Health Status Assessment Method and Device for Isolated Switching Power Supply at the Front End of Aging Bench

Abstract

The embodiments of this specification provide a method and device for evaluating the health status of a front-end isolated switching power supply of an aging platform, which relates to the technical field of aging test platforms. The method includes: during the aging test process, the front-end isolated switching power supply of the aging test platform will generate monitoring signals such as voltage and current; the directly measured monitoring signals have a large amount of data and contain a lot of redundant information. First, feature extraction is performed on the directly measured data to obtain a feature sequence with high information density; the neighborhood component analysis algorithm is used to perform further spatial mapping on the feature sequence to realize the automatic segmentation of the health status and degradation status of new test samples; the multi-distance metric fusion algorithm with automatic weight optimization is used to perform quantitative calculation of the health status of the mapped feature sequence, and the health assessment of the front-end isolated switching power supply of the aging test platform is completed.

Description

Technical Field

[0001] The embodiments of this specification relate to the technical field of non-destructive reliability screening of integrated circuits, and particularly to a method and device for evaluating the health status of a front-end isolated switching power supply of an aging platform. Background Art

[0002] An aging test bench simulates various actual usage environments and working conditions, enabling a product to experience state changes similar to those after long-term use in a short period of time. Aging tests usually require long-term power-on operation of the test equipment to simulate the aging process of the product in actual use. Among them, the front-end isolated switching power supply can convert the input mains power into stable direct current to provide continuous and stable power for many devices under test on the aging test bench. Existing high-temperature aging products can achieve long-term monitoring of the test environment by means such as increasing in-machine long-term monitoring and over-stress protection mechanisms, and can respond in a timely manner when a product fails, realizing after-sales maintenance based on fault data. However, due to the lack of active guarantee technologies for high-temperature aging products in existing aging products, it is difficult to achieve the integrity of the test process and the consistency of test environment stress for related products, which is extremely likely to lead to major property losses such as the forced interruption of the test process due to product failures resulting in the destruction of millions of test devices, or the aging test being recognized as a failed test due to adverse effects such as additional stress introduced during the test period due to product performance degradation, causing ineffective waste of resources. At the same time, ensuring the quality of high-temperature aging tests can avoid over-aging and under-aging failures of integrated circuits, greatly reduce the failure rate of integrated circuits, and avoid large-scale electronic system failures and outages such as new energy vehicles, civil airliners, and energy storage power transmission stations due to integrated circuit failures.

[0003] During the aging test process, if a front-end isolated switching power supply fails, such as a sudden drop or disappearance of the output voltage, it will cause an unexpected interruption of the aging test. And during the use of the aging test bench, problems such as capacitor aging and degradation of the performance of switching tubes will gradually occur. Therefore, how to ensure the health status of the front-end isolated switching power supply of the aging test bench has become an urgent problem to be solved. Summary of the Invention

[0004] In view of this, the embodiments of this specification provide a method for evaluating the health status of a front-end isolated switching power supply of an aging platform. One or more embodiments of this specification also relate to a device for evaluating the health status of a front-end isolated switching power supply of an aging platform, a computing device, a computer-readable storage medium, and a computer program to solve the technical defects existing in the prior art.

[0005] According to the first aspect of the embodiments of this specification, a method for evaluating the health status of a front-end isolated switching power supply of an aging platform is provided, including:

[0006] Obtain the monitoring data of the front-end isolated switching power supply of the aging test bench at the current moment, and through feature extraction of the monitoring data, obtain feature data with high information density;

[0007] By inputting the feature data with high information density into the trained NCA (Neighborhood Components Analysis) algorithm model, obtain the preliminary health status evaluation result of the front-end isolated switching power supply of the aging test bench at the current moment;

[0008] If the preliminary health status evaluation result is a degradation state, use the multi-distance metric fusion method with automatic weight optimization to quantitatively calculate the health status of the feature data with high information density, obtain the health degree of the degradation state at the current moment, and use the health degree of the degradation state at the current moment as the final health status evaluation result of the front-end isolated switching power supply of the aging test bench.

[0009] Preferably, it further includes:

[0010] If the evaluation status result is a healthy state, use the healthy state as the final health status evaluation result of the front-end isolated switching power supply of the aging test bench.

[0011] Preferably, it further includes:

[0012] Obtain the first historical monitoring data and the second historical monitoring data of the full-life degradation cycle of the front-end isolated switching power supply of the aging test bench, use the first historical monitoring data as the training set, and use the second historical monitoring data as the test set;

[0013] Through feature extraction of the historical monitoring data in the training set, obtain the first historical feature data with high information density;

[0014] Determine the state labels corresponding to each moment of the first historical feature data with high information density; among them, the state labels include healthy state labels and degradation state labels;

[0015] Construct an NCA algorithm model, use the first historical feature data with high information density as the model input, and at the same time use the state labels corresponding to each moment of the first historical feature data with high information density as the model output to train and learn the NCA algorithm model, and obtain the trained NCA algorithm model and the optimized mapping matrix.

[0016] Preferably, it further includes:

[0017] Through feature extraction of the historical monitoring data in the test set, obtain the second historical feature data with high information density;

[0018] By inputting the historical feature data with the second highest information density into the trained NCA algorithm model, the state labels corresponding to each moment of the historical feature data with the second highest information density are obtained;

[0019] All healthy state labels are extracted from the state labels corresponding to each moment of the historical feature data with the second highest information density, and all degraded state labels are discarded;

[0020] Based on the historical feature data corresponding to all healthy state labels, healthy baseline data is constructed.

[0021] Preferably, using a multi-distance metric fusion method with automatic weight optimization to quantitatively calculate the health state of the feature data with high information density, obtaining the health degree of the degraded state at the current moment includes:

[0022] Calculate the Mahalanobis distance, cosine similarity, and Manhattan distance between the feature data with high information density and the healthy baseline data respectively;

[0023] Through multi-distance metric fusion with automatic weight optimization of the Mahalanobis distance, the cosine similarity, and the Manhattan distance, the health degree of the degraded state at the current moment is obtained.

[0024] Preferably, calculating the Mahalanobis distance between the feature data with high information density and the healthy baseline data includes:

[0025] Based on the optimized mapping matrix, calculate the Mahalanobis distance between the feature data with high information density and the healthy baseline data.

[0026] Preferably, through multi-distance metric fusion with automatic weight optimization of the Mahalanobis distance, the cosine similarity, and the Manhattan distance, obtaining the health degree of the degraded state at the current moment includes:

[0027]

[0028] Wherein, represents the health degree of the degraded state at the current moment, represents the Mahalanobis distance; represents the cosine similarity; represents the Manhattan distance; 、 and represent weights respectively, and + + = 1;

[0029] Among them, by measuring the monotonicity and smoothness of the health curve in the health degree of the degradation state at the current moment, the monotonicity loss and the smoothness loss are calculated respectively, and taking the monotonicity loss and the smoothness loss as the objective functions, the multi-objective genetic algorithm is used to optimize the fusion weights of multiple distance metrics to obtain the best weight combination.

