Electronic component performance data monitoring method and system

Abstract

The invention discloses an electronic component performance data monitoring method and system, and the method comprises the steps: inputting an LSTM neural network model for training based on the work historical data of an electronic component, and constructing an electronic component performance prediction model; inputting the current electronic component surface temperature field distribution data, the partial discharge characteristic quantity and the leakage current dynamic parameter into the electronic component performance prediction model, and obtaining a plurality of prediction change trend functions of the electronic component performance in a future time period; comparing the data with electronic component performance actual data acquired in real time, calculating an error value, and acquiring prediction data with the minimum error; dynamic error analysis is carried out, and model correction parameters are generated; and performing online parameter updating on the electronic component performance prediction model to form a closed-loop optimization system. The method has the advantages that the performance change of the electronic component is accurately predicted by comparing the prediction model and the multi-dimensional real-time data, and the long-term stability and reliability of the electronic component are ensured through a closed-loop optimization mechanism.

Description

Technical Field

[0001] The present invention relates to data monitoring technology, and particularly to a method and system for monitoring the performance data of electronic components. Background Art

[0002] With the rapid development of electronic technology, electronic components are increasingly widely used in various electronic products, especially in the fields of communication, automotive, aerospace, medical, etc. Electronic components play a crucial role in these fields. For example, microprocessors in communication systems, sensors in automobiles, power management systems in spacecraft, and precision sensors in medical devices all need to work continuously and stably under high-intensity and complex environments. Therefore, ensuring the efficient and stable operation of these components is crucial for the safety and reliability of the entire system.

[0003] Most of the current methods for monitoring the performance data of electronic components rely on traditional static models or single-parameter monitoring means, lacking comprehensive analysis of multi-dimensional data. These methods usually cannot perform performance prediction and dynamic adjustment in real time, and the accuracy and adaptability of the prediction results are poor, being easily affected by factors such as environmental changes, working conditions, and aging. In addition, many existing methods lack self-adaptive capabilities and cannot be optimized and corrected according to real-time data, resulting in difficult effective control of prediction errors. Traditional methods usually rely on fixed models and cannot fully consider multiple influencing factors such as temperature, partial discharge, leakage current, etc., resulting in limited monitoring accuracy and insufficient ability to handle complex situations. More importantly, most of these methods lack dynamic error analysis and online update mechanisms and cannot be adjusted according to the errors and deviations in actual operation, resulting in poor long-term stability of the model and inability to continuously optimize and improve prediction accuracy. Summary of the Invention

[0004] In order to improve the existing methods for monitoring the performance data of electronic components, a method and system for monitoring the performance data of electronic components are provided. This method combines an LSTM neural network and multi-dimensional real-time data to accurately predict the performance changes of electronic components, and dynamically optimizes the prediction model to improve accuracy. Through an online correction and closed-loop optimization mechanism, the long-term stability and reliability of electronic components under different working conditions are ensured.

[0005] To achieve the above object, the technical solution adopted by the present invention is as follows: A method for monitoring the performance data of electronic components, comprising: Based on the working history data of the electronic components, input it into an LSTM neural network model for training to construct a performance prediction model of the electronic components, and obtain the change function of the performance of the electronic components with working time; Obtain the surface temperature field distribution data, partial discharge characteristic quantities, and leakage current dynamic parameters of the current electronic component, input them into the electronic component performance prediction model, and based on the changes of various factors, obtain multiple prediction change trend functions of the electronic component performance in the future time period; Based on the obtained multiple prediction change trend function data, compare it with the actual data of the electronic component performance obtained by real-time acquisition, calculate the error value, and obtain the prediction data with the smallest error among them; Based on the prediction data with the smallest error, conduct dynamic error analysis to generate model correction parameters; Based on the model correction parameters, perform online parameter update on the electronic component performance prediction model to form a closed-loop optimization system.

[0006] Preferably, the input of the working history data of the electronic component into the LSTM neural network model for training to construct an electronic component performance prediction model and obtain the change function of the electronic component performance with working time specifically includes: Transmit the working history data of the electronic component into the LSTM neural network model for training to construct an electronic component performance prediction model; Based on the double-branch structure of the output layer in the electronic component performance prediction model, the main branch predicts the performance parameters at the current moment, and the auxiliary branch outputs the performance degradation rate. Through integral operation, the functional relationship between the performance parameters and the working time is constructed.

