, and whether the updating of the position of the test point is in a stable state or not is checked;
when the error falls within an acceptable range with iterative calculations, i.e When the test point is updated, the updated test point position is regarded as the accurate position, and an updated accurate test position signal is output and calibrated as。
Optionally, the processing steps of the feature extraction are as follows:
Collecting test data after filtering processing of EMI and RFI filtering compensation module Constant environmental state dataAnd accurate test position signalAnd identifying and analyzing characteristic data related to the test parameter signal;
describing test data using polynomial regression algorithm Constant environmental state dataAnd accurate test position signalRelation to test parameter signals and using test dataConstant environmental state dataAnd accurate test position signalFitting polynomial regression algorithm by influence coefficient of (2) to obtain characteristic data by combination or conversion, and calibrating asThe polynomial regression formula isAnd (2) andIn which, in the process,Represented as a signal of a test parameter,Expressed as test dataThe regression coefficient of the corresponding effect is used,Represented as constant environmental state dataThe regression coefficient of the corresponding effect is used,Expressed as an accurate test position signalThe regression coefficient of the corresponding effect is used,Represented as error terms;
the calculation formula of the characteristic data is as follows In which, in the process,Represented as an exponential function of the number of words,Represented as slave test dataConstant environmental state dataAnd accurate test position signalBased on regression coefficientsAnd extracted characteristic coefficients.
Optionally, the data normalization processing steps are as follows:
By combining characteristic data Calculation of feature data using data normalization algorithmMean of (2)And standard deviationThen, carrying out standardized calculation to obtain standardized data, and calibrating asWherein the mean valueThe calculation formula of (2) isAnd (2) andIn which, in the process,Represented as characteristic dataIs used to determine the number of samples in the sample,Expressed as a total sampleMiddle (f)Corresponding characteristic data; Standard deviation ofThe calculation formula of (2) is；
Normalization of dataThe calculation formula of (2) isIn which, in the process,Represented as characteristic data is provided for the purpose of,Represented as characteristic dataIs used for the average value of (a),Represented as characteristic dataStandard deviation of (a) calculated standardized dataFor testing parameter signalsIs input after pretreatment of (a) and testing the parameter signal.
Optionally, the predicting step of the recurrent neural network RNN model is as follows:
collecting test parameter signals after pretreatment by a test parameter acquisition processing module And transmitted into an input layer in the recurrent neural network RNN;
the RNN layer signals test parameters in the input layer through recursive connection Processing, the calculation formula of the RNN layer is as followsIn which, in the process,Expressed as at timeIs used to determine the hidden state of the (c),Represented as an activation function,Expressed as the last timeIs used to determine the hidden state of the (c),Expressed as to the last timeIs hidden state of (a)Is used for the weight matrix of the (c),Represented as a signal to test parametersAt the time ofIs used to determine the weight matrix of the input values of (c),Represented as test parameter signalsAt the time ofIs used to determine the input value of (a),Bias terms denoted RNN layer;
in the output layer of the RNN network, the hidden state is converted into an output predicted value and calibrated as The calculation formula of the prediction output isIn which, in the process,Expressed as at timeIs hidden state of (a)Is used for the weight matrix of the (c),Represented as a bias term for the output layer,Expressed as at timeOutput predictive value of (2);
optionally, the logic for acquiring the fault hidden danger signal is as follows:
by prediction of neural networks And actual timeIs a test parameter signal of (a)The difference between the two signals is used for determining the prediction error, obtaining the fault hidden danger signal and calibrating asThe calculation formula of the fault hidden danger signal isIn which, in the process,Prediction results expressed as RNNAnd actual timeIs a test parameter signal of (a)Is used for detecting the fault hidden trouble signal of the vehicle,Prediction results expressed as RNN，Represented as test parameter signalsAt the time ofIs used to determine the actual value of (c) in the (c),Represented as absolute value symbols;
Defining an error threshold for testing a semiconductor integrated circuit fault, calibrated as Comparing the fault hidden danger signal with the error threshold value whenJudging that the tested semiconductor integrated circuit has faults or anomalies;
according to two cases of semiconductor integrated circuit, including The types of faults are further classified, the occurrence time points of the faults and the occurrence types of the faults are marked, and analysis results are transmitted to an analysis and diagnosis fault response module for analysis and diagnosis.
