Artificial Intelligence-Based Method and System for Identifying the Status Information of Electronic Components

Abstract

This application discloses an artificial intelligence-based method and system for identifying the state information of electronic components. The method first acquires the physical failure mechanism information of the target electronic component, analyzes the failure process, and determines the state representation logic of each physical failure stage. Then, it establishes an interpretable hierarchical state identification algorithm adapted to the failure process through hierarchical mapping processing. Subsequently, it collects state monitoring data and inputs it into the algorithm to obtain the state feature information corresponding to each physical failure stage. Next, it generates state feature transmission paths according to the logical order of the failure process. Finally, based on the paths, it completes layer-by-layer feature association and causal deduction to generate a state identification result containing the path to the root cause of physical failure. This improves the accuracy and interpretability of state identification, enabling full-chain tracing of the root cause of electronic component failures.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a method and system for identifying the status information of electronic components based on artificial intelligence. Background Technology

[0002] In the field of modern electronic equipment operation and maintenance, the status identification of electronic components is one of the core links to ensure stable equipment operation and avoid sudden failures. As electronic equipment develops towards miniaturization, integration, and high reliability, the failure mechanisms of electronic components are becoming increasingly complex. Affected by the coupled effects of multiple environmental factors such as temperature, humidity, vibration, and electromagnetic interference, their failure process exhibits multi-stage and nonlinear evolutionary characteristics.

[0003] In existing technologies, electronic component condition identification methods are mainly divided into two categories: one is the physical model-based method, which establishes physical models of electronic components such as thermal, mechanical, and electrical systems, and deduces the component condition by combining monitoring data. However, this type of method requires precise physical parameters and a complex model building process, places extremely high demands on prior knowledge of component structure and failure mechanisms, and is difficult to adapt to complex failure scenarios under the coupling of multiple factors. The other is the data-driven method, which collects a large amount of condition monitoring data and uses algorithms such as machine learning and deep learning to build a condition identification model. This type of method depends on the scale and quality of the data and can achieve condition classification to a certain extent. However, most algorithms are "black box" models, which can only output condition identification results and cannot explain the derivation process and root cause of failure. Moreover, the model training and identification logic are not deeply bound to the physical failure mechanism of the component. When the monitoring data contains noise or deviates from the distribution of training samples, the identification accuracy will drop significantly. Furthermore, some existing technologies that attempt to integrate physical mechanisms with data-driven approaches only use physical mechanisms as constraints to assist model training. They fail to build a hierarchical identification logic that is fully adapted to the failure process, making it impossible to achieve layer-by-layer association of state features and causal deduction, and thus difficult to generate state identification results containing complete root cause paths.

[0004] It is evident that existing electronic component status identification technologies generally suffer from the inability to balance identification accuracy and interpretability, and the difficulty in achieving full-chain tracing of the root cause of failure. Summary of the Invention

[0005] This application provides a method and system for identifying the status information of electronic components based on artificial intelligence.

[0006] This application provides, in one aspect, an artificial intelligence-based method for identifying the status information of electronic components, applied to an electronic component status information identification system. The method includes:

[0007] The physical failure mechanism information of the target electronic component is obtained, and the physical failure mechanism information reflects the failure mechanism and inherent law of the target electronic component under different environmental conditions;

[0008] The failure process of the target electronic component is analyzed based on the physical failure mechanism information, and the state representation logic corresponding to each physical failure stage is determined based on the evolution characteristics of each physical failure stage in the failure process.

[0009] Based on the state representation logic corresponding to each physical failure stage, a hierarchical mapping process is performed to obtain state identification logic units at different levels. An interpretable hierarchical state identification algorithm is established through the collaborative association of each level of state identification logic units. The hierarchical structure of the hierarchical state identification algorithm matches the division of the physical failure stages of the failure process.

[0010] The status monitoring data of the target electronic component is collected and input into the hierarchical status recognition algorithm. The status monitoring data is then processed by the corresponding hierarchical status recognition logic unit to obtain the status feature information corresponding to each physical failure stage.

[0011] In the hierarchical state recognition algorithm, the state feature information is transmitted according to the logical order of the failure process, and the transmission path of the state feature information is generated. The transmission path corresponds to the evolution order of the failure process and is associated with the state recognition logic units of each level.

[0012] Based on the transmission path, the state feature information is analyzed layer by layer to obtain the layer-by-layer feature association result. Based on the transmission path and the layer-by-layer feature association result, the causal inference is performed layer by layer to obtain the causal inference result. The state identification result containing the physical failure root cause path is generated by combining the layer-by-layer feature association result and the causal inference result.

[0013] One embodiment of this application provides an electronic component status information identification system, including:

[0014] A processor; a storage device having a computer program stored thereon; a network interface for providing network communication functions; when the computer program is executed by the processor, the processor enables the processor to implement any of the aforementioned artificial intelligence-based electronic component status information recognition methods.

[0015] One embodiment of this application provides a readable storage medium storing a program or instructions, which, when executed by a processor, implements the steps of the artificial intelligence-based electronic component status information recognition method.

[0016] This application, starting from the physical failure essence of target electronic components, constructs a hierarchical state recognition system deeply adapted to the failure process, achieving end-to-end logical continuity from failure mechanism to state recognition results. It creatively solves the core problems of "black box" nature and difficulty in root cause tracing in traditional electronic component state recognition. First, it analyzes the failure process and determines the state representation logic based on physical failure mechanism information, ensuring that the underlying logic of state recognition perfectly aligns with the physical failure patterns of the target electronic components. This overcomes the limitations of traditional methods that rely solely on statistical features of monitoring data, guaranteeing the scientific rigor and accuracy of state recognition from the outset. Second, the hierarchical state recognition algorithm established through hierarchical mapping processing has a hierarchical structure that strictly matches the physical failure stages of the failure process, achieving interpretability of the state recognition logic. This addresses the industry pain point of AI-powered state recognition algorithms struggling to trace the decision-making process, making the derivation path of the state recognition results clearly traceable. Furthermore, the state feature transmission path constructed based on the logical sequence of the failure process provides an accurate logical framework for layer-by-layer feature association and causal deduction, ensuring that state feature analysis and causal deduction strictly follow the evolution law of physical failure. The final state identification results containing the root cause path of physical failure can not only accurately determine the current state of the target electronic component, but also fully present the entire root cause chain from the initial failure initiation to the final state formation. This provides a comprehensive and in-depth decision-making basis for the preventive maintenance, root cause location, and reliability optimization of electronic components, thereby improving the intelligence level of electronic component state identification technology. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart illustrating an artificial intelligence-based method for identifying the status information of electronic components, as provided in an embodiment of this application.

[0019] Figure 2 This is a schematic diagram of the basic structure of an electronic component status information identification system provided in an embodiment of this application.

[0020] Figure 3 This is a functional block diagram of an electronic component status information identification device provided in an embodiment of this application. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0022] Please see Figure 1 , Figure 1 This is a flowchart of an artificial intelligence-based electronic component status information identification method provided in an embodiment of this application. The method can be executed by an electronic component status information identification system, or by the electronic component status information identification system and a server. The method may include steps 110-160.

[0023] In the implementation of the embodiments of this application, those skilled in the art will process state monitoring data from multiple source sensors. This data typically involves different physical quantities and dimensions, such as temperature, voltage, current, and frequency, characterized by units like Celsius, volts, amperes, and hertz, with significant differences in their numerical ranges and unit scales. The technical solution described in the claims involves multiple steps, including feature parsing, correlation calculation, and causal deduction. These steps may implicitly require mathematical operations on parameters with different dimensions. For example, in feature correlation analysis, directly comparing or combining parameter values ​​with different physical meanings may lead to inaccurate physical meaning in the calculation results if dimension unification is not performed, affecting the accuracy of state recognition. However, those skilled in the art, based on professional knowledge and conventional data preprocessing techniques, can overcome such dimensional errors by adaptively normalizing parameters with different physical meanings and dimensions. Normalization aims to convert heterogeneous parameters into dimensionless standardized values. For example, methods such as min-max scaling, Z-score normalization, or decimal scaling can map the original data to a unified scale range, thereby eliminating the interference of dimensional differences on subsequent algorithm calculations. When performing state monitoring data acquisition, feature parsing, and hierarchical correlation analysis in the embodiments of this application, those skilled in the art will naturally introduce a normalization step to ensure that multi-dimensional data are comparable and consistent in feature extraction and logical processing, thereby ensuring the feasibility of the overall technical solution and enabling the hierarchical state recognition algorithm to effectively parse the failure process and generate a reliable root cause path.

[0024] Adaptive normalization of parameters with different physical meanings and dimensions to overcome dimensional errors is a common technique implemented by those skilled in the art based on existing technologies. In data science, machine learning, and signal processing, data preprocessing is a standard procedure, and normalization, as a fundamental technique, is widely used to process multi-source heterogeneous data, with mature documentation in existing technical literature and engineering practice. For example, in electronic component condition monitoring applications, sensor data often have different units and dynamic ranges. Engineers typically clean, align, and normalize the data to ensure the stability of subsequent feature engineering and model training. In artificial intelligence algorithms, such as neural networks, support vector machines, or clustering models, the consistency of input data scale is a prerequisite for model convergence and performance optimization; therefore, normalization is considered a standard practice. Those skilled in the art, when implementing the methods in the claims, have sufficient technical motivation and conventional ability to introduce normalization processing, adjusting parameter scales without inventive effort and avoiding computational logic errors caused by dimensional mismatches. Specifically, in the segmentation processing of state monitoring data, the design of feature parsing rules, and the calculation of correlation, normalization can convert different physical quantities into dimensionless feature values. This allows steps such as feature similarity algorithms and causal deduction logic to be performed based on a fair scale, thereby generating physically meaningful layer-by-layer correlation results and causal deduction results. This not only conforms to data standardization specifications but also enhances the interpretability and robustness of the hierarchical state recognition algorithm.

[0025] Furthermore, those skilled in the art can flexibly select normalization methods based on data type and application scenario to adapt to various parameter processing stages involved in the embodiments of this application. For example, for continuous variables in condition monitoring data, linear normalization can be used to scale them to the [0, 1] interval; for data with skewed distribution, logarithmic transformation or Box-Cox transformation can be used to improve distribution characteristics; in the transmission verification stage, parameters such as information capacity and transmission rate may involve information units such as bits and bytes, but normalization can convert them into relative values ​​to facilitate continuity detection and adaptation conversion evaluation in a simulated transmission environment. In addition, when updating the failure mode library and generating early warning schemes, normalization helps maintain the consistency between historical data and real-time data, making root cause path evolution analysis and impact inference based on a unified benchmark. In summary, through conventional normalization techniques, those skilled in the art can effectively integrate parameters of different dimensions, overcome potential dimension calculation problems in the implementation of the claims, and ensure the logical coherence and physical rationality of the entire condition identification method from data acquisition to result output, thereby meeting the accurate needs of electronic component failure prediction and health management.

[0026] Step 110: Obtain the physical failure mechanism information of the target electronic component. The physical failure mechanism information reflects the failure mechanism and inherent law of the target electronic component under different environmental conditions.