[0030] Preferably, after obtaining the final health curve, evaluate its monotonicity and smoothness, calculate the monotonicity loss and the smoothness loss, and use them as the objective functions to optimize 、 and through the multi-objective genetic algorithm.

[0031] According to the second aspect of the embodiments of the present specification, a health state evaluation device for the front-end isolated switching power supply of an aging platform is provided, including:

[0032] A feature extraction module, configured to obtain the monitoring data of the front-end isolated switching power supply of the aging test platform at the current moment, and obtain feature data with high information density by extracting features from the monitoring data;

[0033] A preliminary evaluation module, configured to obtain the preliminary health state evaluation result of the front-end isolated switching power supply of the aging test platform at the current moment by inputting the feature data with high information density into the trained nearest neighbor component analysis (NCA) algorithm model;

[0034] A final evaluation module, configured to, if the preliminary health state evaluation result is a degradation state, use the multi-distance metric fusion method with automatic weight optimization to quantitatively calculate the health state of the feature data with high information density, obtain the health degree of the degradation state at the current moment, and use the health degree of the degradation state at the current moment as the final health state evaluation result of the front-end isolated switching power supply of the aging test platform.

[0035] According to the third aspect of the embodiments of the present specification, a computing device is provided, including:

[0036] A memory and a processor;

[0037] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the steps of any one of the health state evaluation methods for the front-end isolated switching power supply of the aging platform.

[0038] According to the fourth aspect of the embodiments of the present specification, a computer-readable storage medium is provided, which stores computer-executable instructions, and when the instructions are executed by a processor, the steps of any one of the health state evaluation methods for the front-end isolated switching power supply of the aging platform are implemented.

[0039] According to a fifth aspect of the embodiments of the present specification, a computer program is provided, wherein when the computer program is executed in a computer, the computer is made to execute the steps of the above-mentioned method for evaluating the health state of the front-end isolated switching power supply of the aging bench.

[0040] In the embodiments of the present specification, by obtaining the monitoring data of the front-end isolated switching power supply of the aging test bench at the current moment, and through feature extraction of the monitoring data, feature data with high information density is obtained; by inputting the feature data with high information density into the trained nearest neighbor component analysis (NCA) algorithm model, a preliminary health state evaluation result of the front-end isolated switching power supply of the aging test bench at the current moment is obtained; if the preliminary health state evaluation result is a degradation state, then a multi-distance metric fusion method with automatic weight optimization is used to quantitatively calculate the health state of the feature data with high information density, obtaining the health degree of the degradation state at the current moment, and using the health degree of the degradation state at the current moment as the final health state evaluation result of the front-end isolated switching power supply of the aging test bench. By performing a health assessment on the front-end isolated switching power supply of the aging test bench, potential hidden dangers inside the power supply can be discovered in advance, repairs or replacements can be carried out in a timely manner, test interruptions can be avoided, and the aging test can be ensured to proceed smoothly according to the plan. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a flowchart of a method for evaluating the health state of the front-end isolated switching power supply of an aging bench provided by an embodiment of the present specification;

[0042] Figure 2 is a schematic diagram of a health state evaluation process of the front-end isolated switching power supply of an aging bench based on nearest neighbor component analysis and multi-distance metric fusion with automatic weight optimization provided by an embodiment of the present specification;

[0043] Figure 3 is a flowchart of a method for evaluating the health state of the front-end isolated switching power supply of an aging bench provided by an embodiment of the present specification;

[0044] Figure 4 is a schematic diagram of the structure of a device for evaluating the health state of the front-end isolated switching power supply of an aging bench provided by an embodiment of the present specification;

[0045] Figure 5 is a block diagram of the structure of a computing device provided by an embodiment of the present specification. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] In the following description, numerous specific details are set forth in order to provide a thorough understanding of this specification. However, this specification can be implemented in many other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the connotation of this specification. Therefore, this specification is not limited by the specific implementations disclosed below.

[0047] The terms used in one or more embodiments of this specification are for the purpose of describing specific embodiments only and are not intended to limit one or more embodiments of this specification. The singular forms "a", "the", and "said" used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in one or more embodiments of this specification refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0048] It should be understood that although the terms first, second, etc. may be used in one or more embodiments of this specification to describe various information, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of this specification, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "when" or "while" or "in response to determining".

[0049] In this specification, a method for evaluating the health status of a front-end isolated switching power supply of an aging platform is provided. This specification also relates to a device for evaluating the health status of a front-end isolated switching power supply of an aging platform, a computing device, a computer-readable storage medium, and a computer program, which will be described in detail one by one in the following embodiments.

[0050] Figure 1 The flowchart of a method for evaluating the health status of a front-end isolated switching power supply of an aging platform provided according to an embodiment of this specification is shown, which specifically includes the following steps.

[0051] Step 101: Obtain the monitoring data of the front-end isolated switching power supply of the aging test platform at the current moment, and obtain feature data with high information density by performing feature extraction on the monitoring data;

[0052] Step 102: Input the feature data with high information density into the trained nearest component analysis (NCA) algorithm model to obtain the preliminary health status evaluation result of the front-end isolated switching power supply of the aging test platform at the current moment;

[0053] Step 103: If the preliminary health status assessment result is a degradation state, use the multi-distance metric fusion method with automatic weight optimization to quantitatively calculate the health status of the high-information-density feature data, obtain the health degree of the degradation state at the current moment, and use the health degree of the degradation state at the current moment as the final health status assessment result of the front-end isolated switching power supply of the aging test bench.

[0054] In an optional implementation manner, the health status assessment method for the front-end isolated switching power supply of the aging bench further includes:

[0055] If the assessment status result is a healthy state, use the healthy state as the final health status assessment result of the front-end isolated switching power supply of the aging test bench.

[0056] In an optional implementation manner, the health status assessment method for the front-end isolated switching power supply of the aging bench further includes:

[0057] Obtain the first historical monitoring data and the second historical monitoring data of the full-life degradation cycle of the front-end isolated switching power supply of the aging test bench, use the first historical monitoring data as the training set, and use the second historical monitoring data as the test set;

[0058] Through feature extraction of the historical monitoring data in the training set, obtain the first high-information-density historical feature data;

[0059] Determine the state labels corresponding to each moment of the first high-information-density historical feature data; wherein, the state labels include a healthy state label and a degradation state label;

[0060] Construct an NCA algorithm model, use the first high-information-density historical feature data as the model input, and at the same time use the state labels corresponding to each moment of the first high-information-density historical feature data as the model output to train and learn the NCA algorithm model, and obtain the trained NCA algorithm model and the optimized mapping matrix.