[0007] Preferably, the obtaining of the surface temperature field distribution data, partial discharge characteristic quantities, and leakage current dynamic parameters of the current electronic component, inputting them into the electronic component performance prediction model, and based on the changes of various factors, obtaining multiple prediction change trend functions of the electronic component performance in the future time period specifically includes: Scan the surface of the electronic component through a non-contact infrared thermal imager array to obtain a two-dimensional temperature distribution matrix and extract characteristic parameters; Collect pulse signals through a high-frequency current sensor to obtain partial discharge characteristic quantities; Obtain current characteristic parameters through dual-channel synchronous acquisition; Input the preprocessed acquisition data into the electronic component performance prediction model, and based on the possible change trends of the electronic component parameter data, generate multi-scenario electronic component performance prediction change trend functions.

[0008] Preferably, the inputting of the preprocessed acquisition data into the electronic component performance prediction model and generating multi-scenario electronic component performance prediction change trend functions based on the possible change trends of the electronic component parameter data specifically includes: Based on the characteristic parameters affecting the performance of the electronic component, obtain the change ranges of each perturbation dimension; Through orthogonal experimental design, various parameter change scenarios are combined to generate multiple scenarios of the performance change of electronic components; Based on the characteristic parameter data under each scenario, input it into the performance prediction model of electronic components to generate multiple performance prediction change trend functions of electronic components, and obtain the characteristic data including: time series of performance parameters, performance degradation acceleration, and prediction of the arrival time of key thresholds.

[0009] Preferably, based on the obtained multiple prediction change trend function data, compare it with the actual data of the performance of electronic components acquired in real time, calculate the error value, and obtain the prediction data with the smallest error, which specifically includes: Based on the error contribution degree of each characteristic parameter, perform dynamic error weight allocation; Compare the real-time acquired data with the data at the corresponding time nodes in the prediction change trend function, and obtain the error magnitude between the data of each prediction change trend function and the actually acquired data through multi-dimensional error joint calculation; Based on a set of prediction change trend functions with the smallest error value in the error data, select it as the best performance prediction model of electronic components; Based on the actual data of the performance of electronic components acquired at each moment, screen and obtain the prediction result of the best performance prediction model of electronic components.

[0010] Preferably, based on the prediction data with the smallest error, perform dynamic error analysis to generate model correction parameters, which specifically includes: Based on the prediction data with the smallest error, obtain the fluctuation of the error through the error distribution diagram; Through the relationship between the error and the working conditions, obtain the cause of the error under specific conditions; By comparing the error distribution diagram and the output of the model prediction, obtain the deviation in data, model structure, or parameter settings; Based on the results of dynamic error analysis, adjust the model parameters to generate model correction parameters, including: adjustment of the learning rate, addition of regularization terms, and adjustment of the training data volume.

[0011] Preferably, based on the model correction parameters, perform online parameter update on the performance prediction model of electronic components to form a closed-loop optimization system, which specifically includes: Based on the obtained model correction parameters, perform online training on the model through the incremental learning method, and continuously optimize the model using the real-time acquired performance data of electronic components and error correction information; Through an automated parameter adjustment mechanism, when the error exceeds the set threshold, automatically trigger the update of the model parameters; Compare the prediction result of the model after each update with the real-time data again to form a continuous feedback process.

[0012] Furthermore, an electronic component performance data monitoring system is proposed, including: Model training module: The model training module is mainly used to input the working historical data of electronic components into the LSTM neural network for training, construct a performance prediction model, and obtain the function of performance changing with time; Performance prediction model module: The performance prediction model module is mainly used to construct a performance prediction model of electronic components based on the output structure of the LSTM network, and predict various change trend functions in the future time period; Multi-scenario prediction analysis module: The multi-scenario prediction analysis module generates multiple prediction scenarios based on different working conditions and disturbance factors, combined with the orthogonal experimental design; Error calculation module: The error calculation module is mainly used to compare the real-time collected performance data with the results of the prediction model, calculate the error, and perform dynamic weight allocation based on the error contribution degree to obtain the prediction model with the smallest error value; Online update and closed-loop optimization module: The online update and closed-loop optimization module is mainly used to perform incremental learning based on the model correction parameters, update the prediction model online, automatically adjust the parameters, and continuously optimize the performance prediction results; Processor: The processor is mainly used for the calculation process of each formula and the construction calculation process of each model.