[0027] In this embodiment, the target electronic component is a DC-DC converter chip in an airborne power module. The electronic component status information identification system obtains the physical failure mechanism information of the DC-DC converter chip by connecting to the airborne electronic component failure mechanism database. The physical failure mechanism information covers the failure mechanism and inherent laws of the chip under different environmental conditions such as high and low temperature cycling, vibration and shock, and electromagnetic interference. For example, in the high and low temperature cycling environment, the internal solder joints of the chip accumulate thermal stress due to the difference in thermal expansion coefficients, which leads to fatigue cracking of the solder joints; in the vibration and shock environment, the chip package pins and the substrate connection part suffer mechanical damage due to repeated force; and the inherent laws of various failure mechanisms changing with the duration and intensity of environmental effects. The above information is stored in structured text form, including failure triggering conditions, evolution process descriptions, and key influencing factors.

[0028] Step 120: Analyze the failure process of the target electronic component based on the physical failure mechanism information, and determine the state representation logic corresponding to each physical failure stage based on the evolution characteristics of each physical failure stage in the failure process.

[0029] The electronic component status information identification system, based on the acquired physical failure mechanism information of the DC-DC converter chip, breaks down the failure process through a failure process analysis module, dividing the entire failure process into four physical failure stages: initial damage initiation stage, damage accumulation and development stage, performance degradation and aggravation stage, and functional failure triggering stage. The initial damage initiation stage is characterized by minute damage appearing in the local microstructure within the chip; the corresponding state characterization logic is to capture the weak electrical and thermal signal anomalies caused by microstructural changes. The damage accumulation and development stage is characterized by the gradual expansion of minute damage, forming observable local damage areas; the corresponding state characterization logic is to identify the electrical parameter shifts and uneven thermal distribution caused by damage expansion. The performance degradation and aggravation stage is characterized by the local damage area affecting the overall chip performance; the corresponding state characterization logic is to analyze the downward trend of performance parameters and fluctuations in functional stability. The functional failure triggering stage is characterized by the damage reaching a critical value, resulting in the loss of the chip's core function; the corresponding state characterization logic is to determine whether the core functional parameters exceed the normal threshold range.

[0030] Step 130: Perform hierarchical mapping processing based on the state representation logic corresponding to each physical failure stage to obtain state identification logic units at different levels. Establish an interpretable hierarchical state identification algorithm through the collaborative association of state identification logic units at each level. The hierarchical structure of the hierarchical state identification algorithm matches the division of physical failure stages in the failure process.

[0031] In this embodiment of the application, the electronic component status information identification system performs hierarchical mapping processing on the status representation logic corresponding to each physical failure stage, specifically executing the following sub-steps:

[0032] Step 131: Based on the order of evolution of each physical failure stage, the state representation logic corresponding to each physical failure stage is used as the basic mapping unit and mapped to the logic construction space of different levels in sequence according to the evolution order.

[0033] The electronic component status information identification system uses the evolutionary sequence of the initial damage initiation stage, the damage accumulation and development stage, the performance degradation and aggravation stage, and the functional failure triggering stage as the hierarchical sorting basis. The status representation logic corresponding to each stage is used as four basic mapping units, which are sequentially mapped to a four-layer logic construction space distributed from bottom to top. The first layer corresponds to the initial damage initiation stage, the second layer corresponds to the damage accumulation and development stage, the third layer corresponds to the performance degradation and aggravation stage, and the fourth layer corresponds to the functional failure triggering stage.

[0034] Step 132: Logically decompose the basic mapping unit of each level, extract the core representation elements and element association rules in the basic mapping unit, encapsulate the core representation elements and element association rules into independent state recognition logic sub-units, and combine multiple state recognition logic sub-units to form the state recognition logic unit of the corresponding level.

[0035] Furthermore, the electronic component status information identification system logically decomposes the basic mapping unit at each level. Taking the basic mapping unit at the first level, the initial damage initiation stage, as an example, it extracts two core characterizing elements: minute fluctuations in electrical signals and local anomalies in thermal signals, as well as the element association rule that "minute fluctuations in electrical signals and local anomalies in thermal signals are synchronously correlated." These core characterizing elements and the element association rule are encapsulated into two status identification logic sub-units, responsible for identifying minute fluctuations in electrical signals and local anomalies in thermal signals, respectively. The two sub-units are connected through the element association rule, combining to form the status identification logic unit at the first level. Similarly, the basic mapping units at the other three levels are logically decomposed, elements are extracted, and encapsulated to form the status identification logic units at the corresponding levels. Each level's status identification logic unit contains 2 to 3 status identification logic sub-units, which work collaboratively through element association rules.

[0036] Step 133: Based on the evolutionary correlation between physical failure stages, establish a collaborative correlation channel between adjacent level state recognition logic units. The collaborative correlation channel is pre-set with logical transmission rules corresponding to the evolutionary correlation, so that the output of the previous level state recognition logic unit is adapted to the input of the next level state recognition logic unit according to the logical transmission rules.

[0037] The electronic component status information identification system establishes collaborative correlation channels between adjacent level status identification logic units based on the evolutionary correlation between physical failure stages. For example, there is an evolutionary correlation between the initial damage initiation stage and the damage accumulation and development stage, where "initial damage expansion triggers the formation of local damage areas." Therefore, a collaborative correlation channel is established between the first-level and second-level status identification logic units. Within this channel, a pre-defined logical transmission rule states that "the microscopic damage features output by the previous level must be converted into feature forms suitable for identifying local damage areas in the next level." Specifically, the small fluctuations in electrical signals and local anomalies in thermal signals identified by the first level are converted into damage expansion rate and damage area-related feature forms that can be identified by the second level through a feature conversion algorithm. This ensures that the output of the previous level's status identification logic unit can be adapted to the input of the next level's status identification logic unit. Similarly, collaborative correlation channels are established between the second and third levels, and between the third and fourth levels, respectively, with corresponding pre-defined logical transmission rules to achieve collaborative correlation between adjacent levels.

[0038] Step 134: Perform overall logical integration of all level state recognition logic units and collaborative association channels to obtain the initial level state recognition algorithm.

[0039] The electronic component status information identification system integrates four-layer status identification logic units and three adjacent collaborative association channels into a whole logic, clarifies the workflow of each status identification logic subunit, the information transmission sequence and rule calling method of the collaborative association channels, and forms an initial hierarchical status identification algorithm. The hierarchical structure of this algorithm is completely matched with the four physical failure stages of the failure process. Each level corresponds to one physical failure stage, and the collaborative association channels between levels correspond to the evolutionary association relationship between physical failure stages.

[0040] Step 135: By comparing the coverage of each level of logical unit with the evolutionary feature range of the corresponding physical failure stage, calculate the matching degree between the hierarchical structure of the initial hierarchical state recognition algorithm and the physical failure stage division of the failure process. Adjust the transmission rules of the core representation elements of the logical unit and the collaborative association channel according to the matching degree until the matching degree reaches the set matching value, and complete the establishment of an interpretable hierarchical state recognition algorithm.

[0041] In detail, the electronic component condition information identification system uses a hierarchical matching degree evaluation module to compare the coverage of each level of logic unit with the evolutionary feature range of the corresponding physical failure stage, calculating the matching degree between the hierarchical structure of the initial hierarchical condition identification algorithm and the physical failure stage division of the failure process. For example, the evaluation found that the first-level condition identification logic unit only covered the initial damage initiation features related to electrical and thermal signals, but did not cover the microstructural damage features caused by mechanical stress, resulting in the matching degree not reaching the set matching value. To address this, the system adjusts the core representation elements of the first-level condition identification logic unit, adds mechanical stress-related features, and adjusts the transmission rules of the collaborative correlation channel to include the newly added mechanical stress features in the feature transformation range, recalculating the matching degree. This adjustment and calculation process is repeated until the matching degree reaches the set matching value, completing the establishment of an interpretable hierarchical condition identification algorithm. This hierarchical condition identification algorithm can clearly display the condition identification logic of each physical failure stage and the correlation between levels, possessing good interpretability and facilitating subsequent traceability and analysis of condition identification results.

[0042] Step 140: Collect the status monitoring data of the target electronic component and input the status monitoring data into the hierarchical status recognition algorithm. The status monitoring data is then processed by the status recognition logic unit at the corresponding level to obtain the status feature information corresponding to each physical failure stage.

[0043] In this embodiment of the application, the electronic component status information identification system performs status monitoring data acquisition and feature parsing processing on the DC-DC converter chip, specifically executing the following sub-steps:

[0044] Step 141: Continuously collect multi-dimensional status monitoring data of the target electronic components during operation, and organize the multi-dimensional status monitoring data according to the collection time sequence to generate a time-series status monitoring dataset.

[0045] The electronic component status information identification system continuously collects multi-dimensional status monitoring data of the DC-DC converter chip during operation through various types of sensors deployed on and around the chip. This data includes chip output voltage, output current, surface temperature, package vibration acceleration, and ambient electromagnetic radiation intensity. The system organizes the collected multi-dimensional status monitoring data according to the acquisition time sequence, associating various data collected at the same time point to form a time-series status monitoring dataset indexed by timestamps. This dataset contains all dimensions of monitoring data for each timestamp.

[0046] Step 142: The time-series state monitoring dataset is segmented according to the time distribution characteristics of the physical failure stage to obtain segmented state monitoring data corresponding to each physical failure stage. The time range of the segmented state monitoring data is adapted to the evolution time range of the corresponding physical failure stage.

[0047] The electronic component condition information identification system segments the time-series condition monitoring dataset based on the temporal distribution characteristics of each physical failure stage in the failure process of a DC-DC converter chip. For example, the initial damage initiation stage corresponds to the initial period of chip operation, the damage accumulation and development stage corresponds to a period of time after the initial damage initiation, the performance degradation aggravation stage corresponds to the period after the damage accumulation and development, and the functional failure triggering stage corresponds to the period from performance degradation aggravation to functional failure. According to these temporal distribution characteristics, the system divides the time-series condition monitoring dataset into four segments. The time range of each segment matches the evolution time range of the corresponding physical failure stage, ensuring that each segment covers the state change process of its corresponding stage.

[0048] Step 143: Input the segmented state monitoring data into the corresponding level of the state recognition logic unit in the hierarchical state recognition algorithm. The corresponding level of the state recognition logic unit calls the preset feature parsing rules to extract core information and transform features from the segmented state monitoring data.

[0049] The electronic component condition information identification system inputs segmented condition monitoring data corresponding to the initial damage initiation stage into the first-level condition identification logic unit of the hierarchical condition identification algorithm, inputs segmented condition monitoring data corresponding to the damage accumulation and development stage into the second-level condition identification logic unit, and so on. Each level of condition identification logic unit invokes preset feature parsing rules to extract core information and transform features from the input segmented condition monitoring data. The feature parsing rules of the first-level condition identification logic unit extract core information related to the initial damage initiation, such as minor voltage fluctuations, localized temperature anomalies, and sudden changes in vibration acceleration, and transform this information into feature forms that conform to the initial damage initiation state characterization logic. The feature parsing rules of the second-level condition identification logic unit extract core information related to the damage accumulation and development, such as voltage offsets, temperature distribution unevenness, and vibration frequency changes, and transform them into corresponding feature forms.