Optionally, the machine learning diagnostic model is established as follows:
Collecting fault hidden trouble signals As a training set for a machine learning diagnostic model;
training a machine learning diagnosis model by using a support vector machine (SNM) algorithm, and performing fault hidden danger signal Maximizing edges, further constructing a predictive objective function, and calibrating to beThe system is used for carrying out predictive diagnosis on the type of the fault;
the calculation formula of the objective function is In which, in the process,Represented as fault hidden trouble signalThrough the normal vector of the SVM hyperplane,The bias term denoted as SVM,Represented as a minimization function of the SVM,Represented as a regularization parameter,Respectively denoted as the firstCorresponding fault hidden trouble signal in each sampleFeature vector of (2), and fault hidden trouble signalIs a total sample number of (1);
by setting constraint conditions and calibrating as Then further determine the correct classification of the type of fault, then the constraint is expressed asAnd (2) andIs a positive integer, wherein,Represented as normal vectorIs a transposed calculation of (a),Represented as at the firstCorresponding fault hidden trouble signal in each sampleIs used for the feature vector of (a),Represented as consecutive sample points;
And deploying the trained machine learning diagnosis model into a test system to detect and diagnose faults in real time.
In the technical scheme, the invention has the technical effects and advantages that:
According to the invention, by arranging the EMI and RFI filtering compensation module, the environment parameter regulation and control module and the to-be-tested position detection error correction module, the problems of electromagnetic interference or radio frequency interference influence, unstable influence of a test environment and test position error influence in the test process are comprehensively avoided, the measurement result of interference reduction by effective filtering is realized, the accuracy of measured data is improved, the function of regulating and controlling the environment parameters of temperature and humidity in real time is realized, the stability of the test environment is ensured, the function of correcting the test position error is realized, the precision of the test position is improved, and the quality of the whole test is improved;
And a neural network prediction calibration module and an analysis diagnosis fault response module are adopted, the strong prediction capability of the neural network is utilized, the adaptability and the accuracy of the test are enhanced according to the history and real-time data prediction and calibration test process, the test system is further enabled to be capable of measuring the performance and potential faults of the semiconductor integrated circuit more accurately, the possibility of error monitoring and misdiagnosis is reduced, and the speed and the efficiency of the test flow are also improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments described in the present application, and other drawings may be obtained according to these drawings for those skilled in the art.
FIG. 1 is a block diagram of a semiconductor integrated circuit testing system according to the present invention.
Detailed Description
Example embodiments will now be described more fully with reference to the accompanying drawings. However, the exemplary embodiments may be embodied in many forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the example embodiments to those skilled in the art.
The invention provides a semiconductor integrated circuit testing system as shown in fig. 1, which comprises an EMI and RFI filtering compensation module: using an adaptive filtering technology of an active noise control ANC or an adaptive noise cancellation ANE to carry out filtering processing on test data containing EMI and RFI signals which are actually received so as to reduce interference, outputting the filtered test data and transmitting the test data to a neural network prediction calibration module;
specifically, the steps of the filtering processing of the ANC are as follows:
Capturing electromagnetic interference EMI and radiofrequency interference RFI noise in the background using electromagnetic field detectors during semiconductor integrated circuit testing, calibrated as At the same time, obtain the original test data, calibrated as；
Generating a reference signal according to the detected noise signal, and generating a signal with the same amplitude as the reference signal but opposite in phase by using an adaptive filtering algorithm to obtain an inverted phase signal, wherein the calculation formula of the adaptive filtering algorithm is as followsAnd (2) andIn which, in the process,Represented as at time stepsThe filter coefficients of the time-dependent filter coefficients,Expressed as time stepsThe filter coefficients of the time-dependent filter coefficients,Represented by the rate of learning,Represented as an error signal and is provided by a reference signal,Represented as at time stepsNoise signal in time input signalAnd raw test dataIs used as an input function of the (c),Represented by the time step(s) indicated,Represented as a transpose,Represented as a standard matrix of the type described above,Represented as at time stepsA one-dimensional column vector matrix of the input signal;
adding the inverted signal to the original test data, i.e Signals for eliminating EMI and RFI noise portions to obtain filtered test data。
Specifically, the steps of the filtering process of the ANE are as follows:
during testing of semiconductor integrated circuits, environmental noise is collected by a noise sensor while test data containing EMI and RFI signals is acquired ；
The filter is arranged, the filter parameters are adjusted according to the real-time data to optimally match the noise characteristics, and the filter output is calculated in real time and calibrated asIs calibrated with the expected output asError betweenThen；
According to the errorUpdating the coefficients of the filter so that the output is closer to the desired signal, and adjusting the filter parameters using the NLMS algorithm, the NLMS calculation formula isIn which, in the process,Represented as at time stepsThe filter coefficients of the time-dependent filter coefficients,Expressed as time stepsThe filter coefficients of the time-dependent filter coefficients,Represented as an error signal and is provided by a reference signal,Expressed as learning rateAnd input signalThe ratio of the squares of the modes of (c) is used to adjust the filter coefficients,Represented as a small constant that ensures filter stability.