[0013] Compared with the prior art, the advantages of the present invention are: Through the performance prediction model trained based on historical data, the law of the performance of electronic components changing with time can be accurately captured, and potential failure risks can be discovered in time. In addition, the method provides a more comprehensive reference for performance prediction by integrating multi-dimensional data such as surface temperature, partial discharge, and leakage current, ensuring the accuracy and diversity of the prediction. Through the error dynamic analysis and online correction mechanism, the system can continuously optimize the prediction model during actual operation, ensuring its high efficiency and stability under different working conditions. This method not only has the adaptive ability to adjust the prediction results according to real-time data, but also can realize a closed-loop optimization system to ensure the long-term reliability and performance stability of electronic components. Brief Description of the Drawings

[0014] Figure 1 It is a schematic diagram of the method proposed by the present invention; Figure 2 It is a schematic diagram of the construction of the electronic component performance prediction model proposed by the present invention; Figure 3 It is a schematic diagram of the calculation of the electronic component performance prediction model proposed by the present invention; Figure 4 It is a schematic diagram of obtaining various prediction change trend functions proposed by the present invention; Figure 5Schematic diagram for obtaining minimum error prediction data proposed by the present invention; Figure 6 Schematic diagram for generating model correction parameters proposed by the present invention; Figure 7 Schematic diagram for online parameter update proposed by the present invention; Figure 8 Architecture diagram of the electronic device in this solution; Figure 9 Schematic diagram of the structure of the computer-readable storage medium in this solution. Specific implementation manners

[0015] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments in the following description are only examples, and those skilled in the art can think of other obvious variations.

[0016] The electronic component performance data monitoring system includes: Model training module: The model training module is mainly used to input the working historical data of the electronic component into the LSTM neural network for training, construct a performance prediction model and obtain a function of performance changing with time; Performance prediction model module: The performance prediction model module is mainly used to construct a performance prediction model of the electronic component based on the output structure of the LSTM network, and predict various change trend functions in the future time period; Multi-scenario prediction analysis module: The multi-scenario prediction analysis module generates multiple prediction scenarios based on different working conditions and disturbance factors, in combination with the orthogonal experimental design; Error calculation module: The error calculation module is mainly used to compare the real-time collected performance data with the prediction model results, calculate the error and perform dynamic weight allocation based on the error contribution degree, and obtain the prediction model with the smallest error value; Online update and closed-loop optimization module: The online update and closed-loop optimization module is mainly used to perform incremental learning based on the model correction parameters, update the prediction model online, automatically adjust the parameters and continuously optimize the performance prediction results; Processor: The processor is mainly used for the calculation process of each formula and the construction calculation process of each model.

[0017] Refer to Figure 1 As shown, the electronic component performance data monitoring method includes: Step 1: Based on the working historical data of the electronic component, input it into the LSTM neural network model for training, construct an electronic component performance prediction model, and obtain a function of the performance of the electronic component changing with the working time; Step 2: Obtain the surface temperature field distribution data, partial discharge characteristic quantities, and leakage current dynamic parameters of the current electronic component, input them into the electronic component performance prediction model, and based on the changes of various factors, obtain multiple prediction change trend functions of the electronic component performance in the future time period; Step 3: Based on the obtained multiple prediction change trend function data, compare it with the actual data of the electronic component performance obtained by real-time acquisition, calculate the error value, and obtain the prediction data with the smallest error; Step 4: Based on the prediction data with the smallest error, conduct dynamic error analysis to generate model correction parameters; Step 5: Based on the model correction parameters, perform online parameter update on the electronic component performance prediction model to form a closed-loop optimization system.

[0018] Refer to Figure 2 As shown, based on the working historical data of the electronic component, input it into the LSTM neural network model for training, construct the electronic component performance prediction model, and the specific process of obtaining the function of the electronic component performance changing with the working time includes: Transmit the working historical data of the electronic component into the LSTM neural network model for training to construct the electronic component performance prediction model; Based on the double-branch structure of the output layer in the electronic component performance prediction model, the main branch predicts the performance parameters at the current moment, and the auxiliary branch outputs the performance degradation rate. Through integral operation, the functional relationship between the performance parameters and the working time is constructed.