[0050] Step 144: Extract the core information related to the evolution characteristics of the corresponding physical failure stage from the segmented state monitoring data, transform the core information into a feature form that conforms to the state representation logic, and generate initial state feature information.

[0051] The electronic component condition information identification system extracts core information related to the evolution characteristics of the corresponding physical failure stage from segmented condition monitoring data through core information extraction modules of each level of condition identification logic unit. For example, the first layer extracts core information such as the minute fluctuation amplitude of the voltage signal, the range of local abnormal areas of the temperature signal, and the instantaneous change value of vibration acceleration from the segmented condition monitoring data of the initial damage initiation stage. The second layer extracts core information such as the degree of voltage parameter deviation, the non-uniformity coefficient of temperature distribution, and the trend of vibration frequency change from the segmented condition monitoring data of the damage accumulation and development stage. The system transforms the extracted core information into feature forms that conform to the condition characterization logic of the corresponding physical failure stage. For example, the minute voltage fluctuation amplitude is transformed into a dimensionless fluctuation feature vector, and the range of local abnormal temperature areas is transformed into a regional feature matrix to generate initial state feature information. Each initial state feature information is a multi-dimensional feature set containing multiple related feature parameters.

[0052] Step 145: Based on the evolution characteristics of the physical failure stage, perform logical verification on the initial state feature information. If the verification fails, adjust the feature parsing rules and re-process the feature transformation until the verification passes, thereby obtaining the state feature information corresponding to each physical failure stage.

[0053] The electronic component status information identification system performs logical verification on initial state feature information based on the evolution characteristics of each physical failure stage. The verification includes checking whether the initial state feature information conforms to the evolutionary pattern of the corresponding stage, whether the feature dimensions are complete, and whether there are contradictions between feature parameters. For example, if the initial state feature information for the initial damage initiation stage is found to contain a large voltage shift characteristic that only appears in the damage accumulation and development stage, the verification fails. In response, the system adjusts the feature parsing rules of the first-layer status identification logic unit, adds feature filtering conditions, excludes information that does not conform to the evolutionary characteristics of the initial damage initiation stage, re-extracts core information and performs feature transformation processing, generates new initial state feature information, and performs logical verification again. This adjustment and verification process is repeated until the initial state feature information for all physical failure stages passes verification, resulting in state feature information corresponding to each physical failure stage. This state feature information accurately reflects the state changes of the corresponding physical failure stage.

[0054] Step 150: In the hierarchical state recognition algorithm, the state feature information is transmitted according to the logical order of the failure process, and a transmission path for the state feature information is generated. The transmission path corresponds to the evolution order of the failure process and is associated with the state recognition logic units of each level.

[0055] The electronic component status information identification system transmits status feature information according to the logical order of the failure process in the hierarchical status identification algorithm, generates a transmission path, and specifically executes the following sub-steps:

[0056] Step 151: Based on the evolution order of the failure process, determine the basic transmission direction of the state characteristic information. The basic transmission direction is consistent with the evolution order of the physical failure stages.

[0057] The electronic component status information identification system uses the evolution sequence of the failure process of DC-DC converter chips, namely the initial damage initiation stage, the damage accumulation and development stage, the performance degradation aggravation stage, and the functional failure triggering stage, as a benchmark to determine the basic transmission direction of status feature information as transmission from the first-level status identification logic unit to the fourth-level status identification logic unit. The basic transmission direction is consistent with the evolution sequence of the physical failure stages, ensuring that the transmission of status feature information follows the natural evolution law of the failure process.

[0058] Step 152: Determine the transmission adaptation requirements of different types of state feature information based on the type and attributes of the state feature information output by each level of state recognition logic unit; wherein, the state feature information output by the previous level meets the input requirements of the next level.

[0059] The electronic component status information identification system analyzes the type and attributes of the status feature information output by each level of status identification logic unit. For example, the first level outputs feature vectors related to micro-damage, the second level outputs feature matrices related to local damage areas, the third level outputs feature sets related to performance degradation, and the fourth level outputs feature parameters related to functional failure. Based on these types and attributes, the system determines the transmission adaptation requirements for different types of status feature information. For example, micro-damage feature vectors need to be converted into a feature form acceptable to the feature matrix of local damage areas, and the feature matrix of local damage areas needs to be converted into a feature form acceptable to the feature set of performance degradation, ensuring that the status feature information output by the previous level meets the input requirements of the next level.

[0060] Step 153: Based on the transmission adaptation requirements of basic transmission direction and state feature information, a transmission node for state feature information is constructed between adjacent level state recognition logic units. The transmission node has the functions of information reception, adaptation conversion and transmission.

[0061] The electronic component status information identification system, based on the basic transmission direction and status feature information transmission adaptation requirements, constructs transmission nodes for status feature information between adjacent level status identification logic units. Three transmission nodes are constructed between the first and second layers, the second and third layers, and the third and fourth layers of status identification logic units, respectively. Each transmission node has information receiving, adaptation and conversion, and sending functions. The information receiving function receives status feature information output from the previous level status identification logic unit; the adaptation and conversion function performs format conversion and feature mapping on the received status feature information according to transmission adaptation requirements; and the sending function sends the converted status feature information to the next level status identification logic unit.

[0062] Step 154: Connect the corresponding transmission nodes of each level in sequence to obtain the initial transmission path; wherein, the initial transmission path covers the state recognition logic units of all levels, and the path direction follows the evolution order of the failure process.

[0063] The electronic component status information identification system connects three transmission nodes sequentially, forming an initial transmission path that starts from the first-level status identification logic unit, passes through the first transmission node to the second level, then through the second transmission node to the third level, and finally through the third transmission node to the fourth-level status identification logic unit. This initial transmission path covers all levels of status identification logic units, and the path direction follows the evolution sequence of the failure process, ensuring that status characteristic information can be transmitted between levels in the order of the failure process.

[0064] Step 155: By simulating the transmission process of the state feature information along the initial transmission path, the transmission path is verified, and the adaptation and conversion effect and transmission continuity of the state feature information at the transmission node are detected to obtain the detection result. The conversion rules of the transmission node are adjusted according to the detection result until there are no breaks in the transmission of the state feature information in the initial transmission path.

[0065] The electronic component status information identification system uses a transmission path verification module to simulate the transmission process of status feature information along the initial transmission path and verify the transmission of the initial path. The system first determines the basic parameters for simulated transmission, including the transmission rate threshold and information capacity adaptation range. These basic parameters are consistent with the transmission adaptation requirements of the status feature information. Based on these basic parameters, a simulated transmission environment is constructed. In this environment, the hierarchical structure of the initial transmission path and the initial configuration of each transmission node are replicated. The status feature information is then input in batches into the starting transmission node of the initial transmission path according to a preset transmission order.

[0066] During the simulated transmission process, input and output data of state feature information at each transmission node are collected in real time. The adaptation and conversion effect of the transmission node is evaluated by calculating the feature matching degree between the input and output data. The feature matching degree is calculated by a preset feature similarity algorithm and must meet a preset matching threshold. At the same time, the transmission time sequence of state feature information between adjacent transmission nodes is tracked, the transmission interval time is recorded and compared with a preset time sequence threshold, and the continuity of state feature information transmission is judged by the time sequence comparison result.

[0067] The evaluation results of the adaptation and conversion effects of each transmission node, along with the consistency judgment results obtained from time-series comparison, are integrated to generate a detection result that includes both the adaptation and conversion pass status and the transmission consistency status. If the detection result shows any cases of adaptation and conversion failure or transmission inconsistency, the conversion rule parameters of the corresponding transmission node are extracted. The reasons for parameter deviations are analyzed in conjunction with the type attributes of the status feature information and the transmission adaptation requirements. For example, the feature mapping rules of the transmission node may not cover certain feature dimensions, leading to an unqualified adaptation and conversion effect; or the information transmission timing settings of the transmission node may be unreasonable, leading to transmission inconsistency.

[0068] Based on the reasons for the deviation, the data conversion algorithm and timing coordination parameters in the conversion rules are adjusted. After the adjustment is completed, the transmission verification process of state feature information is re-executed in the simulated transmission environment until the adaptation conversion effect of all transmission nodes meets the preset matching threshold and the transmission interval between adjacent transmission nodes meets the preset timing threshold, so that there are no breaks in the transmission of state feature information in the initial transmission path.

[0069] Step 156: Mark the physical failure stage evolution correlation of each transmission node in the initial transmission path so that the initial transmission path defines the evolution order of the corresponding failure process. Combine the marking results with the state recognition logic units of each level to complete the generation of the transmission path.

[0070] The electronic component status information identification system marks the evolutionary relationships of physical failure stages corresponding to each transmission node in the initial transmission path. For example, it marks the evolutionary relationship from the initial damage initiation stage to the damage accumulation and development stage for the first transmission node, from the damage accumulation and development stage to the performance degradation aggravation stage for the second transmission node, and from the performance degradation aggravation stage to the functional failure triggering stage for the third transmission node. Through this marking, the initial transmission path can clearly define the evolutionary sequence of the corresponding failure process. The system combines the marking results with the status identification logic units at each level, clarifying the preceding and following level status identification logic units connected to each transmission node and the corresponding physical failure stage evolutionary relationships, thus completing the generation of the transmission path. This transmission path not only enables the transmission of status feature information according to the logical sequence of the failure process but also visually displays the evolutionary relationships between each physical failure stage.

[0071] Step 160: Perform layer-by-layer correlation analysis on the state feature information based on the transmission path to obtain layer-by-layer feature correlation results; perform layer-by-layer causal inference based on the transmission path and the layer-by-layer feature correlation results to obtain layer-by-layer causal inference results; and combine the layer-by-layer feature correlation results and the layer-by-layer causal inference results to generate a state identification result containing the physical failure root cause path.

[0072] The electronic component status information identification system performs layer-by-layer correlation analysis and causal deduction of status feature information based on the transmission path and generates status identification results. Specifically, it executes the following sub-steps:

[0073] Step 161: Taking the transmission order of the transmission path as the analysis order, starting from the state feature information of the initial level, extract the state feature information corresponding to the two adjacent levels in sequence, and determine the correlation dimension between the state feature information of the adjacent levels.

[0074] The electronic component status information identification system analyzes data according to the transmission order of the transmission path. Starting with the status feature information corresponding to the first layer, it sequentially extracts the status feature information corresponding to adjacent layers. Specifically, it first extracts the status feature information between the first and second layers, then the second and third layers, and finally the third and fourth layers. For each pair of adjacent layers, the system determines the correlation dimension between the status feature information through a correlation dimension analysis module. For example, the correlation dimension between the first and second layers is damage propagation correlation, meaning the expansion of initial damage features triggers the formation of local damage area features; the correlation dimension between the second and third layers is performance impact correlation, meaning the development of local damage area features affects changes in performance degradation features; and the correlation dimension between the third and fourth layers is function triggering correlation, meaning the aggravation of performance degradation features triggers the appearance of functional failure features.