An environmental parameter regulation and control module: receiving environmental parameter data of temperature and humidity from an intelligent sensor, adjusting the temperature and the humidity by using a PID control algorithm according to the environmental parameter data so as to adjust a test environment, obtaining constant environmental state data, and transmitting the constant environmental state data to a neural network prediction calibration module;
specifically, the step of adjusting environmental parameter data by the PID control algorithm is as follows:
The real-time acquisition of the environmental parameter data of the actual temperature and humidity of the semiconductor integrated circuit in the test process from the intelligent sensor is calibrated as And set target values of temperature and humidity in a suitable test environment for the semiconductor integrated circuit, calibrated as；
Calculating the deviation between the actual environment parameter data and the set target value, wherein the calculation formula of the deviation is as followsIn which, in the process,Represented as a result of a deviation between the actual environmental parameter data and the set target value,Expressed as the temperature in the target valueTemperature of the actual environmental parameterThe absolute value of the difference between them,Expressed as humidity in target valueHumidity with actual environmental parametersThe absolute value of the difference between them;
According to the deviation result Calculating PID control action comprising three parts of proportion, integral and derivative, respectively calibrated asThe calculation formula of the PID control quantity isIn which, in the process,Represented as a PID control quantity,Gain values expressed as proportional, integral, differential,Represented by the time step(s) indicated,Represented as at time stepsDeviation results at time；
According to PID control quantityThe output power control element of the heater or cooler, humidifier or dryer is adjusted to dynamically regulate temperature and humidity data, so that the regulated environmental parameters approach the target values.
Specifically, the logic for obtaining constant environmental state data is as follows:
by adjusting PID control quantity, the deviation result is obtained Approaching zero, when the environment state dataWhen reaching a steady-state point, continuously acquiring environmental parameter data through an intelligent sensor to acquire constant environmental state data;
The constant environmental state data is calculated as In which, in the process,Represented as constant environmental state data,Represented as atEnvironmental parameter data acquired by the intelligent sensor in real time,Represented as being infinitely close to the symbol.
The position detection error correction module to be detected: collecting dynamic mechanical position data point signals of which the test points are affected by vibration from an intelligent sensor, correcting errors of the test positions by using a gradient descent optimization algorithm to obtain corrected accurate test position signals, and transmitting the accurate test position signals to a neural network prediction calibration module;
Specifically, the error correction step of the gradient descent optimization algorithm on the test position is as follows:
Semiconductor integrated circuit currently using intelligent sensor acquisition test instrument The test point position signal at the moment is calibrated asAnd sets a position signal of an ideal test as；
According to the currentTest point position signal at timeAnd position signal for ideal testBetween, set an error function, calibrate asThe calculation formula of the error function isIn which, in the process,Expressed as presentTest point position signal at timeAnd position signal for ideal testThe value of the error between them,Represented as in a position signalThe difference in the coordinate points of the axes,Represented as in a position signalDifference of the axis coordinate points;
Selection of Position signal of initial test point at momentIn combination with the currentTest point position signal at timeCalculating the gradient of the error function, and calibrating asThe calculation formula of the gradient isAnd (2) andIn which, in the process,Expressed as presentTest point position signal at timeTo the point ofPosition signal of initial test point at momentIs used to determine the error value of (a),Represented as initialFrom moment to momentA time period of the moment;
Updating the position signal of the test point by using a gradient descent algorithm to reduce errors, and updating the calculation formula of the position signal of the test point to be In which, in the process,Is shown as the nextUpdating the calculated position signal at the moment in time,Expressed as a learning rate;
repeating the steps of calculating the gradient value and updating the position signal of the test point until the maximum iteration number is reached.
