Intelligent fault prediction method and system for power adapter

Abstract

The present invention relates to the field of fault prediction technology, and in particular to an intelligent fault prediction method and system for a power adapter. The method comprises the following steps: obtaining a circuit board structure diagram inside the power adapter; marking circuit component nodes based on the circuit board structure diagram; performing spatial topological structure analysis on the circuit component nodes, thereby obtaining a component spatial topological structure diagram; performing three-dimensional topological reconstruction on the circuit component nodes, and performing iterative optimization of local details, thereby constructing a local optimized three-dimensional topological structure model; obtaining multi-dimensional operating status data of different parts in the power adapter based on multiple sensors; performing state feature analysis on the multi-dimensional operating status data of the different parts, and performing state trend evolution, thereby obtaining state trend evolution data of different parts. The present invention realizes efficient and accurate power adapter fault prediction.

Description

Technical Field

[0001] The present invention relates to the technical field of fault prediction, and in particular to an intelligent fault prediction method and system for a power adapter. Background Art

[0002] In modern electronic devices, power adapters, as key power supply components, undertake the important task of converting and distributing electrical energy. With the popularization of smart devices and the improvement of industrial automation, power adapters are widely used in various electronic products, and their stability and reliability are crucial to the normal operation of the equipment. However, during long-term use, the power adapter may be affected by various factors, such as circuit aging, overload operation, ambient temperature changes, etc., leading to failures.

[0003] Traditional power adapter fault prediction methods mainly rely on regular manual inspection and manual testing, which have the problems of inaccurate fault prediction and low efficiency. With the rapid development of artificial intelligence and data analysis technology, intelligent power adapter fault prediction methods have become the key to solving this problem. In order to meet the needs of modern power adapter fault prediction, an intelligent method is needed to realize real-time monitoring and fault prediction of the working status of the power adapter. Summary of the invention

[0004] In order to solve the above technical problems, the present invention proposes an intelligent fault prediction method and system for a power adapter to solve at least one of the above technical problems.

[0005] To achieve the above object, the present invention provides an intelligent fault prediction method for a power adapter, comprising the following steps:

[0006] Step S1: obtaining a circuit board structure diagram inside a power adapter; marking circuit component nodes based on the circuit board structure diagram; performing spatial topological structure analysis on the circuit component nodes, thereby obtaining a component spatial topological structure diagram;

[0007] Step S2: reconstructing the three-dimensional topology of the circuit component nodes and iteratively optimizing the local details, thereby constructing a local optimized three-dimensional topological structure model;

[0008] Step S3: acquiring multi-dimensional operating status data of different parts of the power adapter based on the multi-sensor; performing state characteristic analysis on the multi-dimensional operating status data of different parts, and performing state trend evolution, thereby obtaining state trend evolution data of different parts;

[0009] Step S4: performing topological positioning matching on the local optimized three-dimensional topological structure model based on the state trend evolution data of different parts, and then performing dynamic mapping processing to construct a dynamic trend evolution twin model;

[0010] Step S5: performing supernormal parameter operation simulation on the dynamic trend evolution twin model, and performing component performance attenuation calculations one by one to generate supernormal performance attenuation data for each component;

[0011] Step S6: Based on the abnormal performance attenuation data of each component, the fault probability of the circuit component node is predicted, and the fault distribution is visualized to build a fault probability situation visualization model.

[0012] By acquiring the circuit board structure diagram, the present invention can deeply understand the internal composition of the power adapter and provide basic data for subsequent analysis. Marking the circuit component nodes and performing spatial topological structure analysis are helpful to establish the overall structural model of the circuit board and provide a basis for fault prediction. Three-dimensional topological reconstruction and local detail iterative optimization can more clearly present the spatial layout and connection relationship of the components inside the circuit board and provide more detailed structural information for subsequent analysis. Constructing a local optimized three-dimensional topological structure model is helpful to deeply understand the relationship between the components inside the circuit board and provide visual support for fault prediction. The acquisition of multi-dimensional operating status data can comprehensively record the operating conditions of different parts of the power adapter and provide comprehensive data support for fault prediction. State feature analysis and state trend evolution are helpful to identify changes in the operating conditions of various parts inside the power adapter and can detect potential fault signs in advance. Topological positioning matching and dynamic mapping processing can match status trends with specific components, helping to understand the relationship between status data and circuit board structure. Building a dynamic trend evolution twin model can help to associate status data with actual components, providing a more accurate basis for subsequent fault prediction. Abnormal parameter operation simulation and performance degradation calculation can help simulate the operation of the power adapter under abnormal conditions, providing more scenario analysis for fault prediction. Generating abnormal performance degradation data of components can help evaluate the health status and life of components, and provide data support for fault probability prediction. Fault probability prediction and fault distribution visualization can help identify potential fault sources and risk locations, take preventive measures in advance, and reduce fault risks. Building a fault probability situation visualization model can help to intuitively display the failure probability of each component of the power adapter, providing guidance and decision support for maintenance personnel.