[0075] Step 162: Construct a feature association analysis model based on the association dimension. The feature association analysis model has preset association judgment rules corresponding to the association dimension, and the state feature information of adjacent levels is input into the feature association analysis model.

[0076] The electronic component status information identification system constructs a feature association analysis model based on defined association dimensions. Each association dimension corresponds to a feature association analysis model. For example, for the damage propagation association dimension, a damage propagation feature association analysis model is constructed. The model pre-sets association judgment rules corresponding to damage propagation, such as whether the changing trend of the initial damage features is consistent with the development trend of the local damage area features, and whether there is a corresponding relationship between the intensity of the initial damage features and the scale of the local damage area features. The system inputs the status feature information of adjacent levels into the corresponding feature association analysis model; for example, the status feature information of the first layer and the second layer are input into the damage propagation feature association analysis model.

[0077] Step 163: Calculate the correlation degree of state feature information of adjacent levels through feature correlation analysis model, select feature correlation combinations that meet the preset requirements, and generate single-level feature correlation results.

[0078] The electronic component status information identification system calculates the correlation degree of status feature information at adjacent levels through a feature association analysis model. This correlation degree calculation is based on preset association judgment rules, analyzing the correspondence and consistency of change trends between status feature information. For example, in the damage propagation feature association analysis model, the system calculates the correlation degree between the initial damage feature vector at the first level and the feature matrix of the local damage region at the second level, filtering out feature association combinations that meet preset requirements. If the correlation degree between a certain initial damage feature and its corresponding local damage region feature reaches a preset threshold, then that combination is included in the single-level feature association result. The system completes the feature association analysis of all adjacent levels in this manner, generating single-level feature association results. Each single-level feature association result contains multiple sets of compliant feature association combinations.

[0079] Step 164: Perform feature association analysis on all adjacent levels in sequence according to the transmission path, and integrate the feature association results of each single level in hierarchical order to obtain the layer-by-layer feature association results.

[0080] The electronic component status information identification system sequentially performs feature association analysis between the first and second layers, the second and third layers, and the third and fourth layers according to the transmission path, obtaining three single-layer feature association results. The system integrates these single-layer feature association results in hierarchical order, using the feature association results between the first and second layers as the first part, the second and third layers as the second part, and the third and fourth layers as the third part, forming a layer-by-layer feature association result. This result clearly demonstrates the correlation between status feature information at each adjacent layer, providing a foundation for subsequent causal inference.

[0081] Step 165: Based on the failure process evolution order corresponding to the transmission path and combined with the layer-by-layer feature association results, determine the causal association direction between adjacent level state feature information. The causal association direction is consistent with the evolution direction of the failure process.

[0082] The electronic component status information identification system determines the causal relationship between adjacent levels of status feature information based on the failure process evolution sequence corresponding to the transmission path and the results of layer-by-layer feature association. Since the failure process is a unidirectional evolution from initial damage initiation to functional failure triggering, the causal relationship direction is consistent with the evolution direction of the failure process. That is, the status feature information of the previous level is the cause, and the information of the next level is the effect. For example, the initial damage feature of the first level is the cause of the local damage area feature of the second level, the local damage area feature of the second level is the cause of the performance degradation feature of the third level, and so on. This determination of the causal relationship direction conforms to the physical laws of the failure process, ensuring the rationality of subsequent causal deductions.

[0083] Step 166: Based on the causal correlation direction and the failure occurrence mechanism in the physical failure mechanism information, generate causal deduction logic and input the layer-by-layer feature correlation results into the causal deduction logic.

[0084] The electronic component status information identification system generates causal deduction logic based on a determined causal relationship direction and the failure mechanism in the physical failure mechanism information. This causal deduction logic encompasses rules for determining the causal relationships between adjacent levels of status feature information. For example, regarding the causal relationship between initial damage features and local damage area features, based on the mechanism of thermal stress accumulation leading to damage propagation in the physical failure mechanism information, the causal deduction rule is that there is a causal relationship between the amount of thermal stress accumulation in the initial damage feature and the scale of local damage area expansion; the greater the amount of thermal stress accumulation, the larger the scale of local damage area expansion. The system inputs the layer-by-layer feature association results into the causal deduction logic, providing specific feature association data for causal deduction.

[0085] Step 167: Analyze the causal relationship between adjacent level feature association combinations through causal deduction logic, identify the influence of the previous level state feature information on the next level state feature information, and generate a single-level causal deduction result.

[0086] The electronic component status information identification system analyzes the causal relationships between adjacent level feature associations using causal reasoning logic, specifically executing the following sub-steps:

[0087] Step 1671: Based on the failure occurrence mechanism in the physical failure mechanism information, extract the evolutionary correlation criteria between adjacent physical failure stages, and integrate the evolutionary correlation criteria into the core judgment rules of the causal deduction logic.

[0088] The electronic component status information identification system, based on the failure mechanism information of DC-DC converter chips' physical failure mechanism, extracts evolutionary correlation criteria between adjacent physical failure stages. For example, the evolutionary correlation criterion from the initial damage initiation stage to the damage accumulation and development stage is "the initial damage gradually expands under environmental influence, forming a local damage area"; the evolutionary correlation criterion from the damage accumulation and development stage to the performance degradation aggravation stage is "the development of the local damage area affects the normal operation of the chip's internal circuitry, leading to performance parameter degradation"; and the evolutionary correlation criterion from the performance degradation aggravation stage to the functional failure trigger stage is "performance degradation reaches a critical value, triggering the failure of the chip's core function." The system integrates these evolutionary correlation criteria into the core judgment rules of causal deduction logic, providing a basis for causal relationship analysis.

[0089] Step 1672: Decompose the adjacent level feature association combination into a set of state feature information of the previous level and a set of state feature information of the next level. Match the attribute parameters of each state feature information in the two sets one by one through the causal inference logic, and filter out feature information pairs with potential correlation of attribute parameters.

[0090] The electronic component status information recognition system decomposes the feature association combination of adjacent levels into a set of status feature information for the previous level and a set of status feature information for the next level. For example, the feature association combination of the first and second levels is decomposed into a set of initial damage feature information and a set of local damage area feature information. The system matches the attribute parameters of each status feature information in the two sets one by one through causal reasoning logic. For example, it matches the thermal stress parameter in the initial damage feature information with the damage expansion parameter in the local damage area feature information, and filters out feature information pairs with potentially correlated attribute parameters, such as the combination of thermal stress accumulation feature and damage area expansion scale feature.

[0091] Step 1673: Analyze the timing of the action of the previous level state feature information and the next level state feature information in the feature information pair according to the evolutionary correlation criterion, and determine the target feature information pair whose action timing conforms to the evolution order of the failure process.

[0092] The electronic component condition information identification system analyzes the timing of the effects of preceding and subsequent level condition features in a feature information pair based on evolutionary correlation criteria. For example, it analyzes the timing of changes in thermal stress accumulation characteristics and damage area expansion scale characteristics to determine whether they conform to the evolutionary correlation criterion of "initial damage expansion triggers the formation of local damage areas," i.e., whether the change in thermal stress accumulation characteristics precedes the change in damage area expansion scale characteristics, and whether the trends are consistent. The system selects feature information pairs whose timing of action conforms to the failure process evolution sequence and uses them as target feature information pairs.

[0093] Step 1674: For each pair of target feature information, combined with the inherent laws in the physical failure mechanism information, explore the way in which the core elements of the previous level state feature information affect the core elements of the next level state feature information, and determine the intermediate transmission elements and parameter change laws involved in the process.

[0094] The electronic component condition information identification system, for each target feature pair, combines the inherent laws in physical failure mechanism information to uncover the interaction between the core elements of the previous level of condition feature information and the core elements of the subsequent level of condition feature information. For example, for the target feature pair of thermal stress accumulation and damage area expansion scale, combining the inherent law of thermal stress accumulation leading to solder joint fatigue cracking in physical failure mechanism information, the system identifies the core elements of thermal stress accumulation as the number of thermal stress cycles and stress amplitude. Its interaction with the core element of damage area expansion scale, the damage expansion length, occurs through thermal stress cycles inducing fatigue in the solder joint material, leading to gradual crack propagation. The system determines that the intermediate transmission element involved in the process is the fatigue damage accumulation of the solder joint material, and the parameter change pattern is that as the number of thermal stress cycles increases, the fatigue damage accumulation increases, and the damage expansion length increases accordingly.

[0095] Step 1675: Based on the mode of action, intermediate transmission elements and parameter change patterns, generate a description of the influence of the previous level state feature information on the next level state feature information. The description of the influence is used to define the intensity of the interaction and the constraints between the previous level state feature information and the next level state feature information.

[0096] The electronic component status information identification system generates a description of the influence of previous-level status feature information on subsequent-level status feature information based on the determined action mode, intermediate transmission elements, and parameter change laws. For example, for the above target feature information pair, the generated influence description is: "The number of thermal stress cycles and stress amplitude of the thermal stress accumulation feature affect the damage propagation length of the damage area expansion feature by inducing fatigue damage accumulation of the solder joint material. The intensity of the effect is positively correlated with the number of thermal stress cycles and stress amplitude. The constraint condition is the fatigue limit of the solder joint material. When the thermal stress accumulation has not reached the fatigue limit, the damage propagation length grows slowly. After reaching the fatigue limit, the damage propagation length grows rapidly." This influence description clearly defines the intensity of the effect and the constraint condition between the previous-level and subsequent-level status feature information, providing a basis for the quantitative analysis of causal relationships.

[0097] Step 1676: Summarize the causal relationship analysis results and corresponding impact descriptions of all target feature information pairs, and organize the summarized results in a structured manner according to the association order of state feature information pairs to generate a single-layer causal derivation result containing causal relationship objects, action sequence, impact descriptions and constraints.

[0098] The electronic component status information identification system summarizes the causal relationship analysis results and corresponding impact descriptions of all target feature information pairs. The summarized results are then structured according to the association order of the status feature information pairs to generate single-level causal deduction results. Each single-level causal deduction result includes the causal relationship objects (i.e., the corresponding preceding and following level status feature information); the action sequence (i.e., the time sequence in which the preceding level status feature information acts on the following level); the impact description (i.e., the specific mode and intensity of the preceding level's influence on the following level); and constraints (i.e., the limiting factors of the impact). The system generates single-level causal deduction results for all adjacent levels in this manner.

[0099] Step 168: Integrate all single-layer causal derivation results to obtain layer-by-layer causal derivation results.

[0100] The electronic component status information identification system integrates all generated single-layer causal derivation results in hierarchical order, taking the first and second layer causal derivation results as the first part, the second and third layer causal derivation results as the second part, and the third and fourth layer causal derivation results as the third part, thus obtaining layer-by-layer causal derivation results. These results can clearly show the causal relationship and action law between the status feature information of each adjacent level, providing detailed causal basis for subsequent physical failure root cause path analysis.

[0101] Step 169: The layer-by-layer feature association results and the layer-by-layer causal derivation results are fused together to extract the feature association and causal relationship information related to the physical failure root cause from the fused results.