Specifically, the logic for obtaining the accurate test position signal is as follows:
continuously updating the test point position through a gradient descent algorithm To reduce the error function value；
According to the error valueThe iterative calculation is reduced to analyze, and whether the updating of the position of the test point is subjected to the maximum iteration times to reach a stable state is checked, so that the position adjustment is ensured to be stable and the precision requirement is met;
when the error falls within an acceptable range with iterative calculations, i.e When the test point is updated, the updated test point position is regarded as the accurate position, and an updated accurate test position signal is output and calibrated as。
Test parameter acquisition processing module: receiving test data after the EMI and RFI filtering compensation module carries out filtering treatment, carrying out constant environmental state data after the environmental parameter regulation and control module carries out the environmental parameter regulation and control, carrying out accurate test position signals after the position error correction by the position detection error correction module to be tested, carrying out preprocessing by using a characteristic extraction and data standardization algorithm, obtaining preprocessed test parameter signals, and transmitting the preprocessed test parameter signals to the neural network prediction calibration module;
specifically, the feature extraction process includes the following steps:
Collecting test data after filtering processing of EMI and RFI filtering compensation module Constant environmental state dataAnd accurate test position signalAnd identifying and analyzing characteristic data related to the test parameter signal;
describing test data using polynomial regression algorithm Constant environmental state dataAnd accurate test position signalRelation to test parameter signals and using test dataConstant environmental state dataAnd accurate test position signalFitting polynomial regression algorithm by influence coefficient of (2) to obtain characteristic data by combination or conversion, and calibrating asThe polynomial regression formula isAnd (2) andIn which, in the process,Represented as a signal of a test parameter,Expressed as test dataThe regression coefficient of the corresponding effect is used,Represented as constant environmental state dataThe regression coefficient of the corresponding effect is used,Expressed as an accurate test position signalThe regression coefficient of the corresponding effect is used,Represented as error terms;
the calculation formula of the characteristic data is as follows In which, in the process,Represented as an exponential function of the number of words,Represented as slave test dataConstant environmental state dataAnd accurate test position signalBased on regression coefficientsAnd extracted characteristic coefficients.
Specifically, the data normalization process steps are as follows:
By combining characteristic data Calculation of feature data using data normalization algorithmMean of (2)And standard deviationThen, carrying out standardized calculation to obtain standardized data, and calibrating asWherein the mean valueThe calculation formula of (2) isAnd (2) andIn which, in the process,Represented as characteristic dataIs used to determine the number of samples in the sample,Expressed as a total sampleMiddle (f)Corresponding characteristic data; Standard deviation ofThe calculation formula of (2) is；
Normalization of dataThe calculation formula of (2) isIn which, in the process,Represented as characteristic data is provided for the purpose of,Represented as characteristic dataIs used for the average value of (a),Represented as characteristic dataStandard deviation of (a) calculated standardized dataFor testing parameter signalsIs input after pretreatment of (a) and testing the parameter signal.
The neural network prediction calibration module: the method comprises the steps of receiving test parameter signals processed by a test parameter acquisition processing module, predicting and calibrating the test parameter signals by using a cyclic neural network (RNN) model to obtain fault hidden danger signals, optimizing test precision, and transmitting the fault hidden danger signals to an analysis diagnosis fault response module;
specifically, the prediction steps of the cyclic neural network RNN model are as follows:
collecting test parameter signals after pretreatment by a test parameter acquisition processing module And transmitted into an input layer in the recurrent neural network RNN;
the RNN layer signals test parameters in the input layer through recursive connection Processing, the calculation formula of the RNN layer is as followsIn which, in the process,Expressed as at timeIs used to determine the hidden state of the (c),Represented as an activation function,Expressed as the last timeIs used to determine the hidden state of the (c),Expressed as to the last timeIs hidden state of (a)Is used for the weight matrix of the (c),Represented as a signal to test parametersAt the time ofIs used to determine the weight matrix of the input values of (c),Represented as test parameter signalsAt the time ofIs used to determine the input value of (a),Bias terms denoted RNN layer;
in the output layer of the RNN network, the hidden state is converted into an output predicted value and calibrated as The calculation formula of the prediction output isIn which, in the process,Expressed as at timeIs hidden state of (a)Is used for the weight matrix of the (c),Represented as a bias term for the output layer,Expressed as at timeOutput predicted values of (a).
Specifically, the logic for acquiring the fault hidden danger signal is as follows:
by prediction of neural networks And actual timeIs a test parameter signal of (a)The difference between the two signals is used for determining the prediction error, obtaining the fault hidden danger signal and calibrating asThe calculation formula of the fault hidden danger signal isIn which, in the process,Prediction results expressed as RNNAnd actual timeIs a test parameter signal of (a)Is used for detecting the fault hidden trouble signal of the vehicle,Prediction results expressed as RNN，Represented as test parameter signalsAt the time ofIs used to determine the actual value of (c) in the (c),Represented as absolute value symbols;
Defining an error threshold for testing a semiconductor integrated circuit fault, calibrated as Comparing the fault hidden danger signal with the error threshold value whenJudging that the tested semiconductor integrated circuit has faults or anomalies;
according to two cases of semiconductor integrated circuit, including The types of faults are further classified, the occurrence time points of the faults and the occurrence types of the faults are marked, and analysis results are transmitted to an analysis and diagnosis fault response module for analysis and diagnosis.