[0019] Specifically, at the output layer of the LSTM network, design a double-branch structure. The main branch is used to predict the performance parameters at the current moment, and the auxiliary branch is used to output the performance degradation rate; The output of the LSTM in the main branch is , and the output of the main branch is the performance parameters at the current moment , and it is predicted through a fully connected layer. The formula is: ; Among them, is the predicted value of the performance parameters at the current moment, is the output of the LSTM, is the trainable weight matrix, which is used to linearly map the high-dimensional hidden state to the target parameter dimension and determine the feature weight distribution, is the trainable bias term; The output of the auxiliary branch is the performance degradation rate and it is predicted through a fully connected layer. The formula is: ; Among them, is the output of the auxiliary branch, which is the performance degradation rate, is the weight matrix of the fully connected layer, is the bias term of the fully connected layer; A functional relationship is established between the performance parameters and the working time through integral operation. Specifically, the performance parameter changes over time through the performance degradation rate and is updated according to the formula: ; Integrating the degradation rate with respect to time, the relationship between the performance parameter and time can be obtained, and the formula is: ; wherein, is the value of the performance parameter at time t, is the integration variable, representing any instant within the time period from to t; Refer to Figure 3 as shown, obtain the surface temperature field distribution data, partial discharge characteristic quantities and leakage current dynamic parameters of the current electronic component, input them into the electronic component performance prediction model, and based on the changes of various factors, obtain multiple prediction change trend functions of the electronic component performance in the future time period, specifically including: Scan the surface of the electronic component through a non-contact infrared thermal imager array to obtain a two-dimensional temperature distribution matrix and extract characteristic parameters; Collect pulse signals through a high-frequency current sensor to obtain partial discharge characteristic quantities; Obtain current characteristic parameters through dual-channel synchronous acquisition; Input the preprocessed acquisition data into the electronic component performance prediction model, and generate multi-scenario electronic component performance prediction change trend functions based on the possible change trends of the electronic component parameter data.

[0020] Specifically, scan the surface of the electronic component through a non-contact infrared thermal imager array to obtain a two-dimensional temperature distribution matrix, and extract characteristic data including the average temperature, temperature variance, and maximum and minimum temperatures from the temperature matrix; collect pulse signals through a high-frequency current sensor to obtain partial discharge characteristic quantities, and extract the characteristic parameters of partial discharge by analyzing the amplitude, frequency, duration, etc. of the current signal; obtain current characteristic parameters through dual-channel synchronous acquisition of current signals, including the maximum current value and the average value of the waveform; Fuse the characteristics extracted from different sensors and use them as input data to input into the electronic component performance prediction model.

[0021] Refer to Figure 4As shown, the preprocessed collected data is input into the electronic component performance prediction model, and based on the possible change trends of the electronic component parameter data, a multi-scenario electronic component performance prediction change trend function is generated, specifically including: Based on the characteristic parameters affecting the performance of the electronic component, obtain the change range of each perturbation dimension; Through orthogonal experimental design, combine the parameter change scenarios to generate multiple scenarios of the electronic component performance change; Based on the characteristic parameter data under each scenario, input it into the electronic component performance prediction model to generate multiple electronic component performance prediction change trend functions, and the obtained characteristic data includes: performance parameter time series, performance degradation acceleration, and key threshold arrival time prediction.

[0022] Specifically, the performance of the electronic component is affected by multiple factors, such as temperature, current, voltage, operating frequency, etc. Through historical data and according to the changes in the component operating state, determine different perturbation dimensions, and obtain the minimum value, maximum value, and change range of each characteristic; According to the perturbation dimension and its change range, select a suitable orthogonal table for experimental design to ensure that each experimental combination can effectively evaluate the influence of different perturbation dimensions, and obtain multiple different experimental scenarios (combinations), that is, the specific values of the characteristic parameters under each experimental condition; Based on the characteristic parameter data under different scenarios, input it into the electronic component performance prediction model to generate different performance prediction change trend functions; By predicting the time series data under each scenario, the trend of the performance parameter changing with time can be obtained. The formula is: ; where is the performance parameter at the initial moment, is the degradation rate under scenario ; The performance degradation acceleration refers to the change of the performance degradation rate with time. For each scenario, the LSTM model can output the degradation rate , and further derive its acceleration. The formula is: ; where is the performance degradation acceleration; By predicting the change of the performance parameter with time, the arrival time of the key threshold (such as the performance drops to a certain critical value) can be predicted.