[0102] The electronic component status information recognition system fuses layer-by-layer feature association results with layer-by-layer causal derivation results. Through an information fusion module, feature association combinations are linked with corresponding causal relationship information to form a fusion result. The system extracts feature associations and causal relationship information related to the root causes of physical failure from the fusion result. For example, it extracts feature associations and causal relationships related to thermal stress accumulation in the initial damage initiation stage. This information reflects the initial triggering factors and evolution process of physical failure.

[0103] Step 1610: Determine the evolution path of the physical failure root cause based on the extracted feature association and causal relationship information, and combine the evolution path with the layer-by-layer feature association results and the layer-by-layer causal deduction results to generate a state identification result containing the physical failure root cause path.

[0104] The electronic component status information identification system determines the evolution path of physical failure root causes based on extracted feature correlation and causal relationship information, specifically by executing the following sub-steps:

[0105] Step 16101: Based on the failure process evolution order corresponding to the transmission path, the extracted feature association and causal relationship information are hierarchically classified to obtain feature association subsets and causal relationship subsets corresponding to each physical failure stage.

[0106] The electronic component status information identification system classifies the extracted feature associations and causal relationships based on the failure process evolution sequence corresponding to the transmission path. It categorizes the feature associations and causal relationships related to the initial damage initiation stage into subsets of initial damage feature associations and subsets of initial damage causal relationships; those related to the damage accumulation and development stage into subsets of damage accumulation feature associations and subsets of damage accumulation causal relationships; those related to the performance degradation aggravation stage into subsets of performance degradation feature associations and subsets of performance degradation causal relationships; and those related to the functional failure triggering stage into subsets of functional failure feature associations and subsets of functional failure causal relationships.

[0107] Step 16102: From the feature association subset and causal relationship subset corresponding to the initial physical failure stage, select the core feature association and key causal relationship related to the initial failure signs, and use the core feature association and key causal relationship as the starting point for tracing the root cause of physical failure.

[0108] The electronic component status information identification system filters out the core feature associations and key causal relationships related to the initial failure signs from the subset of initial damage feature associations and the subset of initial damage causal relationships. For example, it filters out the association between thermal stress accumulation features and initial micro-damage features, as well as the causal relationship between thermal stress accumulation and initial micro-damage. The above core feature associations and key causal relationships are used as the starting point for tracing the root cause of physical failure, that is, the initial triggering factors of physical failure.

[0109] Step 16103: Based on the evolution order of the failure process, sequentially associate the feature association subsets and causal relationship subsets of adjacent physical failure stages, track the transmission trajectory of core feature associations between each level, and analyze the continuation logic of key causal relationships between adjacent levels.

[0110] The electronic component condition information identification system, based on the evolution sequence of the failure process, sequentially associates feature subsets and causal relationship subsets of adjacent physical failure stages. Specifically, it first associates the initial damage feature subset with the damage accumulation feature subset, then the damage accumulation feature subset with the performance degradation feature subset, and finally the performance degradation feature subset with the functional failure feature subset. The system tracks the transmission trajectory of core feature associations between each level. For example, it tracks how the initial microscopic damage features caused by thermal stress accumulation expand into local damage area features, then affect performance degradation features, and finally trigger functional failure features. Simultaneously, it analyzes the continuation logic of key causal relationships between adjacent levels. For instance, it analyzes how the causal relationship of thermal stress accumulation causing initial damage continues into the causal relationship of initial damage expansion causing local damage, then into the causal relationship of local damage development causing performance degradation, and finally into the causal relationship of performance degradation aggravating functional failure.

[0111] Step 16104: Based on the transmission trajectory of core features and the continuation logic of key causal relationships between adjacent levels, create an association chain from the starting tracing point to the final failure stage. Mark the key feature nodes and causal links that trigger the failure evolution at each stage in the complete association chain to obtain the evolution path of the physical failure root cause.

[0112] The electronic component condition information identification system creates a chain of connections from the initial tracing point to the final failure stage based on the transmission trajectory of core feature associations and the continuation logic of key causal relationships between adjacent levels. Starting with the core feature associations and key causal relationships of the initial damage initiation stage, the chain sequentially connects the relevant feature associations and causal relationships of the damage accumulation and development stage, the performance degradation aggravation stage, and the functional failure triggering stage, forming a complete chain. Within this complete chain, the system marks the key feature nodes and causal links that trigger the failure evolution at each stage. For example, the feature node of thermal stress accumulation in the initial damage initiation stage and its corresponding causal links, and the feature node of local damage area expansion in the damage accumulation and development stage and its corresponding causal links, etc., thus obtaining the evolution path of the physical failure root cause. This evolution path clearly demonstrates the complete process and key nodes of physical failure from initial triggering to final occurrence.

[0113] Step 16105: Compare and calibrate the evolution path of the physical failure root cause with the layer-by-layer feature association results so that each feature node in the evolution path corresponds to the feature association combination in the layer-by-layer feature association results.

[0114] The electronic component condition information identification system compares and calibrates the evolution path of physical failure root causes with the layer-by-layer feature association results, checking whether each feature node in the evolution path corresponds to the feature association combination in the layer-by-layer feature association results. For example, if a feature node of a local damage area in the evolution path is found to have a corresponding feature association combination in the layer-by-layer feature association results, but the association description is inconsistent, the system adjusts the association description of that node in the evolution path to make it consistent with the content in the layer-by-layer feature association results. This comparison and calibration process is repeated until every feature node in the evolution path corresponds to the feature association combination in the layer-by-layer feature association results, ensuring the accuracy of the evolution path.

[0115] Step 16106: Match and optimize the causal links in the evolutionary path with the single-layer causal derivation results in the layer-by-layer causal derivation results.

[0116] The electronic component status information identification system matches and optimizes the causal interactions in the evolution path with the single-layer causal derivations in the layer-by-layer causal derivation results. It checks whether each causal interaction in the evolution path is consistent with the single-layer causal derivation result. For example, if a discrepancy is found between the description of the impact of a certain causal interaction in the evolution path and the content in the layer-by-layer causal derivation result, the system adjusts the description of that causal interaction in the evolution path to match the description of the impact, constraints, and other content in the single-layer causal derivation result. Through matching and optimization, the system ensures a complete match between the causal interactions in the evolution path and the layer-by-layer causal derivation results.

[0117] Step 16107: Combining the initial conditions of the physical failure root cause, the key triggering factors in the evolution process, the characteristic changes at each stage, and the formation logic of the final failure state, the evolution path of the calibrated and verified physical failure root cause, the layer-by-layer feature association results, and the layer-by-layer causal deduction results after matching optimization are integrated to generate a state identification result containing the physical failure root cause path.

[0118] The electronic component condition information identification system combines the initial conditions of physical failure root causes (i.e., initial triggering factors); key triggering factors in the evolution process, such as thermal stress accumulation reaching the fatigue limit or the expansion of local damage areas to a critical scale; characteristic changes at each stage, such as changes from initial damage characteristics to local damage area characteristics and then to performance degradation characteristics; and the formation logic of the final failure state. It integrates the evolution path of the calibrated and verified physical failure root causes, the layer-by-layer feature association results, and the matched and optimized layer-by-layer causal deduction results to generate a condition identification result containing the physical failure root cause path. The condition identification result not only includes the current failure state of the target electronic component but also the complete physical failure root cause path from initial failure triggering to final failure, as well as characteristic association and causal relationship information at each stage.

[0119] In an optional embodiment, the method further includes:

[0120] Step 210, after generating the state identification result containing the physical failure root cause path by combining the layer-by-layer feature association result and the layer-by-layer causal inference result, further includes: based on the state identification result containing the physical failure root cause path, associating the operation association link information of the target electronic component, and logically mapping the physical failure root cause path in the state identification result with the operation association link information to obtain the set of association link nodes corresponding to the root cause path.

[0121] After generating status identification results containing the physical failure root cause paths, the electronic component status information identification system first associates the operational link information of the DC-DC converter chip. This operational link information covers the connection relationships, signal transmission paths, and functional cooperation relationships between the chip and other components in the airborne power module, such as input filter capacitors, output inductors, and control chips. The system then logically maps the physical failure root cause paths in the status identification results to the operational link information. For example, it maps thermal stress accumulation nodes in the physical failure root cause path to the heat dissipation system link of the power module, and local damage area expansion nodes to the connection link between the chip and the substrate, obtaining a set of associated link nodes corresponding to the root cause path. This set contains all operational link nodes related to the physical failure root cause.

[0122] Step 220: Using the state identification result as input, start the root cause influence deduction logic, and deduce the influence transmission logic of each root cause node on the corresponding node in the set of associated link nodes in sequence according to the evolution sequence of the physical failure root cause path.

[0123] The electronic component status information identification system takes the status identification results containing the root cause paths of physical failures as input and initiates the root cause impact deduction logic. This logic pre-defines the impact rules of different types of failure root causes on operational related link nodes. For example, the impact rule of thermal stress accumulation on heat dissipation system nodes is that thermal stress accumulation leads to an increase in chip temperature, which in turn affects the heat dissipation load of the heat dissipation system. The system deduces the impact transmission logic of each root cause node on the corresponding node in the set of related link nodes according to the evolution sequence of the physical failure root cause paths, that is, from the initial damage initiation stage to the functional failure triggering stage. For example, it first deduces the impact transmission logic of the thermal stress accumulation root cause node on the heat dissipation system related link nodes, that is, thermal stress accumulation causes the chip temperature to rise, resulting in an increase in the speed of the heat dissipation fan and the temperature of the heat sink fins, etc.; then it deduces the impact transmission logic of the local damage area expansion root cause node on the chip-substrate connection link node, that is, the expansion of the local damage area leads to an increase in the connection resistance between the chip and the substrate, affecting the stability of signal transmission, etc.

[0124] Step 230: Integrate the layer-by-layer causal deduction results from the influence transmission logic and state identification results to generate time-series representation data of the root cause influence range; construct a dynamic evolution model of the root cause influence based on the time-series representation data, and output the diffusion trend information of the root cause influence through the dynamic evolution model.

[0125] The electronic component status information recognition system integrates the derived influence transmission logic with the layer-by-layer causal derivation results from the status recognition results. It associates the influence process in the influence transmission logic with the causal relationships in the layer-by-layer causal derivation results, generating time-series representation data of the root cause influence range. Based on a time axis, the time-series representation data displays the range and degree of influence of each root cause node on associated link nodes within different time periods. The system constructs a dynamic evolution model of root cause influence based on the time-series representation data. This model employs a time-series convolutional network structure, taking the time-series representation data as input and outputting the diffusion trend information of the root cause influence. By learning the changing patterns in the time-series representation data, the model predicts the diffusion trend of root cause influence in the operational associated links over a future period. For example, it predicts how the effects of accumulated thermal stress will further diffuse to other components of the power module, leading to more performance degradation or functional failure.

[0126] Step 240: The diffusion trend information is fused with the physical failure root cause path in the state identification result to generate a root cause impact assessment report. The root cause impact assessment report is used to characterize the overall impact and time series distribution of the physical failure root cause on the target electronic component and related links.