An analysis and diagnosis fault response module: and receiving the predicted fault hidden danger signals in the neural network prediction calibration module, establishing a machine learning diagnosis model, analyzing and diagnosing according to the fault hidden danger signals, generating a diagnosis report and corresponding response measures, and feeding back to an operator for early warning response.
Specifically, the machine learning diagnostic model is built as follows:
Collecting fault hidden trouble signals As a training set for a machine learning diagnostic model;
training a machine learning diagnosis model by using a support vector machine (SNM) algorithm, and performing fault hidden danger signal Maximizing edges, further constructing a predictive objective function, and calibrating to beThe system is used for carrying out predictive diagnosis on the type of the fault;
the calculation formula of the objective function is In which, in the process,Represented as fault hidden trouble signalThrough the normal vector of the SVM hyperplane,The bias term denoted as SVM,Represented as a minimization function of the SVM,Represented as regularization parameters, for controlling the degree of penalty of the fault classification points,Respectively denoted as the firstCorresponding fault hidden trouble signal in each sampleFeature vector of (2), and fault hidden trouble signalIs a total sample number of (1);
by setting constraint conditions and calibrating as Then further determine the correct classification of the type of fault, then the constraint is expressed asAnd (2) andIs a positive integer, wherein,Represented as normal vectorIs a transposed calculation of (a),Represented as at the firstCorresponding fault hidden trouble signal in each sampleIs used for the feature vector of (a),Represented as consecutive sample points;
And deploying the trained machine learning diagnosis model into a test system to detect and diagnose faults in real time.
Specifically, the logic for generating the diagnostic report and response measures is as follows:
receiving fault hidden danger signals through a machine learning diagnosis model, identifying the fault type and severity, and obtaining a detection result;
the test system automatically generates detailed diagnosis reports, including time, reason, influence and solutions of corresponding faults;
according to the type and severity of the fault, setting repair suggestions, replacing components and adjusting response measures of operation parameters;
feeding back the diagnosis report and the response measures to a tested operator through a user interface UI, providing references and taking corresponding operation actions;
The performance of the machine learning diagnostic model is periodically monitored and iterative optimization is performed according to feedback, so that the diagnostic accuracy and response speed of the machine learning diagnostic model are improved.
The specific method and flow of the semiconductor integrated circuit testing system provided by the embodiment of the present invention are detailed in the embodiment of the above-mentioned semiconductor integrated circuit testing system, and are not repeated here.
The above formulas are all formulas with dimensions removed and numerical values calculated, the formulas are formulas with a large amount of data collected for software simulation to obtain the latest real situation, and preset parameters in the formulas are set by those skilled in the art according to the actual situation.
The above embodiments may be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented in software, the above-described embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the processes or functions described in accordance with embodiments of the present application are produced in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center by wired or wireless means (e.g., infrared, wireless, microwave, etc.). The computer readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that contains one or more sets of available media. The usable medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium may be a solid state disk.
It should be understood that, in various embodiments of the present application, the sequence numbers of the foregoing processes do not mean the order of execution, and the order of execution of the processes should be determined by the functions and internal logic thereof, and should not constitute any limitation on the implementation process of the embodiments of the present application.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
The foregoing is merely illustrative of the present application, and the present application is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (6)
1. A semiconductor integrated circuit testing system comprising an EMI and RFI filter compensation module: using an adaptive filtering technology of an active noise control ANC or an adaptive noise cancellation ANE to carry out filtering processing on test data containing EMI and RFI signals which are actually received so as to reduce interference, outputting the filtered test data and transmitting the test data to a neural network prediction calibration module;
An environmental parameter regulation and control module: receiving environmental parameter data of temperature and humidity from an intelligent sensor, adjusting the temperature and the humidity by using a PID control algorithm according to the environmental parameter data so as to adjust a test environment, obtaining constant environmental state data, and transmitting the constant environmental state data to a neural network prediction calibration module;
the step of adjusting the environmental parameter data by the PID control algorithm is as follows:
The real-time acquisition of the environmental parameter data of the actual temperature and humidity of the semiconductor integrated circuit in the test process from the intelligent sensor is calibrated as And set target values of temperature and humidity in a suitable test environment for the semiconductor integrated circuit, calibrated as；