[0061] In an optional implementation manner, the health status assessment method for the front-end isolated switching power supply of the aging bench further includes:

[0062] Through feature extraction of the historical monitoring data in the test set, obtain the second high-information-density historical feature data;

[0063] By inputting the second high-information-density historical feature data into the trained NCA algorithm model, obtain the state labels corresponding to each moment of the second high-information-density historical feature data;

[0064] Extract all the healthy state labels from the state labels corresponding to each moment of the historical feature data with the second highest information density, and discard all the degraded state labels at the same time;

[0065] Construct healthy baseline data based on the historical feature data corresponding to all the healthy state labels.

[0066] In an alternative embodiment, a multi-distance metric fusion method with automatically optimized weights is used to quantitatively calculate the health state of the feature data with high information density, and the health degree of the degraded state at the current moment is obtained, including:

[0067] Calculate the Mahalanobis distance, cosine similarity and Manhattan distance between the feature data with high information density and the healthy baseline data respectively;

[0068] Through multi-distance metric fusion with automatically optimized weights for the Mahalanobis distance, the cosine similarity and the Manhattan distance, obtain the health degree of the degraded state at the current moment.

[0069] In an alternative embodiment, calculating the Mahalanobis distance between the feature data with high information density and the healthy baseline data includes:

[0070] Based on the optimized mapping matrix, calculate the Mahalanobis distance between the feature data with high information density and the healthy baseline data.

[0071] In an alternative embodiment, through multi-distance metric fusion of the Mahalanobis distance, the cosine similarity and the Manhattan distance, obtaining the health degree of the degraded state at the current moment includes:

[0072]

[0073] Wherein, represents the health degree of the degraded state at the current moment, represents the Mahalanobis distance; represents the cosine similarity; represents the Manhattan distance; , and represent the weights respectively, and + + = 1;

[0074] Wherein, by measuring the monotonicity and smoothness of the health degree curve in the health degree of the degraded state at the current moment, calculate the monotonicity loss and the smoothness loss respectively, and use the monotonicity loss and the smoothness loss as the objective functions, and use the multi-objective genetic algorithm to optimize the fusion weights of the multi-distance metrics to obtain the best weight combination.

[0075] In the embodiment of this specification, by obtaining the monitoring data of the front-end isolated switching power supply of the aging test bench at the current moment, and through feature extraction of the monitoring data, feature data with high information density is obtained; by inputting the feature data with high information density into the trained nearest neighbor component analysis (NCA) algorithm model, a preliminary health status evaluation result of the front-end isolated switching power supply of the aging test bench at the current moment is obtained; if the preliminary health status evaluation result is a degraded state, then a multi-distance metric fusion method with automatically optimized weights is used to quantitatively calculate the health status of the feature data with high information density, obtaining the health degree of the degraded state at the current moment, and taking the health degree of the degraded state at the current moment as the final health status evaluation result of the front-end isolated switching power supply of the aging test bench. By performing a health assessment on the front-end isolated switching power supply of the aging test bench, potential hidden dangers inside the power supply can be discovered in advance, repaired or replaced in a timely manner, avoiding test interruption, and ensuring that the aging test can proceed smoothly according to the plan.

[0076] Figure 2 The figure shows a schematic diagram of the health status evaluation process of the front-end isolated switching power supply of the aging bench based on nearest neighbor component analysis and multi-distance metric fusion with automatically optimized weights provided by an embodiment of this specification. As Figure 2 shown, during the aging test process, signals such as current and voltage of the front-end isolated switching power supply of the test bench are monitored by sensors to obtain a large amount of monitoring data for the state evaluation of the switching power supply. Since the directly measured monitoring signal data is large in quantity and contains a large amount of redundant information, it is first necessary to perform feature extraction on the directly monitored data to obtain a feature sequence with high information density.

[0077] Based on the extracted feature sequence, the health assessment process is further realized: using the nearest neighbor component analysis (NCA) metric learning algorithm to achieve adaptive state partitioning and distance metric of the feature data. The main goal of the NCA algorithm is to learn a Mahalanobis distance metric such that, under this distance metric, the distance between samples of the same class is as small as possible, and the distance between samples of different classes is as large as possible. The NCA algorithm will consider the neighbor relationship of the feature data sample points. For a set of feature data samples , where represents the i-th feature sample. NCA will calculate the distance between each feature sample and other feature samples , and this distance is calculated through a transformed space. The calculation formula is as follows:

[0078]

[0079] Among them, the optimized mapping matrix L is a learnable positive semi - definite metric matrix. The core of the NCA algorithm lies in adjusting the optimized mapping matrix L through training, mapping the feature samples into a new space, so that the distance between samples of the same class is as small as possible after mapping, and the distance between samples of different classes is as large as possible. This process is completed by maximizing the probability of determining the same - class feature samples as neighbors through the gradient optimization algorithm.

[0080] Compared with the health assessment method based on traditional distance metrics, which requires artificially setting a health baseline for constructing health indicators and has certain instability. The NCA metric learning algorithm, through the above - mentioned training methods of in - class aggregation and inter - class separation, learns and optimizes the mapping matrix L, can effectively separate and map the feature samples into a new space, making the healthy feature samples and degraded feature samples show obvious clustering and separation in the new space, and automatically cutting and determining the health baseline in this process.

[0081] The health assessment based on the NCA metric learning algorithm can be trained and learned according to the characteristics of samples, can perform adaptive recognition and mapping of features, and has strong adaptability; during the training process, the health and degradation states of the training feature samples are set, and the mapping matrix is obtained through training. Then, during the test process, it is applied to the test feature samples, and the mapping and state separation of the test data can be carried out.

[0082] Specifically, in the training stage of the NCA metric learning algorithm model, the training data with health - state labels is input , h = 0 or 1; where is the feature sample, is the corresponding label, 0 and 1 represent the healthy state and the degraded state respectively, and n is the number of feature samples obtained by resampling the entire degradation curve sliding window. The NCA model continuously optimizes the mapping matrix of the input data to the low - dimensional feature space in this process, strengthens the classification ability of the healthy - state data and the degraded - state data, so that the NCA model obtains the ability to automatically segment the health baseline of the test data; in the test stage of the algorithm model, the test data is input, and the NCA model can map the health - baseline data and the degraded data into a space with a larger inter - class distance according to the feature mapping matrix optimized and learned during the training process, and easily segment the health baseline of the test data through the unsupervised clustering algorithm to obtain the test data set , represents the test data set of all healthy labels, represents the test data set of all degraded labels, avoiding the blindness of artificial segmentation of the health baseline.