[0023] Refer to Figure 5As shown, based on the obtained data of various prediction change trend functions, compare them with the actual data of the electronic component performance obtained through real-time acquisition, calculate the error value, and obtain the prediction data with the smallest error, which specifically includes: Perform dynamic error weight allocation based on the error contribution degrees of each characteristic parameter; Compare the real-time acquisition data with the data at the corresponding time nodes in the prediction change trend function, and through multi-dimensional error joint calculation, obtain the error magnitudes between the data of each prediction change trend function and the actual acquisition data; Based on a set of prediction change trend functions with the smallest error value in the error data, select it as the optimal electronic component performance prediction model; Based on the actual data of the electronic component performance acquired at each moment, screen and obtain the prediction result of an optimal electronic component performance prediction model.

[0024] Specifically, in the calculation of the error contribution degree of the characteristic parameter, calculate its contribution degree according to the square of each characteristic error , and the formula is: ; Among them, is the error contribution degree, is the error contribution degree of the i-th characteristic parameter at time t, is the sum of the squares of the errors of all n characteristic parameters at time t; According to the error contribution degree , dynamically adjust the weight of each characteristic; After real-time data acquisition, it is necessary to compare with the change trend function predicted by the model, calculate the error between the real-time data and the predicted data, and through multi-dimensional error joint calculation, obtain the error magnitude at each moment, including absolute error, relative error, and trend error; By comparing the errors of different prediction models, select the prediction model with the smallest error as the optimal prediction model. Once the optimal model is selected, based on the actual data at each moment, use the prediction result of this model for subsequent performance prediction; In the actual application process, as the real-time data is continuously updated, the error contribution degree and dynamic weight can be recalculated according to the new error data, and the parameters of the model can be adjusted. In this way, the model will be gradually optimized and can more accurately predict the performance of the electronic component.

[0025] Refer to Figure 6 As shown, based on the prediction data with the smallest error, perform dynamic error analysis and generate model correction parameters, which specifically includes: Based on the prediction data with the smallest error, obtain the fluctuation situation of the error through the error distribution diagram; Obtain the cause of error under specific conditions through the relationship between error and working conditions; Obtain the deviation in data, model structure, or parameter settings by comparing the error distribution map and the output predicted by the model; Based on the results of dynamic error analysis, adjust the model parameters to generate model correction parameters, including: adjustment of the learning rate, addition of regularization terms, and adjustment of the amount of training data.

[0026] Specifically, obtain the distribution of errors by calculating the statistics of the error sequence. By analyzing the relationship between errors and working conditions, the cause of errors under specific conditions can be deeply understood; By comparing the error distribution map and the model prediction output, analyze the deviation in data, model structure, or parameter settings, including: Error offset: If the errors mostly concentrate on one side of positive or negative values, it indicates that there may be a deviation in the prediction model. For example, the model may be more optimistic in prediction under low-temperature conditions and more pessimistic under high-temperature conditions; Error fluctuation: If the errors fluctuate greatly over time, it may mean that the prediction of the model is unstable at certain time points; Degree of matching between model output and data: If the distribution of the model output is inconsistent with the distribution of the actual data, there may be a problem that the model structure or parameters do not adapt to the current data; Based on the error analysis results, dynamically adjust the model parameters. Adjust the amplitude of parameter adjustment each time the model is updated by adjusting the learning rate, prevent model overfitting by adding regularization terms, and improve the generalization ability of the model by increasing the amount of training data; According to the results of the above steps, a set of model correction parameters can finally be generated to improve the prediction ability of the model.