Calculating the deviation between the actual environment parameter data and the set target value, wherein the calculation formula of the deviation is as followsIn which, in the process,Represented as a result of a deviation between the actual environmental parameter data and the set target value,Expressed as the temperature in the target valueTemperature of the actual environmental parameterThe absolute value of the difference between them,Expressed as humidity in target valueHumidity with actual environmental parametersThe absolute value of the difference between them;
According to the deviation result Calculating PID control action comprising three parts of proportion, integral and derivative, respectively calibrated asThe calculation formula of the PID control quantity isIn which, in the process,Represented as a PID control quantity,Gain values expressed as proportional, integral, differential,Represented by the time step(s) indicated,Represented as at time stepsDeviation results at time；
According to PID control quantityThe output power control element of the heater or the cooler, the humidifier or the dryer is adjusted, and the data of the temperature and the humidity are dynamically regulated, so that the regulated environmental parameters approach to the target values;
The constant environmental state data acquisition logic is as follows:
by adjusting PID control quantity, the deviation result is obtained Approaching zero, when the environment state dataWhen reaching a steady-state point, continuously acquiring environmental parameter data through an intelligent sensor to acquire constant environmental state data;
The constant environmental state data is calculated as In which, in the process,Represented as constant environmental state data,Represented as atEnvironmental parameter data acquired by the intelligent sensor in real time,Represented as infinitely close to the symbol;
The position detection error correction module to be detected: collecting dynamic mechanical position data point signals of which the test points are affected by vibration from an intelligent sensor, correcting errors of the test positions by using a gradient descent optimization algorithm to obtain corrected accurate test position signals, and transmitting the accurate test position signals to a neural network prediction calibration module;
The error correction step of the gradient descent optimization algorithm on the test position is as follows:
Semiconductor integrated circuit currently using intelligent sensor acquisition test instrument The test point position signal at the moment is calibrated asAnd sets a position signal of an ideal test as；
According to the currentTest point position signal at timeAnd position signal for ideal testBetween, set an error function, calibrate asThe calculation formula of the error function isIn which, in the process,Expressed as presentTest point position signal at timeAnd position signal for ideal testThe value of the error between them,Represented as in a position signalThe difference in the coordinate points of the axes,Represented as in a position signalDifference of the axis coordinate points;
Selection of Position signal of initial test point at momentIn combination with the currentTest point position signal at timeCalculating the gradient of the error function, and calibrating asThe calculation formula of the gradient isAnd (2) andIn which, in the process,Expressed as presentTest point position signal at timeTo the point ofPosition signal of initial test point at momentIs used to determine the error value of (a),Represented as initialFrom moment to momentA time period of the moment;
Updating the position signal of the test point by using a gradient descent algorithm to reduce errors, and updating the calculation formula of the position signal of the test point to be In which, in the process,Is shown as the nextUpdating the calculated position signal at the moment in time,Expressed as a learning rate;
repeating the steps of calculating the gradient value and updating the position signal of the test point until the maximum iteration number is reached;
the acquisition logic for the accurate test position signal is as follows:
continuously updating the test point position through a gradient descent algorithm To reduce the error function value；
According to the error valueThe iteration calculation is reduced to analyze, and whether the updating of the position of the test point is in a stable state or not is checked;
when the error falls within an acceptable range with iterative calculations, i.e When the test point is updated, the updated test point position is regarded as the accurate position, and an updated accurate test position signal is output and calibrated as；
Test parameter acquisition processing module: receiving test data after the EMI and RFI filtering compensation module carries out filtering treatment, carrying out constant environmental state data after the environmental parameter regulation and control module carries out the environmental parameter regulation and control, carrying out accurate test position signals after the position error correction by the position detection error correction module to be tested, carrying out preprocessing by using a characteristic extraction and data standardization algorithm, obtaining preprocessed test parameter signals, and transmitting the preprocessed test parameter signals to the neural network prediction calibration module;
The neural network prediction calibration module: receiving the test parameter signals processed by the test parameter acquisition processing module, predicting and calibrating the test parameter signals by using a cyclic neural network (RNN) model to obtain fault hidden danger signals, and transmitting the fault hidden danger signals to the analysis and diagnosis fault response module;
An analysis and diagnosis fault response module: and receiving the predicted fault hidden danger signals in the neural network prediction calibration module, establishing a machine learning diagnosis model, analyzing and diagnosing according to the fault hidden danger signals, generating a diagnosis report and corresponding response measures, and feeding back to an operator for early warning response.