[0083] To further achieve the health status assessment of the front-end isolated switching power supply at the current moment, a multi-distance metric fusion method with automatic weight optimization is used. The Mahalanobis distance, cosine similarity, and Manhattan distance between sample data are comprehensively considered and weighted fusion is performed to obtain the health degree at the current moment. At the same time, the health degree degradation curve up to this moment is output, providing reliable data support for the further study of the remaining service life prediction of the front-end isolated switching power supply. Among them, the Mahalanobis distance is calculated as shown in the following formula:

[0084]

[0085] For two n-dimensional space vectors and , the calculation formula of their cosine similarity is as follows:

[0086]

[0087] The calculation formula of the Manhattan distance is as follows:

[0088]

[0089] On this basis, further multi-distance metric fusion with automatic weight optimization is carried out. The calculation process of the final distance metric result is as follows:

[0090]

[0091] Among them, , and represent weights respectively, and + + = 1.

[0092] Furthermore, by measuring the monotonicity and smoothness of the health degree curve in the health degree of the degradation state at the current moment, the monotonicity loss and the smoothness loss are calculated respectively. Taking the monotonicity loss and the smoothness loss as the objective functions, the multi-objective genetic algorithm is used to optimize the fusion weights of the multi-distance metrics to obtain the best weight combination.

[0093] It should be noted that the evaluation indicators include the monotonicity and smoothness of the health degree curve, and the monotonicity loss and the smoothness loss are calculated. The health state of the equipment gradually degrades over time. Assuming the obtained health degree curve is , then:

[0094] Monotonicity loss:

[0095]

[0096] Smoothness loss:

[0097]

[0098] Furthermore, the pseudo-code of the health assessment algorithm for multi-distance metric fusion with automatic optimization of NCA and weights is shown in Table 1:

[0099] Table 1: Pseudo-code of the health assessment algorithm for multi-distance metric fusion with automatic optimization of NCA and weights

[0100]

[0101] Furthermore, the automatic optimization of the weights is reflected in the consistency evaluation of the degradation trend of the health curve, and finally the fusion weights are optimized based on the evaluation results.

[0102] Figure 3 is a flowchart of the method for evaluating the health status of the front-end isolated switching power supply of the aging platform provided by an embodiment of this specification, as Figure 3 shown, including:

[0103] Step S301: Preprocessing the voltage signal of the front-end isolated switching power supply of the aging test bench;

[0104] The health assessment case of the front-end isolated switching power supply of the aging test bench uses the voltage data during the degradation process of the front-end isolated switching power supply. In this case, two voltage curves that complete the full-life degradation cycle are used for training and testing respectively. First, the training curve is subjected to maximum-minimum normalization, and the same normalization operation is performed on the test curve using the normalization parameters of the training curve. Then, sliding window resampling is performed on the two voltage degradation data to achieve sample cutting. The window size of the sliding window is 10, and the sliding window step is 10, obtaining two sets of resampled voltage degradation data sets for the training of the nearest neighbor component analysis algorithm and the testing of the health assessment method respectively. The state of the front-end isolated switching power supply of the aging test bench can be divided into a healthy state and a degradation state, and the training data set is marked according to the state of the samples as shown in Table 2.

[0105] Table 2: Marking of training samples

[0106]

[0107] Step S302: Training the nearest neighbor component analysis model;

[0108] Use the healthy samples and degraded samples of the training dataset as the input of the NCA metric learning model, and their corresponding calibrated state feature labels as the output to train the metric learning model. Set the maximum number of iterations to 100, and use the cross-validation method for parameter adjustment. Continuously optimize the random state random_state and learning rate learning_rate of the algorithm model to make the training effect accuracy optimal. Table 3 below shows the NCA model parameters.

[0109] Table 3: NCA metric learning model parameters

[0110]

[0111] Based on the trained NCA metric learning model and the learned feature matrix L, the voltage samples of the front-end isolated switching power supply can be adaptively mapped to a new distribution space, so as to completely separate the healthy and degraded features, reduce the data aliasing of different states, and facilitate the calculation of subsequent health indicators.

[0112] Step S303: Health assessment of test samples.

[0113] Input the test samples into the trained NCA model, and it can be obtained whether the test samples are in a healthy state or a degraded state. Perform mean processing on the samples in the healthy state to construct a healthy baseline, and calculate the health indicators for the samples in the degraded state. Calculate the Mahalanobis distance, cosine similarity, and Manhattan distance between the degraded samples and the healthy baseline in sequence. The learnable feature matrix used in calculating the Mahalanobis distance is the L obtained in step 302. Considering the normalization processing of the data, the difference in degraded features is mainly reflected in the inclination angle. The weight distribution in the final distance metric result formula is shown in Table 4 below. , cosine similarity and Manhattan distance , where the learnable feature matrix used in calculating the Mahalanobis distance is the L obtained in step 302. Considering the normalization processing of the data, the difference in degraded features is mainly reflected in the inclination angle. The weight distribution in the final distance metric result formula is shown in Table 4 below.

[0114] Table 4: Initial weight distribution of multi-distance metric fusion with automatic weight optimization

[0115]

[0116] The embodiment of the present application also provides a method for extracting health features of a primary (first-level) switching power supply of an aging platform, which is used to extract features from the monitoring data in the process of evaluating the health state of the primary (first-level) switching power supply of the aging platform.

[0117] Furthermore, the flowchart of a method for extracting health features of a first-level switching power supply of an aging platform in the embodiment of the present application specifically includes the following steps.

[0118] Step 402: Obtain multiple historical signal values of the analog signal of the first-level switching power supply of the aging platform.

[0119] Specifically, the aging platform is an integrated circuit high-temperature aging test bench. By continuously applying a certain electrical stress to components for a long time, various physical and chemical reaction processes inside the components are accelerated, so that various potential faults inside the components are exposed early, thereby eliminating early failure products and enabling electronic components to enter a period with low failure rate and relatively stable performance from the beginning of use. The primary switching power supply is used to provide the power required for the entire aging test bench.

[0120] Since monitoring signals such as current and voltage are generated during the operation of the primary switching power supply of the aging test bench, the analog signals described in the embodiments of this specification may include voltage signals and current signals, and the current signals may further include capacitive current signals and inductive current signals.

[0121] In addition, the multiple historical signal values described in the embodiments of this specification may be historical time-series data of analog signals.

[0122] Step 404: Encode the multiple historical signal values through the Transformer model encoder in the multi-layer model encoder to obtain the self-attention features in the multiple historical signal values. The multi-layer model encoder is a neural network structure that fuses context features constructed by the Transformer model encoder and the stacked autoencoder, including an input layer, an intermediate layer, and an output layer. The intermediate layer includes a Transformer model encoder structure and a stacked autoencoder connected in sequence.