[0127] The electronic component status information identification system integrates the diffusion trend information output by the dynamic evolution model with the physical failure root cause path from the status identification results. It overlays the diffusion trend information onto the physical failure root cause path, displaying the future diffusion direction and scope of the root cause impact. Based on the fused information, the system generates a root cause impact assessment report. This report includes detailed information on the physical failure root cause path, the impact transmission logic of each root cause node on related link nodes, and a prediction of the diffusion trend of the root cause impact. It characterizes the overall impact and time-series distribution of physical failure root causes on the target electronic component and related links, providing crucial decision-making support for maintenance planning and operational strategy adjustments for airborne power modules.

[0128] In an optional embodiment, the method further includes:

[0129] Step 310, after generating a state identification result containing the physical failure root cause path by combining the layer-by-layer feature association result and the layer-by-layer causal deduction result, further includes: extracting the core logical association information of the physical failure root cause path, layer-by-layer feature association result and layer-by-layer causal deduction result from the state identification result; inputting the core logical association information into a preset failure mode library, and determining the failure mode category corresponding to the core logical association information according to the classification logic of the failure mode library.

[0130] After generating a status identification result containing the physical failure root cause path, the electronic component status information identification system extracts the core logical association information from the status identification result, including the physical failure root cause path, layer-by-layer feature association results, and layer-by-layer causal deduction results. This core logical association information covers the initial triggering factors of the physical failure, key relationships in the evolution process, causal interaction laws, and the formation logic of the final failure state. The system inputs this core logical association information into a preset failure mode library, which contains various electronic component failure mode categories, such as thermal fatigue failure, mechanical damage failure, and electromagnetic interference failure. Each failure mode category corresponds to specific failure characteristics and evolution laws. The system determines the failure mode category corresponding to the core logical association information according to the classification logic of the failure mode library—that is, based on the failure triggering factors, evolution process, and failure characteristics. For example, it determines that the failure mode category corresponding to the current failure of the DC-DC converter chip is thermal fatigue failure.

[0131] Step 320: Based on the core logic association information, adjust the feature association weights and causal deduction logic under the corresponding failure mode category to ensure that the internal logic of the failure mode library is compatible with the core rules of the state identification results.

[0132] The electronic component condition information identification system adjusts the feature association weights and causal deduction logic under the corresponding failure mode category, namely the thermal fatigue failure category, based on the extracted core logical association information. For example, if it finds that the correlation between the thermal stress accumulation feature and the initial damage feature in the core logical association information is higher than the preset weight in the failure mode library, the system adjusts the weight value of this feature association to make it consistent with the actual correlation. At the same time, according to the causal action law in the core logical association information, the system adjusts the causal deduction logic under the thermal fatigue failure category, such as updating the specific mechanism description of thermal stress accumulation causing initial damage, so that the internal logic of the failure mode library and the core law of the condition identification results are kept consistent, thereby improving the accuracy and practicality of the failure mode library.

[0133] Step 330: Extract unique root cause paths and associated features not covered by the failure mode library from the state identification results, and construct new failure mode logical units; integrate the new failure mode logical units into the failure mode library, and update the classification logic and retrieval rules of the failure mode library; output the updated failure mode library and adaptation description, wherein the adaptation description is used to characterize the adaptation and association relationship between the updated failure mode library and the current state identification results.

[0134] The electronic component condition information identification system extracts unique root cause paths and associated features from the condition identification results that are not covered by the failure mode library. For example, in the failure process of the DC-DC converter chip, there is a unique root cause path in which thermal stress and electromagnetic interference jointly cause initial damage, and this path is not covered by any category in the existing failure mode library. The system constructs new failure mode logic units based on this unique information. These new failure mode logic units include descriptions of unique root cause paths, definitions of associated features, and causal deduction rules. The system integrates the new failure mode logic units into the failure mode library, updates the classification logic of the failure mode library, adds new failure mode categories, and adjusts the retrieval rules to ensure accurate retrieval of new failure modes. Finally, the system outputs the updated failure mode library and adaptation descriptions. The adaptation descriptions characterize the compatibility and association between the updated failure mode library and the current condition identification results, for example, explaining that the newly added failure mode logic units correspond to the unique root cause paths in the current condition identification results.

[0135] In an optional embodiment, the method further includes:

[0136] Step 410, after generating the state identification result containing the physical failure root cause path by combining the layer-by-layer feature association result and the layer-by-layer causal inference result, further includes: taking the state identification result containing the physical failure root cause path as the core input, analyzing the hierarchical node distribution and causal association strength between nodes of the physical failure root cause path in the state identification result, and obtaining the root cause path node set and node association strength matrix.

[0137] After generating a status identification result containing the root cause path of physical failure, the electronic component status information identification system uses this status identification result as the core input. Through a path parsing module, it analyzes the hierarchical node distribution and the causal correlation strength between nodes in the physical failure root cause path, obtaining a root cause path node set and a node correlation strength matrix. The root cause path node set contains all key feature nodes and causal link nodes in the physical failure root cause path; the node correlation strength matrix displays the causal correlation strength between each node in matrix form. The element values ​​in the matrix represent the degree of correlation between corresponding two nodes, with higher values ​​indicating stronger correlation.

[0138] Step 420: Based on the root cause path node set and node association strength matrix, map the failure warning triggering conditions corresponding to electronic components, and take the node combination in the node association strength matrix that exceeds the preset association threshold as the warning core triggering unit to obtain the node-level warning triggering logic.

[0139] The electronic component status information identification system maps failure warning triggering conditions for DC-DC converter chips based on root cause path node sets and node association strength matrices. The system pre-sets association thresholds and uses node combinations in the node association strength matrix that exceed these thresholds as core warning triggering units. For example, a combination of a thermal stress accumulation node and an initial damage node, where the association strength between the two exceeds a preset threshold, is used as a core warning triggering unit. For each core warning triggering unit, the system defines corresponding warning triggering conditions. For instance, when the thermal stress accumulation characteristic parameter reaches a certain level and its association strength with the initial damage characteristic exceeds a preset threshold, a warning is triggered, resulting in node-level warning triggering logic.

[0140] Step 430: The node-level early warning triggering logic is hierarchically integrated according to the evolution order of the physical failure root cause path to construct a full-link early warning logic chain from the initial failure node to the final failure state. The full-link early warning logic chain includes the early warning triggering sequence and associated response relationship of each level node.

[0141] The electronic component status information identification system hierarchically integrates node-level early warning triggering logic according to the evolution sequence of the physical failure root cause path. It takes the node-level early warning triggering logic of the initial damage initiation stage as the starting part of the chain, and sequentially integrates the node-level early warning triggering logic of the damage accumulation development stage, the performance degradation aggravation stage, and the functional failure triggering stage, constructing a full-link early warning logic chain from the initial failure node to the final failure state. The full-link early warning logic chain includes the early warning triggering sequence of each level of node, i.e., the order in which different nodes trigger early warnings, as well as the associated response relationships, i.e., the response rules of subsequent nodes after a node triggers an early warning, such as whether to trigger early warnings of subsequent nodes in conjunction with the early warning of a certain node, and how to adjust the early warning thresholds of subsequent nodes.

[0142] Step 440: Establish an adaptation association between the full-link early warning logic chain and the real-time status monitoring data transmission link of electronic components, so that the real-time status monitoring data can be synchronously input into the full-link early warning logic chain according to the preset time sequence; perform layer-by-layer node matching and association strength verification on the input real-time status monitoring data through the full-link early warning logic chain to generate real-time early warning judgment results.

[0143] The electronic component status information identification system establishes an adaptive association between the end-to-end early warning logic chain and the real-time status monitoring data transmission link of the DC-DC converter chip, configuring a data transmission interface to enable real-time status monitoring data to be synchronously input into the end-to-end early warning logic chain according to a preset time sequence. The system performs layer-by-layer node matching and association strength verification on the input real-time status monitoring data through the end-to-end early warning logic chain. First, it matches the real-time status monitoring data with the characteristics of the initial failed node to verify whether it meets the node-level early warning triggering conditions; if so, it triggers the initial early warning. Then, it matches the real-time status monitoring data with the characteristics of the next level of nodes to verify whether the association strength reaches a preset threshold. This process is repeated for all levels of nodes to complete the matching and verification, generating a real-time early warning judgment result. The real-time early warning judgment result includes whether an early warning has been triggered, the location of the node that triggered the warning, and the severity of the warning.

[0144] Step 450: Based on the real-time early warning judgment results, output a failure early warning scheme that includes the early warning level, root cause node location, and preliminary intervention direction. The early warning level of the failure early warning scheme is positively correlated with the correlation strength value in the node correlation strength matrix.

[0145] The electronic component status information identification system outputs a failure warning scheme based on real-time early warning judgment results, including the warning level, root cause node location, and preliminary intervention direction. The warning level is determined by the correlation strength value in the node correlation strength matrix; the higher the correlation strength value, the higher the warning level, showing a positive correlation. For example, a high correlation strength value corresponds to a Level 1 warning, indicating an extremely high failure risk; a medium correlation strength value corresponds to a Level 2 warning, indicating a relatively high failure risk. Root cause node location identifies the specific node that triggers the warning, such as a thermal stress accumulation node in the initial damage initiation stage. The preliminary intervention direction is given based on the type of root cause node and the failure mechanism. For example, for a warning triggered by a thermal stress accumulation node, the preliminary intervention direction would be to optimize the heat dissipation system or reduce the chip operating temperature. The failure warning scheme provides maintenance personnel with timely warning information and preliminary handling suggestions, helping to take preventative measures to avoid functional failure of target electronic components and ensure the stable operation of airborne power modules.

[0146] Based on the above, this application takes an airborne (28V to 5V) DC-DC converter chip (model: LMxxxx-xx) as the target object. By connecting to the failure mechanism database of avionics components, the specific failure mechanism file of this chip model is retrieved to determine its core failure mechanism in the airborne environment: In the high and low temperature cycling environment (-55℃~125℃), the internal solder joints of the chip accumulate thermal stress due to the difference in thermal expansion coefficients between the ceramic substrate and the solder (ceramic substrate α≈4ppm / ℃, solder α≈22ppm / ℃). When the number of cycles exceeds 500, fatigue cracking of the solder joint is likely to occur; In the vibration and shock environment (10~2000Hz, acceleration 15g), mechanical damage is likely to occur at the connection between the package pins and the substrate; Electromagnetic interference (200V / m, 800MHz~2GHz) will cause distortion of internal circuit signals. The above information is stored in structured text and includes core contents such as failure triggering conditions, evolution process and key influencing factors.