2. The semiconductor integrated circuit test system according to claim 1, wherein the feature extraction process steps are as follows:
Collecting test data after filtering processing of EMI and RFI filtering compensation module Constant environmental state dataAnd accurate test position signalAnd identifying and analyzing characteristic data related to the test parameter signal;
describing test data using polynomial regression algorithm Constant environmental state dataAnd accurate test position signalRelation to test parameter signals and using test dataConstant environmental state dataAnd accurate test position signalFitting polynomial regression algorithm by influence coefficient of (2) to obtain characteristic data by combination or conversion, and calibrating asThe polynomial regression formula isAnd (2) andIn which, in the process,Represented as a signal of a test parameter,Expressed as test dataThe regression coefficient of the corresponding effect is used,Represented as constant environmental state dataThe regression coefficient of the corresponding effect is used,Expressed as an accurate test position signalThe regression coefficient of the corresponding effect is used,Represented as error terms;
the calculation formula of the characteristic data is as follows In which, in the process,Represented as an exponential function of the number of words,Represented as slave test dataConstant environmental state dataAnd accurate test position signalBased on regression coefficientsAnd extracted characteristic coefficients.
3. A semiconductor integrated circuit testing system according to claim 2, wherein said data normalization process comprises the steps of:
By combining characteristic data Calculation of feature data using data normalization algorithmMean of (2)And standard deviationThen, carrying out standardized calculation to obtain standardized data, and calibrating asWherein the mean valueThe calculation formula of (2) isAnd (2) andIn which, in the process,Represented as characteristic dataIs used to determine the number of samples in the sample,Expressed as a total sampleMiddle (f)Corresponding characteristic data; Standard deviation ofThe calculation formula of (2) is；
Normalization of dataThe calculation formula of (2) isIn which, in the process,Represented as characteristic data is provided for the purpose of,Represented as characteristic dataIs used for the average value of (a),Represented as characteristic dataStandard deviation of (a) calculated standardized dataFor testing parameter signalsIs input after pretreatment of (a) and testing the parameter signal.
4. A semiconductor integrated circuit testing system according to claim 3, wherein the cyclic neural network RNN model prediction step is as follows:
collecting test parameter signals after pretreatment by a test parameter acquisition processing module And transmitted into an input layer in the recurrent neural network RNN;
the RNN layer signals test parameters in the input layer through recursive connection Processing, the calculation formula of the RNN layer is as followsIn which, in the process,Expressed as at timeIs used to determine the hidden state of the (c),Represented as an activation function,Expressed as the last timeIs used to determine the hidden state of the (c),Expressed as to the last timeIs hidden state of (a)Is used for the weight matrix of the (c),Represented as a signal to test parametersAt the time ofIs used to determine the weight matrix of the input values of (c),Represented as test parameter signalsAt the time ofIs used to determine the input value of (a),Bias terms denoted RNN layer;
in the output layer of the RNN network, the hidden state is converted into an output predicted value and calibrated as The calculation formula of the prediction output isIn which, in the process,Expressed as at timeIs hidden state of (a)Is used for the weight matrix of the (c),Represented as a bias term for the output layer,Expressed as at timeOutput predicted values of (a).
5. The semiconductor integrated circuit test system of claim 4, wherein the logic for obtaining the fault signal is as follows:
by prediction of neural networks And actual timeIs a test parameter signal of (a)The difference between the two signals is used for determining the prediction error, obtaining the fault hidden danger signal and calibrating asThe calculation formula of the fault hidden danger signal isIn which, in the process,Prediction results expressed as RNNAnd actual timeIs a test parameter signal of (a)Is used for detecting the fault hidden trouble signal of the vehicle,Prediction results expressed as RNN，Represented as test parameter signalsAt the time ofIs used to determine the actual value of (c) in the (c),Represented as absolute value symbols;
Defining an error threshold for testing a semiconductor integrated circuit fault, calibrated as Comparing the fault hidden danger signal with the error threshold value whenJudging that the tested semiconductor integrated circuit has faults or anomalies;
according to two cases of semiconductor integrated circuit, including The types of faults are further classified, the occurrence time points of the faults and the occurrence types of the faults are marked, and analysis results are transmitted to an analysis and diagnosis fault response module for analysis and diagnosis.