[0123] In an alternative embodiment, the multi-layer model encoder further includes a multi-head attention layer, a frequency-domain feature extraction layer, and a fusion layer. The frequency-domain feature extraction layer is used to extract frequency-domain features, and the fusion layer is used to fuse the data features extracted by the Transformer model encoder and the frequency-domain feature extraction layer.

[0124] Correspondingly, the method further includes: obtaining the transformation result generated by the multi-head attention layer performing a linear transformation on the multiple historical signal values; encoding the transformation result through the Transformer model encoder to generate the self-attention features in the multiple historical signal values; extracting the frequency-domain features of the multiple historical signal values output by the frequency-domain feature extraction layer through the frequency-domain feature extraction layer; and fusing the self-attention features and the frequency-domain features through the fusion layer to obtain the target fusion features output by the fusion layer.

[0125] Among them, the process of fusing the self-attention feature and the frequency-domain feature through the fusion layer to obtain the target fusion feature output by the fusion layer includes: splicing the self-attention feature and the frequency-domain feature through the fusion layer to obtain the target fusion feature output by the fusion layer.

[0126] Specifically, the multi-layer model encoder is a multi-layer Transformer model encoder, including an input layer, an intermediate layer, and an output layer. The intermediate layer includes a Transformer model encoder structure and a stacked autoencoder connected in sequence. Among them, the Transformer model encoder includes a multi-head attention sub-layer, a fully connected layer, and a feed-forward neural network. In addition, the multi-layer Transformer model encoder also includes a frequency-domain feature extraction layer and a fusion layer. The multi-head attention sub-layer is used to extract the multi-head attention feature in the historical time-series data. The frequency-domain feature extraction layer is used to extract the frequency-domain feature in the historical time-series data. The fusion layer is used to fuse the multi-head attention feature extracted by the multi-head attention sub-layer and the frequency-domain feature extracted by the frequency-domain feature extraction layer to achieve the purpose of fully extracting features.

[0127] In an alternative embodiment, after obtaining multiple historical signal values of the analog signal of the aging bench's primary switching power supply, it further includes: performing sliding window cutting on the multiple historical signal values according to the first window width and the first step length to generate a first sliding window cutting result, and determining an initial sample data set based on the first sliding window cutting result; performing normalization processing on the initial sample data set to generate a target sample data set. Correspondingly, encoding the multiple historical signal values through the Transformer model encoder in the multi-layer model encoder includes: encoding the target sample data set through the Transformer model encoder in the multi-layer model encoder.

[0128] A processing procedure for feature extraction provided by an embodiment of this specification includes the following steps.

[0129] Step 502: Obtain the fault prediction data of the aging bench's primary switching power supply.

[0130] Step 504: Perform comprehensive preprocessing on the fault data.

[0131] Step 506: Perform sliding window cutting on the preprocessing result according to the first window width and the first step length to generate a first sliding window cutting result, and determine an initial sample data set based on the first sliding window cutting result.

[0132] Step 508: Perform normalization processing on the initial sample data set to generate a target sample data set.

[0133] Step 510: Construct a training dataset and a test dataset based on the target sample dataset.

[0134] Step 512: Extract multi-head attention features from the training dataset through the Transformer model encoder.

[0135] Step 514: Extract frequency-domain features from the training dataset.

[0136] Step 516: Perform deep feature fusion on the concatenated features through a stacked autoencoder to obtain the health features corresponding to the primary switch power supply of the aging platform.

[0137] Step 518: Output the health features.

[0138] For obtaining the fault prediction data of the primary switch power supply of the aging platform, the voltage and current signals of the primary switch power supply in the integrated circuit high-temperature dynamic aging detection system can be monitored by sensors to obtain the original fault prediction data of the primary switch power supply of the aging platform, that is, the historical time-series data of the voltage and current signals.

[0139] For comprehensively preprocessing the fault data, the collected original fault prediction data can be sent to the fault data preprocessing unit for comprehensive processing to obtain a training dataset and a test dataset. Specifically, it can be achieved through the following steps:

[0140] Step 602: Perform sliding window cutting on the historical time-series data to construct an initial sample dataset.

[0141] Specifically, the historical time-series data of any primary switch power supply collected by the sensor is X, , and sliding window cutting is performed on X to generate the corresponding initial sample dataset. When the first window width is W and the first step length is S, the number of samples generated is:

[0142]

[0143] Then the corresponding initial sample dataset generated is , for each sample in take the data with a length of as the training data, and take the data with a length of W - as the prediction data corresponding to this training data.

[0144] Step 604: Perform min-max normalization processing on the training dataset.

[0145] To improve the data expression ability and accelerate the convergence speed of the subsequent model training, it is necessary to normalize the training data set. Specifically, the amplitude of the original parameters is scaled by the min-max normalization method to complete the linear transformation of the data. For a single sample data , the normalization process is achieved through the following formula to obtain the normalized sample data set .

[0146]

[0147] Step 608: Construct the training data set and the test data set.

[0148] Select the first r% of the data from all the data as the training data set, and the remaining data as the test data set to verify the model prediction performance. Generally speaking, r is usually taken as 60 - 80, and in the embodiments of this specification, r is taken as 70.

[0149] For the multi-head attention feature extraction of the processed fault data based on the Transformer model encoder, the training data set obtained after being processed by the comprehensive data preprocessing module can be respectively sent into the Transformer model encoder and the frequency domain feature extraction layer to obtain self-attention features and frequency domain features.

[0150] Another processing process of feature extraction provided by the embodiments of this specification specifically includes the following steps.

[0151] Step 702: Obtain the training data set.

[0152] Step 704: Construct the Transformer model encoder.

[0153] Step 706: Train the constructed Transformer model encoder.

[0154] Step 708: Perform self-attention feature extraction on the training data set based on the trained Transformer model encoder.

[0155] Step 710: Perform sliding window cutting on the normalized target sample data set.

[0156] Step 712: Extract at least two initial frequency domain features based on the sliding window cutting result.

[0157] Step 714: Normalize at least two initial frequency domain features.

[0158] Step 716: Perform feature splicing on the self-attention features and the frequency domain features.

[0159] When performing multi-head attention feature extraction on the processed fault data using the Transformer model encoder, first, a Transformer model encoder can be constructed using the training dataset. Specifically, the model can be pre-trained using the training dataset. The Transformer model usually requires the input data to be three-dimensional, so a training dataset needs to be constructed. The purpose of constructing the training dataset is to meet the input requirements of the model. The method is to convert the data format of the training dataset into , where is the number of samples, is the data length of each sample, and 3 is the number of features, representing three features: the output voltage signal, the capacitor current signal, and the inductor current signal at each time point. After the training dataset is constructed, the constructed training sample dataset can be input into the Transformer model encoder.