[0013] Preferably, step S1 comprises the following steps:

[0014] Step S11: obtaining a circuit board structure diagram inside the power adapter;

[0015] Step S12: performing high-frequency filtering and noise reduction optimization on the circuit board structure diagram to generate a filtering optimization structure diagram;

[0016] Step S13: visually identify components of the filter optimization structure diagram and mark circuit component nodes;

[0017] Step S14: calculating the spatial position of the circuit component node in the circuit board to obtain the spatial position coordinates of the component node;

[0018] Step S15: performing component layout distribution analysis on the component node spatial position coordinates to obtain component layout distribution data;

[0019] Step S16: Performing a spatial topological structure analysis on the filtering optimization structure diagram based on the component layout distribution data, thereby obtaining a component spatial topological structure diagram.

[0020] The present invention obtains the circuit board structure diagram in order to deeply understand the circuit design and layout inside the power adapter, and provide basic data for subsequent analysis and optimization. High-frequency filtering noise reduction optimization helps to reduce interference and noise in the circuit image, and improves the accuracy and reliability of subsequent processing. Generating a filtering optimization structure diagram can provide a clear image basis for subsequent identification and analysis. Component visual recognition can identify various electronic components on the circuit board, and help identify the composition and function of the circuit board. Marking circuit component nodes is helpful for subsequent spatial position calculation and fault prediction analysis. By calculating the spatial position coordinates of the component nodes, the accurate position of each electronic component on the circuit board can be determined, and basic data is provided for subsequent analysis. Obtaining the spatial position coordinates of the component nodes is helpful for fault prediction and layout optimization. Component layout distribution analysis can evaluate the arrangement of electronic components on the circuit board, and find possible layout problems or optimization space. Obtaining component layout distribution data is helpful for optimizing the circuit board design and improving the performance and reliability of the power adapter. Spatial topological structure analysis can deeply understand the connection relationship and layout characteristics between the components on the circuit board, and provide an important basis for fault prediction and optimization. Obtaining the component spatial topological structure diagram is helpful for finding potential fault points and improvement space, and improving the reliability and stability of the power adapter.

[0021] Preferably, step S2 comprises the following steps:

[0022] Step S21: Identify the circuit connections between components of the circuit component nodes to obtain component connection relationships;

[0023] Step S22: performing logical association mining on component connection relationships based on the spatial position coordinates of component nodes to obtain logical association data between components;

[0024] Step S23: reconstructing the component space topology structure diagram in three dimensions based on the logical association data between components to generate a component three-dimensional topology structure model;

[0025] Step S24: Iteratively optimize the local details of the component three-dimensional topological structure model, so as to construct a local optimized three-dimensional topological structure model.

[0026] The present invention helps to understand the overall structure of the circuit board by identifying the circuit connection relationship between components, providing key information for subsequent fault prediction. Obtaining the component connection relationship can help determine the dependency relationship between components and provide input data for the fault prediction algorithm. Logical association mining can discover the logical relationship between components based on the spatial position coordinates of the component nodes, which helps to understand the functional structure of the circuit board. Obtaining the logical association data between components can provide deeper data support for the subsequent fault prediction algorithm. Three-dimensional topology reconstruction can generate a three-dimensional topological structure model that reflects the actual physical structure based on the logical association data between components. Generating a component three-dimensional topological structure model helps to understand the structure of the circuit board more intuitively and provide a more accurate basis for fault prediction. Local detail iterative optimization can fine-tune the three-dimensional topological structure model to improve the accuracy and reliability of the model. Constructing a local optimized three-dimensional topological structure model can make the model closer to the actual situation and provide more reliable input data for the fault prediction algorithm.

[0027] Preferably, the specific steps of step S24 are:

[0028] Detect the topological connections of the component's three-dimensional topological structure model and extract the topological connection curve;

[0029] Perform multi-point sampling on the topological connection curve to extract multiple connection curve sampling points;

[0030] Calculate the geometric curvature of multiple connection curve sampling points to obtain the curvature of each sampling point;

[0031] According to the preset standard parameters of the curve, the curvature of each sampling point is identified for deviation, and the deviation sampling points are extracted and marked as curve sawtooth points;

[0032] Correct the sawtooth parameters of the curve sawtooth points to obtain the correction curve sampling points;

[0033] Smoothing the connection port according to the sampling points of the correction curve to obtain a smooth topological connection curve;

[0034] Iteratively smooth all the curve jagged points and extract all the smooth topological connection curves;

[0035] Based on all the smooth topological connection curves, the local details of the component three-dimensional topological structure model are optimized to construct a local optimized three-dimensional topological structure model.