[0027] Refer to Figure 7 As shown, perform online parameter update on the electronic component performance prediction model based on the model correction parameters to form a closed-loop optimization system, which specifically includes: Based on the obtained model correction parameters, perform online training on the model through the incremental learning method, and continuously optimize the model by using the real-time obtained electronic component performance data and error correction information; Through an automated parameter adjustment mechanism, when the error exceeds the set threshold, automatically trigger the update of the model parameters; Compare the predicted result of the model after each update with the real-time data again to form a continuous feedback process.

[0028] Specifically, use gradient descent for online update of model parameters, and the formula is: ; Where, is the current parameter of the model, is the gradient of the loss function with respect to the parameters, is the current learning rate; After updating the model, based on generate new prediction results, compare the error between the updated predicted value and the actual value, and add the error history for the next step of dynamic weight or model correction judgment; Automatically adjust the learning rate parameter based on the error trend to avoid continuous increase or decrease of the error.

[0029] Furthermore, the method according to the embodiment of the present application can also be implemented by means of Figure 8 the architecture of the electronic device shown. As Figure 8 shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to the network, an input / output component 506, a hard disk 507, etc. The storage device in the electronic device 500, such as the ROM 503 or the hard disk 507, may store the method and system for monitoring the performance data of electronic components provided by the present application. The electronic device 500 may also include a terminal interface 508. Of course, Figure 8 the architecture shown is only exemplary. When implementing different devices, one or more components in the Figure 8 shown electronic device may be omitted according to actual needs.

[0030] Figure 9 is a schematic diagram of the structure of a computer-readable storage medium provided by an embodiment of the present application. As Figure 9 shown, it is a computer-readable storage medium 600 according to an embodiment of the present application. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are run by a processor, the method and system for monitoring the performance data of electronic components according to the embodiment of the present application described with reference to the above drawings can be executed. The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory, etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0031] It should be noted that: the above sequence of the embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. And the above description of specific embodiments of this specification has been made. Additionally, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0032] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other, and the key point of each embodiment is to illustrate the differences from other embodiments.

[0033] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for monitoring the performance data of electronic components, characterized in that, Including: Based on the working historical data of the electronic component, input it into the LSTM neural network model for training, construct an electronic component performance prediction model, and obtain the variation function of the performance of the electronic component with working time; Obtain the current surface temperature field distribution data, partial discharge characteristic quantity and leakage current dynamic parameters of the electronic component, input them into the electronic component performance prediction model, and based on the changes of various factors, obtain multiple predicted variation trend functions of the performance of the electronic component in the future time period; Based on the obtained multiple predicted variation trend function data, compare it with the actual data of the performance of the electronic component collected in real time, calculate the error value, and obtain the predicted data with the smallest error; Based on the predicted data with the smallest error, conduct dynamic error analysis and generate model correction parameters; Based on the model correction parameters, perform online parameter update on the electronic component performance prediction model to form a closed-loop optimization system.

2. The method for monitoring the performance data of an electronic component according to claim 1, characterized in that, The specific steps of "Based on the working historical data of the electronic component, input it into the LSTM neural network model for training, construct an electronic component performance prediction model, and obtain the variation function of the performance of the electronic component with working time" include: Transmit the working historical data of the electronic component into the LSTM neural network model for training to construct an electronic component performance prediction model; Based on the double-branch structure of the output layer in the electronic component performance prediction model, the main branch predicts the performance parameters at the current moment, and the auxiliary branch outputs the performance degradation rate. Through integral operation, the functional relationship between the performance parameters and the working time is constructed.

3. The method for monitoring the performance data of an electronic component according to claim 1, wherein The specific steps of "Obtain the current surface temperature field distribution data, partial discharge characteristic quantity and leakage current dynamic parameters of the electronic component, input them into the electronic component performance prediction model, and based on the changes of various factors, obtain multiple predicted variation trend functions of the performance of the electronic component in the future time period" include: Scan the surface of the electronic component through a non-contact infrared thermal imager array to obtain a two-dimensional temperature distribution matrix and extract characteristic parameters; Collect pulse signals through a high-frequency current sensor to obtain partial discharge characteristic quantities; Obtain current characteristic parameters through dual-channel synchronous acquisition; Input the preprocessed collected data into the electronic component performance prediction model, and based on the possible variation trends of the electronic component parameter data, generate multi-scenario predicted variation trend functions of the performance of the electronic component.