[0160] The Transformer model is a model established based on the Seq-to-Seq framework. Compared with classical deep learning models, the most prominent advantage of the Transformer model is the use of the multi-head attention mechanism. The purpose of the multi-head attention sublayer is to assign different importance to words / tokens in the sequence from multiple aspects. The Transformer model mainly consists of parts such as input, encoder, decoder, and output.

[0161] The position encoding layer in the Transformer model is to determine the position information of the sequence. Since there are no recurrent layers and convolutional layers in RNN and CNN, and only relying on the self-attention mechanism cannot obtain the order information of the input, the order information of the sequence needs to be actively passed to the model. The Transformer model uses a combination of sine and cosine functions to perform position encoding on the sequence. The calculation method is as follows:

[0162]

[0163]

[0164] where pos is the position of the current sequence; i is the dimension; is the dimension of the input features.

[0165] The multi-head attention mechanism in the Transformer model performs operations in parallel using multiple attention mechanisms, and then splices the operation results through a linear transformation. The core technology in the Transformer model is the multi-head attention mechanism. The multi-head attention mechanism is used to extract the dependency relationship features between data, capture the correlation between data, and establish a context prediction model. The calculation method is as follows:

[0166]

[0167]

[0168]

[0169]

[0170]

[0171]

[0172] Among them, Q is the query matrix; K is the key matrix; V is the value matrix; 、 、 are trainable parameter matrices; X is the processed input; is the dimension of the key matrix; 、 、 、 are learnable parameter matrices.

[0173] In the encoding part and the decoding part of the Transformer model, there are also feed-forward networks and summation and normalization. The calculation formula of the feed-forward neural network is as follows:

[0174]

[0175] Among them, x is the input; 、 、 、 are parameters that can be obtained through training.

[0176] The calculation formula of summation and normalization is as follows:

[0177]

[0178] Among them, x is the input; is the result after being processed by the module.

[0179] The model processing flow is as follows: The input data is sent to the encoding part after position encoding, and the output of the encoding part is flattened and then sent to the decoder to map the feature parameters in the high-dimensional hidden layer to the original input data, so as to train the feature extraction ability of the model.

[0180] Secondly, select appropriate iteration times and loss functions, and input the constructed dataset into the feature extraction model to repeatedly perform the forward propagation and backward propagation iterative calculation process; during this process, continuously adjust the model parameters of the embedding dimension, the number of attention heads, the number of encoder and decoder layers to complete the pre-training of the model.

[0181] Next, extract the encoding layer and the decoding layer of the pre-trained model, retain their weight parameters, and construct them into the encoder of the trained Transformer model.

[0182] Finally, based on the encoder of the pre-trained Transformer model for the training dataset perform self-attention feature extraction to obtain a self-attention feature set .

[0183] Step 406: Extract the frequency-domain features of the multiple historical signal values, and fuse the self-attention features with the frequency-domain features to generate target fusion features.

[0184] In an alternative embodiment, the extracting the frequency-domain features of the multiple historical signal values includes:

[0185] Perform sliding window cutting on the target sample dataset according to the second window width and the second step length to generate a second sliding window cutting result, and extract the frequency-domain features of the multiple historical signal values based on the second sliding window cutting result.

[0186] Further, the extracting the frequency-domain features of the multiple historical signal values based on the second sliding window cutting result includes: performing Fourier transform on each to-be-processed sample dataset included in the second sliding window cutting result to generate a time-series energy spectrum corresponding to each to-be-processed sample dataset; respectively extracting the eigenvalues corresponding to at least two initial frequency-domain features of the time-series energy spectrum; using the min-max normalization algorithm to normalize the eigenvalues to generate the frequency-domain features of the multiple historical signal values.

[0187] Wherein, the at least two initial frequency-domain features include: peak value, bandwidth, center frequency, shape factor, power spectral density.

[0188] Specifically, to extract the frequency-domain features of the multiple historical signal values, it is specifically possible to perform frequency-domain feature extraction based on expert knowledge on the training dataset generated by sliding window cutting .

[0189] Further, it is also possible to perform sliding window cutting on the aforementioned target sample dataset generated by sliding window cutting according to the second window width and the second step length to generate a second sliding window cutting result, and extract the frequency-domain features of the multiple historical signal values based on the second sliding window cutting result.

[0190] Wherein, when performing sliding window cutting on the target sample dataset, if the second window width is , the second step length is 1, for the sample can be cut out samples, and the length of each sample is , that is, obtain .

[0191] After the sliding window cutting is completed, the Fourier transform can be performed on each sample data set to be processed included in the second sliding window cutting result to convert it to the frequency domain, and the corresponding frequency domain representation F' (temporal energy spectrum) can be obtained.

[0192] For the frequency domain representation F' of each sample data set to be processed, the peak value, bandwidth, center frequency, shape factor, and power spectral density of the energy spectrum are respectively extracted as five initial frequency domain features. Each initial frequency domain feature includes a frequency domain feature and the corresponding eigenvalue of the frequency domain feature. Among them, for the window data The initial frequency domain features extracted are , so for the sample The initial frequency domain features extracted are . After the extraction is completed, the eigenvalue of each initial frequency domain feature can be normalized by using the maximum-minimum value normalization algorithm to generate the frequency domain features of multiple historical signal values.

[0193] After the frequency domain features and self-attention features (high-dimensional hidden layer features) are extracted, the extracted frequency domain features and self-attention features can be feature concatenated.

[0194] For the training data set in each sample self-attention feature extraction and frequency domain feature extraction are performed and feature fusion is carried out. Let the dimension of the fusion feature be , then the training data set can be reorganized into two-dimensional fusion feature matrix.

[0195] Step 408: Encode the target fusion feature through the stacked autoencoder in the multi-layer model encoder to generate the health feature corresponding to the aging bench primary switch power supply.

[0196] Specifically, the autoencoder is an unsupervised learning method, whose purpose is to learn the effective representation or encoding of data and achieve self-learning by reconstructing the input data. The stacked autoencoder SAE is an extension based on this. Through multi-layer non-linear transformation, the original data is mapped to an increasingly abstract and compact feature space, thereby realizing deeper feature learning and representation learning, which is of great significance for the dimensionality reduction, feature extraction, and classification tasks of complex data.