[0036] The present invention can accurately capture the connection relationship between components through the detection of topological connections and curve extraction, providing a basis for subsequent analysis and processing. Multi-point sampling can more comprehensively describe the shape and characteristics of the curve, and provide more data points for subsequent curvature calculation and analysis. Calculating curvature can help understand the bending degree and shape change of the curve, and provide basic data for subsequent deviation identification and correction. Identifying deviation points in the curve can help discover potential problems and provide guidance for further optimization and correction. Correcting the curve can improve the smoothness and accuracy of the curve, and provide a better basis for subsequent smooth processing and optimization. Smoothing can improve the appearance and quality of the curve, making the overall structure more readable and accurate. Through iterative smoothing, the quality and accuracy of the curve can be further improved, ensuring the smoothness and continuity of the connection curve. Local detail optimization can improve the details of the model, making the overall structure more accurate and complete, and providing a more reliable basis for fault prediction.

[0037] Preferably, step S3 specifically comprises the following steps:

[0038] Step S31: acquiring multi-dimensional operating status data of different parts of the power adapter based on multiple sensors;

[0039] Step S32: performing state characteristic analysis on the multi-dimensional operation state data of the different parts to obtain state characteristic data of the different parts;

[0040] Step S33: performing time series change analysis on the state characteristic data of different parts to obtain time series state change data of different parts;

[0041] Step S34: performing state association mining between different parts on the state feature data of different parts to obtain state association features of different parts;

[0042] Step S35: performing state trend evolution on the time series state change data of different parts according to the state association characteristics of different parts, so as to obtain the state trend evolution data of different parts.

[0043] The multidimensional data obtained by the present invention through multiple sensors can fully reflect the operating status of each part inside the power adapter, providing a rich information basis. The multidimensional acquisition of the operating status data is helpful to fully understand the working condition of the power adapter and provide necessary data support for subsequent analysis. The state feature analysis can extract key state features from the multidimensional data to help understand the working state characteristics of each part. The state feature data obtained is helpful to discover potential anomalies or laws, providing a basis for subsequent analysis and prediction. The time series change analysis can reveal the evolution of the state characteristics of each part over time, help understand the state change law of each part, and obtain the time series state change data to provide historical data support for fault prediction and discover the trend and law of state change. The state association mining can discover the state association characteristics between different parts and reveal the mutual influence and dependence between the parts. The state association characteristics obtained are helpful to comprehensively analyze the state changes between the parts and provide a more comprehensive perspective for comprehensive fault prediction. The state trend evolution can predict the state trend development of different parts according to the state association characteristics, help to discover potential faults in advance, and the state trend evolution data obtained can provide a prediction basis in the time dimension for the fault prediction algorithm, thereby improving the accuracy and timeliness of the prediction.

[0044] Preferably, the specific steps of step S4 are:

[0045] Step S41: performing multi-part subdivision processing on the local optimized three-dimensional topological structure model, thereby obtaining a plurality of three-dimensional topological units;

[0046] Step S42: performing topological positioning matching on multiple three-dimensional topological units based on the state trend evolution data of different parts to obtain state trend positioning unit data;

[0047] Step S43: dynamically mapping the state trend evolution data of different parts to the local optimized three-dimensional topological structure model according to the state trend positioning unit data to construct a dynamic trend evolution twin model.

[0048] The present invention can refine the three-dimensional topological structure model of the overall power adapter into three-dimensional topological units of multiple parts through multi-part subdivision processing, thereby improving the fineness and accuracy of the structure. Obtaining multiple three-dimensional topological units helps to decompose complex structural problems into smaller units, providing more specific data support for subsequent analysis and positioning. Topological positioning matching can match the state trend evolution data of different parts with the three-dimensional topological units to determine the position of each part in the structural model. Obtaining the state trend positioning unit data helps to correspond the actual state data with the structural model, and establish a connection for subsequent analysis and prediction. Dynamic mapping processing can dynamically map the state trend evolution data of different parts to the locally optimized three-dimensional topological structure model according to the state trend positioning unit data. Constructing a dynamic trend evolution twin model helps to associate the actual state data with the structural model, and realize dynamic prediction and presentation of state trends.

[0049] Preferably, the specific steps of step S5 are:

[0050] Step S51: performing normalized mean calculation on the multi-dimensional operating state data of the different parts to obtain a normal range of the multi-dimensional operating state;

[0051] Step S52: defining super-normal state parameters based on the normal range of the multi-dimensional operating state;

[0052] Step S53: performing an ultra-normal parameter operation simulation on the dynamic trend evolution twin model based on the ultra-normal state parameters to generate ultra-normal operation simulation data;

[0053] Step S54: performing multi-component operation response identification on the super-normal operation simulation data, and extracting multi-component operation response data;

[0054] Step S55: Calculate the performance attenuation of each component on the multi-component operation response data to generate abnormal performance attenuation data for each component.