[0147] Based on the failure mechanism information analysis, the failure process is divided into four physical failure stages and corresponding state characterization logic is determined: the initial damage initiation stage (0~200 high and low temperature cycles), which requires capturing weak electrical signals (small fluctuations in output voltage ripple ±5mV) and thermal signals (local temperature anomalies ±0.3℃) caused by microstructural changes; the damage accumulation and development stage (200~400 high and low temperature cycles), which focuses on identifying electrical parameter deviations (output voltage deviation ±10mV) and uneven heat distribution (temperature difference >0.8℃); the performance degradation intensification stage (400~500 high and low temperature cycles), which analyzes the decline in performance parameters (conversion efficiency reduction >3%) and functional stability fluctuations (output voltage fluctuation >15mV); and the functional failure triggering stage (>500 high and low temperature cycles), which determines the core functional parameters (output voltage deviation from the 5V±5% threshold). Subsequently, a hierarchical state recognition algorithm was constructed through hierarchical mapping. The state representation logic of the four stages was mapped into four layers of logical units according to the evolution order. The first layer (initial damage) encapsulates two sub-units: identification of small fluctuations in electrical signals and identification of local anomalies in thermal signals. They work together through the "signal synchronization association" rule. The remaining layers are similarly decomposed and encapsulated. A collaborative association channel is established between adjacent layers. The feature conversion rule is preset. The algorithm is constructed by verifying the matching degree (setting the matching value ≥ 0.92, and finally adjusting the matching degree to 0.95).

[0148] During the condition monitoring data acquisition and feature analysis process, 1000 hours of operational data (sampling frequency 100Hz) were continuously collected using miniature thermocouples, vibration sensors, and voltage monitoring modules deployed on the chip surface, and compiled into a time-series dataset. The data was segmented according to the time distribution characteristics of the failure stages: 0~250 hours (corresponding to initial damage), 250~500 hours (damage accumulation), 500~750 hours (performance degradation), and 750~1000 hours (functional failure). The data for each segment was input into the corresponding logic unit. Through feature analysis and logic verification, the core state characteristics of each stage were extracted: In the initial damage stage, the output voltage ripple exhibits a small fluctuation of 3mV, and the temperature in the chip's central region is 0.2℃ higher than the edge; in the damage accumulation stage, the output voltage deviates by 8mV, and the temperature distribution unevenness reaches 0.9℃; in the performance degradation stage, the conversion efficiency decreases from 94% to 90.5%, and the output voltage fluctuates by 16mV; in the functional failure stage, the output voltage drops to 4.6V (exceeding the 5V±5% threshold).

[0149] Transmit state characteristic information according to the logical sequence of the failure process, construct the transmission path and complete transmission verification (transmission breakpoint is 0), and mark the evolutionary correlation relationship corresponding to each transmission node. Based on the transmission path, conduct layer-by-layer correlation analysis and causal deduction, determine the correlation dimensions of adjacent level features (damage propagation correlation, performance impact correlation, function triggering correlation), calculate the correlation degree through the feature correlation analysis model (preset correlation degree threshold ≥ 0.85), screen out effective correlation combinations, and generate layer-by-layer feature correlation results. Combine the failure mechanism to generate causal deduction logic, determine the correlation direction where the previous level feature is the cause and the next level is the effect, for example, "in the initial damage stage, 3mV voltage ripple fluctuation and 0.2℃ local temperature rise, through thermal stress accumulation, trigger 8mV voltage shift in the damage accumulation stage", generating layer-by-layer causal deduction results including causal correlation objects and action sequence. Finally, the two types of results are merged, root cause-related information is extracted, and the initial thermal stress accumulation is identified as the root cause of failure. The root cause evolution path is constructed as follows: thermal stress accumulation (initial damage) - solder joint damage propagation (damage accumulation) - conversion efficiency decline (performance degradation) - output voltage exceeding the standard (functional failure). Finally, a state identification result containing this root cause path is generated, which determines that the target chip is currently in the stage of accelerated performance degradation and has about H hours of operating life before functional failure. The root cause of failure is thermal stress accumulation caused by high and low temperature cycling.

[0150] This application, starting from the physical failure essence of target electronic components, constructs a hierarchical state recognition system deeply adapted to the failure process, achieving end-to-end logical continuity from failure mechanism to state recognition results. It creatively solves the core problems of "black box" nature and difficulty in root cause tracing in traditional electronic component state recognition. First, it analyzes the failure process and determines the state representation logic based on physical failure mechanism information, ensuring that the underlying logic of state recognition perfectly aligns with the physical failure patterns of the target electronic components. This overcomes the limitations of traditional methods that rely solely on statistical features of monitoring data, guaranteeing the scientific rigor and accuracy of state recognition from the outset. Second, the hierarchical state recognition algorithm established through hierarchical mapping processing has a hierarchical structure that strictly matches the physical failure stages of the failure process, achieving interpretability of the state recognition logic. This addresses the industry pain point of AI-powered state recognition algorithms struggling to trace the decision-making process, making the derivation path of the state recognition results clearly traceable. Furthermore, the state feature transmission path constructed based on the logical sequence of the failure process provides an accurate logical framework for layer-by-layer feature association and causal deduction, ensuring that state feature analysis and causal deduction strictly follow the evolution law of physical failure. The final state identification results containing the root cause path of physical failure can not only accurately determine the current state of the target electronic component, but also fully present the entire root cause chain from the initial failure initiation to the final state formation. This provides a comprehensive and in-depth decision-making basis for the preventive maintenance, root cause location, and reliability optimization of electronic components, thereby improving the intelligence level of electronic component state identification technology.

[0151] Please see Figure 2 The figure is a schematic diagram of the basic structure of an electronic component status information identification system 200 provided in an embodiment of this application. The electronic component status information identification system 200 includes: a processor 201; a storage device 202 on which a computer program 2020 is stored; and a network interface 203 for providing network communication functions. When the computer program 2020 is executed by the processor 201, the processor 201 implements any of the aforementioned artificial intelligence-based electronic component status information identification methods.

[0152] Please see Figure 3 This application provides a functional block diagram of an electronic component status information identification device. The electronic component status information identification device includes:

[0153] The mechanism information acquisition module is used to acquire the physical failure mechanism information of the target electronic component. The physical failure mechanism information reflects the failure mechanism and inherent law of the target electronic component under different environmental conditions.

[0154] The state characterization determination module is used to analyze the failure process of the target electronic component based on the physical failure mechanism information, and determine the state characterization logic corresponding to each physical failure stage based on the evolution characteristics of each physical failure stage in the failure process.

[0155] The hierarchical algorithm establishment module is used to perform hierarchical mapping processing based on the state representation logic corresponding to each physical failure stage to obtain state identification logic units at different levels. Through the collaborative association of each level of state identification logic units, an interpretable hierarchical state identification algorithm is established. The hierarchical structure of the hierarchical state identification algorithm matches the division of the physical failure stages of the failure process.

[0156] The feature parsing and processing module is used to collect the status monitoring data of the target electronic component and input the status monitoring data into the hierarchical status recognition algorithm. The status monitoring data is then processed by the corresponding hierarchical status recognition logic unit to obtain the status feature information corresponding to each physical failure stage.

[0157] The transmission path generation module is used to transmit the state feature information in the logical order of the failure process in the hierarchical state recognition algorithm, and generate the transmission path of the state feature information. The transmission path corresponds to the evolution order of the failure process and is associated with the state recognition logic units of each level.

[0158] The state recognition processing module is used to perform layer-by-layer correlation analysis on the state feature information based on the transmission path to obtain layer-by-layer feature correlation results, perform layer-by-layer causal inference based on the transmission path and the layer-by-layer feature correlation results to obtain layer-by-layer causal inference results, and combine the layer-by-layer feature correlation results and the layer-by-layer causal inference results to generate a state recognition result containing the physical failure root cause path.

[0159] Based on the above, a readable storage medium is provided, on which a program or instructions are stored, and when the program or instructions are executed by a processor, the steps of the above method are implemented.

[0160] Furthermore, it should be noted that this application also provides a computer program product, which may include a computer program that can be stored in a computer-readable storage medium. The processor of the electronic component status information identification system reads the computer program from the computer-readable storage medium, and the processor can execute the computer program, causing the electronic component status information identification system to perform the aforementioned... Figure 1 The methods described in the corresponding embodiments are already known, and therefore will not be repeated here. Furthermore, the beneficial effects of using the same method will also not be repeated. For technical details not disclosed in the computer program product embodiments related to this application, please refer to the description of the method embodiments of this application.

[0161] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems or apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and relevant parts can be referred to the method section.

Claims

1. A method for identifying the status information of electronic components based on artificial intelligence, characterized in that, The method includes: The physical failure mechanism information of the target electronic component is obtained, and the physical failure mechanism information reflects the failure mechanism and inherent law of the target electronic component under different environmental conditions; The failure process of the target electronic component is analyzed based on the physical failure mechanism information, and the state representation logic corresponding to each physical failure stage is determined based on the evolution characteristics of each physical failure stage in the failure process. Based on the state representation logic corresponding to each physical failure stage, a hierarchical mapping process is performed to obtain state identification logic units at different levels. An interpretable hierarchical state identification algorithm is established through the collaborative association of each level of state identification logic units. The hierarchical structure of the hierarchical state identification algorithm matches the division of the physical failure stages of the failure process. The status monitoring data of the target electronic component is collected and input into the hierarchical status recognition algorithm. The status monitoring data is then processed by the corresponding hierarchical status recognition logic unit to obtain the status feature information corresponding to each physical failure stage. In the hierarchical state recognition algorithm, the state feature information is transmitted according to the logical order of the failure process, and the transmission path of the state feature information is generated. The transmission path corresponds to the evolution order of the failure process and is associated with the state recognition logic units of each level. Based on the transmission path, the state feature information is analyzed layer by layer to obtain the layer-by-layer feature association result. Based on the transmission path and the layer-by-layer feature association result, the causal inference is performed layer by layer to obtain the causal inference result. The state identification result containing the physical failure root cause path is generated by combining the layer-by-layer feature association result and the causal inference result.

2. The method as described in claim 1, characterized in that, The hierarchical mapping process based on the state representation logic corresponding to each physical failure stage yields state identification logic units at different levels. An interpretable hierarchical state identification algorithm is then established through the collaborative association of these state identification logic units, including: Based on the order of evolution of each physical failure stage, the state representation logic corresponding to each physical failure stage is used as the basic mapping unit and mapped to the logic construction space of different levels in the order of evolution. The basic mapping unit of each level is logically decomposed, and the core representation elements and element association rules in the basic mapping unit are extracted. The core representation elements and element association rules are encapsulated into independent state recognition logic sub-units. Multiple state recognition logic sub-units are combined to form the state recognition logic unit of the corresponding level. Based on the evolutionary correlation between physical failure stages, a collaborative correlation channel is established between adjacent level state recognition logic units. The collaborative correlation channel is preset with logical transmission rules corresponding to the evolutionary correlation, so that the output of the previous level state recognition logic unit is adapted to the input of the next level state recognition logic unit according to the logical transmission rules. The initial hierarchical state recognition algorithm is obtained by integrating all levels of state recognition logic units and collaborative association channels. By comparing the coverage of each level of logical unit with the evolutionary feature range of the corresponding physical failure stage, the matching degree between the hierarchical structure of the initial hierarchical state recognition algorithm and the physical failure stage division of the failure process is calculated. Based on the matching degree, the transmission rules of the core representation elements of the logical unit and the collaborative association channel are adjusted until the matching degree reaches the set matching value, thus completing the establishment of an interpretable hierarchical state recognition algorithm.