[0197] Corresponding to the above method embodiment, this specification also provides an embodiment of an aging bench front-end isolated switch power supply health state evaluation device, Figure 4 showing the structural schematic diagram of an aging bench front-end isolated switch power supply health state evaluation device provided by an embodiment of this specification. AsFigure 4 As shown in the figure, the device includes:

[0198] A feature extraction module 401, configured to obtain the monitoring data of the front-end isolated switching power supply of the aging test bench at the current moment, and obtain feature data with high information density by extracting features from the monitoring data;

[0199] A preliminary evaluation module 402, configured to obtain a preliminary health status evaluation result of the front-end isolated switching power supply of the aging test bench at the current moment by inputting the feature data with high information density into the trained Nearest Neighbor Component Analysis (NCA) algorithm model;

[0200] A final evaluation module 403, configured to, if the preliminary health status evaluation result is a degradation state, use a multi-distance metric fusion method with automatic weight optimization to quantitatively calculate the health status of the feature data with high information density, obtain the health degree of the degradation state at the current moment, and use the health degree of the degradation state at the current moment as the final health status evaluation result of the front-end isolated switching power supply of the aging test bench.

[0201] Furthermore, the final evaluation module 403 is further configured to, if the evaluation status result is a healthy state, use the healthy state as the final health status evaluation result of the front-end isolated switching power supply of the aging test bench.

[0202] Optionally, it further includes a training module 404, configured to:

[0203] Obtain the first historical monitoring data and the second historical monitoring data of the full-life degradation cycle of the front-end isolated switching power supply of the aging test bench, use the first historical monitoring data as the training set, and use the second historical monitoring data as the test set;

[0204] Obtain the first historical feature data with high information density by extracting features from the historical monitoring data in the training set;

[0205] Determine the state labels corresponding to each moment of the first historical feature data with high information density; wherein, the state labels include a healthy state label and a degradation state label;

[0206] Construct an NCA algorithm model, use the first historical feature data with high information density as the model input, and use the state labels corresponding to each moment of the first historical feature data with high information density as the model output to train and learn the NCA algorithm model, and obtain the trained NCA algorithm model and the optimized mapping matrix.

[0207] Optionally, the training module 404 is further configured to:

[0208] By extracting features from the historical monitoring data in the test set, historical feature data with the second-highest information density is obtained;

[0209] By inputting the historical feature data with the second-highest information density into the trained NCA algorithm model, the state labels corresponding to each moment of the historical feature data with the second-highest information density are obtained;

[0210] All healthy state labels are extracted from the state labels corresponding to each moment of the historical feature data with the second-highest information density, while all degraded state labels are discarded;

[0211] Based on the historical feature data corresponding to all healthy state labels, healthy baseline data is constructed.

[0212] Optionally, the final evaluation module 403 is specifically configured to calculate the Mahalanobis distance, cosine similarity, and Manhattan distance between the feature data with high information density and the healthy baseline data respectively;

[0213] Through multi-distance metric fusion with automatic weight optimization for the Mahalanobis distance, cosine similarity, and Manhattan distance, the health degree of the degraded state at the current moment is obtained, and its calculation formula is:

[0214]

[0215] where, represents the health degree of the degraded state at the current moment, represents the Mahalanobis distance; represents the cosine similarity; represents the Manhattan distance; , and represent weights respectively, and + + = 1.

[0216] Furthermore, by measuring the monotonicity and smoothness of the health degree curve in the health degree of the degraded state at the current moment, the monotonicity loss and smoothness loss are calculated respectively, and taking the monotonicity loss and the smoothness loss as the objective functions, the multi-objective genetic algorithm is used to optimize the fusion weights of the multi-distance metrics to obtain the best weight combination.

[0217] The evaluation indicators include the monotonicity and smoothness of the health degree curve, and the monotonicity loss and smoothness loss are calculated. The health state of the device gradually degrades over time. Assuming the obtained health degree curve is , then:

[0218] Monotonicity loss:

[0219]

[0220] Smoothing loss:

[0221]

[0222] The above is a schematic solution of a health status evaluation device for a front-end isolated switching power supply of an aging platform in this embodiment. It should be noted that the technical solution of the health status evaluation device for the front-end isolated switching power supply of the aging platform belongs to the same concept as the technical solution of the above-mentioned health status evaluation method for the front-end isolated switching power supply of the aging platform. For the details not described in the technical solution of the health status evaluation device for the front-end isolated switching power supply of the aging platform, reference can be made to the description of the technical solution of the above-mentioned health status evaluation method for the front-end isolated switching power supply of the aging platform.

[0223] Figure 5 The structural block diagram of a computing device 500 provided according to an embodiment of this specification is shown. The components of the computing device 500 include, but are not limited to, a memory 510 and a processor 520. The processor 520 is connected to the memory 510 through a bus 530, and a database 550 is used to store data.

[0224] The computing device 500 further includes an access device 540, and the access device 540 enables the computing device 500 to communicate via one or more networks 560. Examples of these networks include the Public Switched Telephone Network (PSTN), Local Area Network (LAN), Wide Area Network (WAN), Personal Area Network (PAN), or a combination of communication networks such as the Internet. The access device 540 may include one or more of any type of wired or wireless network interfaces (for example, Network Interface Card (NIC)), such as IEEE802.11 Wireless Local Area Network (WLAN) wireless interface, Worldwide Interoperability for Microwave Access (Wi-MAX) interface, Ethernet interface, Universal Serial Bus (USB) interface, cellular network interface, Bluetooth interface, Near Field Communication (NFC) interface, and so on.

[0225] In an embodiment of this specification, the above components of the computing device 500 and Figure 5 other components not shown in Figure 5 may also be connected to each other, for example, through a bus. It should be understood that

[0226] The computing device 500 can be any type of stationary or mobile computing device, including mobile computers or mobile computing devices (e.g., tablet computers, personal digital assistants, laptop computers, notebook computers, netbooks, etc.), mobile phones (e.g., smart phones), wearable computing devices (e.g., smart watches, smart glasses, etc.) or other types of mobile devices, or stationary computing devices such as desktop computers or PCs. The computing device 500 can also be a mobile or stationary server.

[0227] Wherein, the processor 520 is used to execute the following computer-executable instructions, and when the computer-executable instructions are executed by the processor, the steps of the above-mentioned health state evaluation method for the isolated switching power supply at the front end of the aging platform are implemented.

[0228] The above is a schematic solution of a computing device according to this embodiment. It should be noted that the technical solution of this computing device and the technical solution of the above-mentioned health state evaluation method for the isolated switching power supply at the front end of the aging platform belong to the same concept. For the details not described in detail in the technical solution of the computing device, reference can be made to the description of the technical solution of the above-mentioned health state evaluation method for the isolated switching power supply at the front end of the aging platform.

[0229] This specification also provides a computer-readable storage medium in one embodiment, which stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, the steps of the above-mentioned health state evaluation method for the isolated switching power supply at the front end of the aging platform are implemented.