4. The method for monitoring the performance data of an electronic component according to claim 3, wherein The specific steps of "Input the preprocessed collected data into the electronic component performance prediction model, and based on the possible variation trends of the electronic component parameter data, generate multi-scenario predicted variation trend functions of the performance of the electronic component" include: Based on the characteristic parameters affecting the performance of the electronic component, obtain the variation ranges of each perturbation dimension; Through orthogonal experimental design, combine the parameter change scenarios to generate multiple scenarios of the performance change of the electronic component; Based on the characteristic parameter data under each scenario, input it into the electronic component performance prediction model to generate multiple predicted variation trend functions of the performance of the electronic component, and the obtained characteristic data includes: performance parameter time series, performance degradation acceleration, prediction of the time to reach the key threshold.

5. The method for monitoring the performance data of an electronic component according to claim 1, characterized in that, The specific steps of "Based on the obtained multiple predicted variation trend function data, compare it with the actual data of the performance of the electronic component collected in real time, calculate the error value, and obtain the predicted data with the smallest error" include: Perform dynamic error weight allocation based on the error contribution degrees of each characteristic parameter; Compare the real-time collected data with the data at the corresponding time nodes in the predicted change trend function, and obtain the error magnitude between the data of each predicted change trend function and the actually collected data through multi-dimensional error joint calculation; Based on the set of predicted change trend functions with the smallest error value in the error data, select it as the optimal electronic component performance prediction model; Based on the actual data of the electronic component performance collected at each moment, screen and obtain the prediction result of an optimal electronic component performance prediction model.

6. The method for monitoring the performance data of an electronic component according to claim 1, wherein The dynamic error analysis based on the predicted data with the smallest error to generate model correction parameters specifically includes: Based on the predicted data with the smallest error, obtain the fluctuation of the error through the error distribution diagram; Obtain the cause of the error under specific conditions through the relationship between the error and the working conditions; Obtain the deviation in data, model structure or parameter settings by comparing the error distribution diagram and the output of the model prediction; Based on the results of the dynamic error analysis, adjust the model parameters to generate model correction parameters, including: adjustment of the learning rate, addition of regularization terms, and adjustment of the training data volume.

7. The method for monitoring the performance data of an electronic component according to claim 1, wherein The online parameter update of the electronic component performance prediction model based on the model correction parameters to form a closed-loop optimization system specifically includes: Based on the obtained model correction parameters, perform online training on the model by the incremental learning method, and continuously optimize the model by using the real-time obtained electronic component performance data and error correction information; Through an automated parameter adjustment mechanism, when the error exceeds the set threshold, automatically trigger the update of the model parameters; Compare the prediction result of the model after each update with the real-time data again to form a continuous feedback process.

8. An electronic component performance data monitoring system, which is used to implement the electronic component performance data monitoring method according to any one of claims 1-7, is characterized in that Include: Model training module: The model training module is mainly used to input the working historical data of the electronic component into the LSTM neural network for training, construct a performance prediction model and obtain the function of the performance changing with time; Performance prediction model module: The performance prediction model module is mainly used to construct an electronic component performance prediction model based on the output structure of the LSTM network and predict multiple change trend functions in the future time period; Multi-scenario prediction analysis module: The multi-scenario prediction analysis module generates multiple prediction scenarios based on different working conditions and perturbation factors in combination with the orthogonal experimental design; Error calculation module: The error calculation module is mainly used to compare the real-time collected performance data with the prediction model results, calculate the error and perform dynamic weight allocation based on the error contribution degree to obtain the prediction model with the smallest error value; Online update and closed-loop optimization module: The online update and closed-loop optimization module is mainly used to perform incremental learning based on the model correction parameters, update the prediction model online, automatically adjust the parameters and continuously optimize the performance prediction results; Processor: The processor is mainly used for the calculation process of each formula and the construction calculation process of each model.

9. An electronic device, characterized in that, Include: At least one processor; And a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method for monitoring electronic component performance data as described in any one of claims 1-7.

10. A computer-readable storage medium storing computer-readable instructions, characterized in that, When the computer-readable instructions are executed by a processor, the method for monitoring electronic component performance data as described in any one of claims 1-7 is implemented.

Images