3. The method as described in claim 1, characterized in that, The process involves collecting state monitoring data from the target electronic component and inputting the state monitoring data into the hierarchical state recognition algorithm. The state recognition logic unit at the corresponding level then performs feature parsing processing on the state monitoring data to obtain state feature information corresponding to each physical failure stage, including: Continuously collect multi-dimensional status monitoring data of the target electronic components during operation, and organize the multi-dimensional status monitoring data according to the collection time sequence to generate a time-series status monitoring dataset. The time-series state monitoring dataset is segmented according to the time distribution characteristics of the physical failure stage to obtain segmented state monitoring data corresponding to each physical failure stage. The time range of the segmented state monitoring data is adapted to the evolution time range of the corresponding physical failure stage. Each segment of the state monitoring data is input into the corresponding level of the state recognition logic unit in the hierarchical state recognition algorithm. The corresponding level of the state recognition logic unit calls the preset feature parsing rules to extract core information and transform features from the segmented state monitoring data. Extract core information related to the evolution characteristics of the corresponding physical failure stage from the segmented state monitoring data, transform the core information into a feature form that conforms to the state representation logic, and generate initial state feature information; Based on the evolution characteristics of the physical failure stage, the initial state feature information is logically verified. If the verification fails, the feature parsing rules are adjusted and the feature transformation process is repeated until the verification passes, thereby obtaining the state feature information corresponding to each physical failure stage.

4. The method as described in claim 1, characterized in that, The state feature information is transmitted in the hierarchical state recognition algorithm according to the logical order of the failure process, generating a transmission path for the state feature information. This transmission path corresponds to the evolution order of the failure process and is associated with the state recognition logic units at each level, including: Based on the evolution sequence of the failure process, the basic transmission direction of the state characteristic information is determined, and the basic transmission direction is consistent with the evolution sequence of the physical failure stages. Based on the type and attributes of the state feature information output by each level of state recognition logic unit, the transmission adaptation requirements of different types of state feature information are determined; wherein, the state feature information output by the previous level meets the input requirements of the next level. Based on the transmission adaptation requirements of basic transmission direction and state feature information, a transmission node for state feature information is constructed between state recognition logic units at adjacent levels. The transmission node has the functions of information reception, adaptation conversion and transmission. The initial transmission path is obtained by sequentially connecting the corresponding transmission nodes of each level; wherein, the initial transmission path covers the state recognition logic units of all levels, and the path direction follows the evolution order of the failure process; By simulating the transmission process of the state feature information along the initial transmission path, the transmission path is verified, and the adaptation and conversion effect and transmission continuity of the state feature information at the transmission node are detected to obtain the detection result. Based on the detection result, the conversion rules of the transmission node are adjusted until there are no breaks in the transmission of the state feature information in the initial transmission path. The physical failure stage evolution correlation of each transmission node in the initial transmission path is marked so that the initial transmission path defines the evolution order of the corresponding failure process. The state recognition logic unit of each level is associated with the marking results to complete the generation of the transmission path.

5. The method as described in claim 4, characterized in that, The process of simulating the transmission of the state feature information along the initial transmission path, verifying the transmission of the initial transmission path, detecting the adaptation and conversion effect and transmission continuity of the state feature information at the transmission nodes to obtain detection results, and adjusting the conversion rules of the transmission nodes based on the detection results until there are no breakpoints in the transmission of the state feature information in the initial transmission path, includes: Based on the type and attributes of the state feature information, the basic parameters of the simulated transmission are determined. The basic parameters include the transmission rate threshold and the information capacity adaptation range. The basic parameters are consistent with the transmission adaptation requirements of the state feature information. A simulated transmission environment is constructed based on the basic parameters. The hierarchical structure of the initial transmission path and the initial configuration of each transmission node are replicated in the simulated transmission environment. The status feature information is then input into the starting transmission node of the initial transmission path in batches according to a preset transmission order. During the simulated transmission process, input and output data of status feature information at each transmission node are collected in real time. The adaptation and conversion effect of the transmission node is evaluated by calculating the feature matching degree between the input and output data. The feature matching degree is calculated by a preset feature similarity algorithm and must meet a preset matching threshold. The transmission timing of the state feature information between adjacent transmission nodes is tracked, the transmission interval time is recorded and compared with a preset timing threshold, and the continuity of the state feature information transmission is judged by the timing comparison result. The adaptation and conversion effect evaluation results of each transmission node and the consistency judgment results obtained by time sequence comparison are integrated to generate a detection result that includes the adaptation and conversion qualified status and the transmission consistency status. If the detection results show that the adaptation conversion is unqualified or the transmission is not continuous, the conversion rule parameters of the corresponding transmission node are extracted. The reasons for the parameter deviation are analyzed in combination with the type attributes of the state feature information and the transmission adaptation requirements. Based on the reasons for the deviation, the data conversion algorithm and timing coordination parameters in the conversion rules are adjusted. After the adjustment is completed, the transmission verification process of the state feature information is re-executed in the simulated transmission environment until the adaptation and conversion effects of all transmission nodes meet the preset matching threshold and the transmission interval between adjacent transmission nodes meets the preset timing threshold, so that there are no breaks in the transmission of the state feature information in the initial transmission path.

6. The method as described in claim 1, characterized in that, The process involves performing a layer-by-layer correlation analysis on the state feature information based on the transmission path to obtain layer-by-layer feature correlation results, performing layer-by-layer causal inference based on the transmission path and the layer-by-layer feature correlation results to obtain layer-by-layer causal inference results, and combining the layer-by-layer feature correlation results and the layer-by-layer causal inference results to generate a state identification result containing the physical failure root cause path, including: Taking the transmission order of the transmission path as the analysis order, starting from the state feature information of the initial level, the state feature information corresponding to the two adjacent levels is extracted in sequence to determine the correlation dimension between the state feature information of adjacent levels. A feature association analysis model is constructed based on the association dimension. The feature association analysis model has preset association judgment rules corresponding to the association dimension and inputs the state feature information of adjacent levels into the feature association analysis model. The correlation degree of state feature information of adjacent levels is calculated by using a feature correlation analysis model, and feature correlation combinations that meet the preset requirements are selected to generate single-level feature correlation results. The feature association analysis of all adjacent levels is completed sequentially according to the transmission path. The feature association results of each single level are then integrated in hierarchical order to obtain the layer-by-layer feature association results. Based on the failure process evolution order corresponding to the transmission path, and combined with the layer-by-layer feature association results, the causal association direction between adjacent level state feature information is determined, and the causal association direction is consistent with the evolution direction of the failure process. Based on the failure occurrence mechanism in the causal correlation direction and physical failure mechanism information, a causal deduction logic is generated, and the layer-by-layer feature correlation results are input into the causal deduction logic; By analyzing the causal relationships between adjacent level feature combinations through causal deduction logic, the influence of the previous level state feature information on the next level state feature information is identified, generating single-level causal deduction results; and by integrating all single-level causal deduction results, layer-by-layer causal deduction results are obtained. The layer-by-layer feature association results and the layer-by-layer causal derivation results are fused together to extract the feature association and causal relationship information related to the physical failure root cause from the fused results. Based on the extracted feature association and causal relationship information, the evolution path of the physical failure root cause is determined. The evolution path is combined with the layer-by-layer feature association results and the layer-by-layer causal deduction results to generate a state recognition result containing the physical failure root cause path.

7. The method as described in claim 6, characterized in that, The process involves analyzing the causal relationships between adjacent level feature combinations using causal deduction logic, identifying the influence of previous level state feature information on subsequent level state feature information, and generating single-level causal deduction results, including: Based on the failure occurrence mechanism in the physical failure mechanism information, the evolutionary correlation criteria between adjacent physical failure stages are extracted, and the evolutionary correlation criteria are integrated into the core judgment rules of the causal deduction logic. The adjacent-level feature association combination is decomposed into a set of state feature information of the previous level and a set of state feature information of the next level. The attribute parameters of each state feature information in the two sets are matched one by one through the causal inference logic to filter out feature information pairs with potential correlation of attribute parameters. Based on the evolutionary correlation criteria, the timing of the action of the previous level state feature information and the next level state feature information in the feature information pair is analyzed to determine the target feature information pair whose timing of action conforms to the evolution order of the failure process; For each pair of target feature information, combined with the inherent laws in the physical failure mechanism information, we explore the way in which the core elements of the previous level state feature information affect the core elements of the next level state feature information, and determine the intermediate transmission elements and parameter change laws involved in the process. Based on the mode of action, intermediate transmission elements and parameter change rules, a description of the influence of the previous level state feature information on the next level state feature information is generated. The influence description is used to define the intensity of the interaction and the constraints between the previous level state feature information and the next level state feature information. The causal relationship analysis results and corresponding impact descriptions of all target feature information pairs are summarized. The summarized results are then structured according to the association order of state feature information pairs to generate a single-layer causal derivation result containing causal relationship objects, action sequence, impact descriptions, and constraints.

8. The method as described in claim 6, characterized in that, The evolution path of the physical failure root cause is determined based on the extracted feature association and causal relationship information. This evolution path is then combined with layer-by-layer feature association results and layer-by-layer causal derivation results to generate a state identification result containing the physical failure root cause path, including: Based on the failure process evolution order corresponding to the transmission path, the extracted feature association and causal relationship information are hierarchically classified to obtain feature association subsets and causal relationship subsets corresponding to each physical failure stage. From the feature association subset and causal relationship subset corresponding to the initial physical failure stage, core feature associations and key causal relationships related to the initial failure signs are screened out, and the core feature associations and key causal relationships are used as the starting point for tracing the root cause of physical failure. Based on the evolution sequence of the failure process, the feature association subsets and causal relationship subsets of adjacent physical failure stages are sequentially associated, the transmission trajectory of core feature associations between each level is tracked, and the continuation logic of key causal relationships between adjacent levels is analyzed. Based on the transmission trajectory of core features and the continuation logic of key causal relationships between adjacent levels, an association chain is created from the starting tracing point to the final failure stage. In the complete association chain, the key feature nodes and causal links that trigger the failure evolution at each stage are marked to obtain the evolution path of the physical failure root cause. The evolution path of the physical failure root cause is compared and calibrated with the layer-by-layer feature association results so that each feature node in the evolution path corresponds to the feature association combination in the layer-by-layer feature association results. The causal links in the evolutionary path are matched and optimized with the single-layer causal deduction results in the layer-by-layer causal deduction results; By combining the initial conditions of physical failure root causes, key triggering factors in the evolution process, characteristic changes at each stage, and the formation logic of the final failure state, the evolution path of the calibrated and verified physical failure root causes, the layer-by-layer feature association results, and the layer-by-layer causal deduction results after matching optimization are integrated to generate a state identification result containing the physical failure root cause path.

9. A system for identifying the status information of electronic components, characterized in that, include: A processor; a storage device having a computer program stored thereon; a network interface for providing network communication functions; when the computer program is executed by the processor, the processor enables the processor to implement the artificial intelligence-based electronic component status information recognition method as described in any one of claims 1-8.

10. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions, which, when executed by a processor, implement the artificial intelligence-based electronic component status information recognition method as described in any one of claims 1-8.

Images