6. The semiconductor integrated circuit test system of claim 5, wherein the machine learning diagnostic model is built by:
Collecting fault hidden trouble signals As a training set for a machine learning diagnostic model;
training a machine learning diagnosis model by using a support vector machine (SNM) algorithm, and performing fault hidden danger signal Maximizing edges, further constructing a predictive objective function, and calibrating to beThe system is used for carrying out predictive diagnosis on the type of the fault;
the calculation formula of the objective function is In which, in the process,Represented as fault hidden trouble signalThrough the normal vector of the SVM hyperplane,The bias term denoted as SVM,Represented as a minimization function of the SVM,Represented as a regularization parameter,Respectively denoted as the firstCorresponding fault hidden trouble signal in each sampleFeature vector of (2), and fault hidden trouble signalIs a total sample number of (1);
by setting constraint conditions and calibrating as Then further determine the correct classification of the type of fault, then the constraint is expressed asAnd (2) andIs a positive integer, wherein,Expressed as normal vectorIs a transposed calculation of (a),Represented as at the firstCorresponding fault hidden trouble signal in each sampleIs used for the feature vector of (a),Represented as consecutive sample points;
And deploying the trained machine learning diagnosis model into a test system to detect and diagnose faults in real time.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410760811.5A CN118330447B (en) | 2024-06-13 | 2024-06-13 | Semiconductor integrated circuit test system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410760811.5A CN118330447B (en) | 2024-06-13 | 2024-06-13 | Semiconductor integrated circuit test system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118330447A CN118330447A (en) | 2024-07-12 |
| CN118330447B true CN118330447B (en) | 2024-09-03 |
Family
ID=91782937
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410760811.5A Active CN118330447B (en) | 2024-06-13 | 2024-06-13 | Semiconductor integrated circuit test system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118330447B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112630620A (en) * | 2020-12-14 | 2021-04-09 | 清华大学 | Testing device, testing system and testing method for semiconductor sample |
| CN117970211A (en) * | 2023-11-23 | 2024-05-03 | 广西电网有限责任公司电力科学研究院 | Self-adaptive deep learning transformer test instrument calibration system |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1999045408A1 (en) * | 1998-03-06 | 1999-09-10 | Btg International Limited | Apparatus for and method of nuclear quadrupole resonance testing a sample in the presence of interference |
| DE102021201618A1 (en) * | 2021-02-19 | 2022-08-25 | Hyundai Motor Company | Method and system for active noise control in vehicles |
| CN117874523A (en) * | 2024-01-25 | 2024-04-12 | 苏州旭澜达通讯技术有限公司 | Filter auxiliary debugging method, system and medium based on PID controller |
-
2024
- 2024-06-13 CN CN202410760811.5A patent/CN118330447B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112630620A (en) * | 2020-12-14 | 2021-04-09 | 清华大学 | Testing device, testing system and testing method for semiconductor sample |
| CN117970211A (en) * | 2023-11-23 | 2024-05-03 | 广西电网有限责任公司电力科学研究院 | Self-adaptive deep learning transformer test instrument calibration system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118330447A (en) | 2024-07-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111474510B (en) | Error evaluation method and system for voltage transformer with non-stable output | |
| US20050182581A1 (en) | Testing of wire systems and end devices installed in industrial processes | |
| CN111665066B (en) | Equipment fault self-adaptive upper and lower early warning boundary generation method based on convolutional neural network | |
| CN116735804A (en) | Intelligent sensor precision monitoring system based on Internet of things | |
| Chen et al. | Status self-validation of sensor arrays using gray forecasting model and bootstrap method | |
| JP2011204264A (en) | Method, system and storage medium for performing online valve diagnosis | |
| CN108761281B (en) | Method and system for monitoring state and positioning partial discharge of gas insulated transmission line | |
| CN117798744B (en) | Method for monitoring running state of numerical control machine tool | |
| CN116678489B (en) | Quality control method, system, equipment and medium for force sensor | |
| Li et al. | Signal frequency domain analysis and sensor fault diagnosis based on artificial intelligence | |
| CN114646680A (en) | Automatic test system for gas sensor | |
| CN102342582B (en) | Verification method, verification system and calculation processor for detection precision of scan detection head | |
| CN118330447B (en) | Semiconductor integrated circuit test system | |
| CN117369349B (en) | Management system of remote monitoring intelligent robot | |
| CN117687461B (en) | Environment regulation and control system of animal laboratory | |
| CN117969935A (en) | Current measurement method, device and test equipment under ATE leakage current scene | |
| RU2714039C1 (en) | Smart sensor development system | |
| CN118275034B (en) | Intelligent anti-interference pressure transmitter | |
| CN118408734B (en) | High-precision breather valve detector base number calibration method | |
| CN116929990A (en) | Quality inspection system and method for dehydration treatment of aluminum powder | |
| CN118482817B (en) | Digital textile equipment management method and system based on data analysis | |
| CN118034136A (en) | Digital twin method and system applied to industrial research and development design | |
| CN118033301B (en) | Intelligent detection system and method for detecting conductive back adhesive performance | |
| CN118091522B (en) | Testing and checking system and method for transformer substation | |
| CN118249810B (en) | Method and system for testing multichannel AD/DA (analog to digital) chip |
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 | ||
| GR01 | Patent grant |