[0230] The above is a schematic solution of a computer-readable storage medium according to this embodiment. It should be noted that the technical solution of this storage medium and the technical solution of the above-mentioned health state evaluation method for the isolated switching power supply at the front end of the aging platform belong to the same concept. For the details not described in detail in the technical solution of the storage medium, reference can be made to the description of the technical solution of the above-mentioned health state evaluation method for the isolated switching power supply at the front end of the aging platform.

[0231] This specification also provides a computer program in one embodiment, wherein when the computer program is executed on a computer, the computer is made to execute the steps of the above-mentioned health state evaluation method for the isolated switching power supply at the front end of the aging platform.

[0232] The above is a schematic solution of a computer program according to this embodiment. It should be noted that the technical solution of this computer program and the technical solution of the above-mentioned health state evaluation method for the isolated switching power supply at the front end of the aging platform belong to the same concept. For the details not described in detail in the technical solution of the computer program, reference can be made to the description of the technical solution of the above-mentioned health state evaluation method for the isolated switching power supply at the front end of the aging platform.

[0233] The above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the acts or steps recited in the claims may be performed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the drawings do not necessarily require the particular order or sequential order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0234] The computer instructions include computer program code, which may be in source code form, object code form, executable file, or some intermediate form, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0235] It should be noted that for the foregoing method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the embodiments of this specification are not limited by the described order of actions, because according to the embodiments of this specification, certain steps may be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the embodiments of this specification.

[0236] In the above embodiments, the descriptions of the various embodiments have their own focuses. For the parts not detailed in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0237] The preferred embodiments of this specification disclosed above are only used to help explain this specification. The alternative embodiments do not elaborate on all the details and do not limit the invention to the specific embodiments described. Obviously, many modifications and variations can be made according to the content of the embodiments of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the embodiments of this specification, so that those skilled in the art can understand and utilize this specification well. This specification is only limited by the claims and their full scope and equivalents.

Claims

1. A method for evaluating the health status of a front-end isolated switching power supply of a burn-in station, characterized in that: include: Acquire the monitoring data of the isolated switching power supply at the front end of the aging test bench at the current moment, and obtain feature data with high information density by extracting features from the monitoring data; By inputting the high information density feature data into the trained nearest neighbor component analysis (NCA) algorithm model, a preliminary health status assessment result of the front-end isolated switching power supply of the aging test bench at the current moment is obtained; If the preliminary health status assessment result is a degraded state, the Mahalanobis distance, cosine similarity and Manhattan distance between the high information density feature data and the health baseline data are calculated respectively, and the final health status assessment result of the front-end isolated switching power supply of the aging test bench is obtained by multi-distance metric fusion with automatic weight optimization of the Mahalanobis distance, the cosine similarity and the Manhattan distance, which includes: ； in, It represents the final health status assessment result of the isolated switching power supply at the front end of the aging test bench. represents the Mahalanobis distance; represents cosine similarity; represents Manhattan distance; , and denote weights, and + + =1; Among them, by measuring the monotonicity and smoothness of the health curve in the health state of the current degradation state, the monotonicity loss and the smoothness loss are calculated respectively, and the monotonicity loss and the smoothness loss are used as objective functions. The multi-objective genetic algorithm is used to optimize the fusion weights of multiple distance metrics to obtain the best weight combination.

2. The method according to claim 1, characterized in that Also includes: If the evaluation status result is a healthy status, the healthy status is used as the final healthy status evaluation result of the front-end isolated switching power supply of the aging test bench.

3. The method according to claim 1, characterized in that Also includes: Acquire first historical monitoring data and second historical monitoring data of the full life cycle degradation cycle of the front-end isolated switching power supply of the aging test bench, and use the first historical monitoring data as a training set and the second historical monitoring data as a test set; By performing feature extraction on the historical monitoring data in the training set, first historical feature data with high information density is obtained; Determine the state label corresponding to each moment of the first high information density historical feature data; wherein the state label includes a healthy state label and a degraded state label; An NCA algorithm model is constructed, and the historical feature data with the first high information density is used as the model input. The state label corresponding to each moment of the historical feature data with the first high information density is used as the model output to train and learn the NCA algorithm model, so as to obtain a trained NCA algorithm model and an optimized mapping matrix.

4. The method according to claim 3, characterized in that Also includes: By performing feature extraction on the historical monitoring data in the test set, historical feature data with the second highest information density is obtained; By inputting the historical feature data with the second highest information density into the trained NCA algorithm model, the state label corresponding to each moment of the historical feature data with the second highest information density is obtained; Extracting all healthy state labels from the state labels corresponding to each moment of the historical feature data with the second highest information density, and discarding all degraded state labels; Build health baseline data based on the historical feature data corresponding to all health status tags.

5. The method according to claim 4, characterized in that Calculating the Mahalanobis distance between the high information density feature data and the health baseline data includes: Based on the optimized mapping matrix, the Mahalanobis distance between the feature data with high information density and the healthy baseline data is calculated.

6. A health status assessment device for the front-end isolated switching power supply of a burn-in station, characterized in that: include: The feature extraction module is configured to obtain the monitoring data of the isolated switching power supply at the front end of the aging test bench at the current moment, and obtain feature data with high information density by extracting features from the monitoring data; A preliminary evaluation module is configured to obtain a preliminary health status evaluation result of the front-end isolated switching power supply of the aging test bench at the current moment by inputting the high information density feature data into a trained neighbor component analysis (NCA) algorithm model; The final evaluation module is configured to respectively calculate the Mahalanobis distance, cosine similarity and Manhattan distance between the feature data with high information density and the healthy baseline data if the preliminary health status evaluation result is a degraded state, and obtain the final health status evaluation result of the front-end isolated switching power supply of the aging test bench by multi-distance metric fusion with automatic weight optimization of the Mahalanobis distance, the cosine similarity and the Manhattan distance, which includes: ； in, It represents the final health status assessment result of the isolated switching power supply at the front end of the aging test bench. represents the Mahalanobis distance; represents cosine similarity; represents Manhattan distance; , and denote weights, and + + =1; Among them, by measuring the monotonicity and smoothness of the health curve in the health state of the current degradation state, the monotonicity loss and the smoothness loss are calculated respectively, and the monotonicity loss and the smoothness loss are used as objective functions. The multi-objective genetic algorithm is used to optimize the fusion weights of multiple distance metrics to obtain the best weight combination.

7. A computing device, characterized in that include: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the health status assessment method of the front-end isolated switching power supply of the aging station described in any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium, characterized in that: It stores computer executable instructions, which, when executed by a processor, implement the steps of the health status assessment method of the front-end isolated switching power supply of the aging station as described in any one of claims 1 to 5.

Images