[0055] The present invention can obtain the normal range of the multi-dimensional operating state of each part through normalized mean calculation, which is helpful to determine the reference range of the normal operating state. Obtaining the normal range is helpful for subsequent abnormal detection and fault prediction, and can more effectively identify abnormal situations. Defining super-normal state parameters can help identify states beyond the normal range and help capture potential abnormal situations. The definition of super-normal state parameters provides an important basis for subsequent abnormal detection and prediction. The operation simulation can simulate the working conditions of the power adapter under super-normal parameters, which is helpful to understand the device response under abnormal conditions. Generating super-normal operation simulation data can provide data support for fault characteristics under abnormal conditions, which is conducive to abnormal detection and prediction. Multi-component operation response identification can analyze the response of each component in the super-normal operation simulation data and discover abnormal responses. Extracting multi-component operation response data is helpful to identify possible fault modes or abnormal situations and provide a basis for further analysis. Component-by-component performance attenuation calculation can evaluate the super-normal performance attenuation of each component, which helps to understand the health status of each component. Generating super-normal performance attenuation data can provide more specific component-level information for fault prediction, which is helpful to achieve refined fault prediction and maintenance.

[0056] Preferably, the specific steps of step S6 are:

[0057] Step S61: performing component life prediction on the abnormal performance attenuation data of each component to generate a component life prediction value;

[0058] Step S62: performing failure probability prediction on circuit component nodes based on component life prediction values, and generating failure probability prediction data for each component;

[0059] Step S63: performing failure probability situation evolution on the failure probability prediction data of each component to obtain a failure probability situation curve of each component;

[0060] Step S64: Visualize the fault distribution of the dynamic trend evolution twin model based on the fault probability situation curve of each component, and build a fault probability situation visualization model.

[0061] The present invention can predict the life of each component based on the abnormal performance attenuation data through component life prediction, which is helpful to discover components that may fail in advance. Generating component life prediction values ​​can help formulate a more reasonable maintenance plan and extend the service life of the power adapter. The failure probability prediction can predict the probability of failure of each component based on the component life prediction value, and provides an assessment of the component failure risk. The generation of failure probability prediction data can help determine which components are more prone to problems and perform targeted maintenance and replacement. The failure probability situation evolution can analyze the changing trend of the failure probability of each component over time, and help understand the development law of the component failure probability. Obtaining the failure probability situation curve can provide maintenance personnel with dynamic failure risk information, which is helpful to take timely measures to avoid failures. The failure probability situation visualization model can display the failure probability situation of each component in an intuitive way, helping users to quickly understand the failure distribution. Constructing a failure probability situation visualization model helps to monitor the failure risk of each component of the power adapter in real time and improve maintenance efficiency and timeliness.

[0062] In this specification, a smart fault prediction system for a power adapter is provided, which is used to execute the smart fault prediction method for the power adapter as described above, including:

[0063] A topology module is used to obtain a circuit board structure diagram inside a power adapter; mark circuit component nodes based on the circuit board structure diagram; and perform spatial topology analysis on the circuit component nodes to obtain a component spatial topology diagram;

[0064] The 3D topology reconstruction module reconstructs the 3D topology of the circuit component nodes and performs iterative optimization of local details to build a local optimized 3D topology structure model.

[0065] A state trend evolution module, which obtains multi-dimensional operation state data of different parts of the power adapter based on multiple sensors; performs state feature analysis on the multi-dimensional operation state data of different parts, and performs state trend evolution, thereby obtaining state trend evolution data of different parts;

[0066] The dynamic mapping module performs topological positioning and matching on the locally optimized three-dimensional topological structure model based on the state trend evolution data of different parts, and then performs dynamic mapping processing to construct a dynamic trend evolution twin model;

[0067] The super-normal simulation module simulates the super-normal parameter operation of the dynamic trend evolution twin model, calculates the performance attenuation of each component, and generates the super-normal performance attenuation data of each component;

[0068] The fault probability prediction module predicts the fault probability of circuit component nodes based on the abnormal performance attenuation data of each component, visualizes the fault distribution, and builds a fault probability situation visualization model.

[0069] The present invention can help understand the composition and layout of the circuit board by obtaining the internal circuit board structure diagram of the power adapter, mark the circuit component nodes and perform spatial topological structure analysis to help establish the spatial structure model of the circuit board, and provide a basis for subsequent analysis. Three-dimensional topological reconstruction and local detail iterative optimization can more intuitively present the three-dimensional structure of the circuit board, and construct a local optimized three-dimensional topological structure model to help deeply understand the spatial relationship between the internal components of the circuit board, and provide better visualization support for subsequent analysis. Performing state feature analysis and trend evolution based on multi-dimensional operating state data obtained by multiple sensors can help understand the state changes of different parts of the power adapter, and obtaining state trend evolution data can help predict possible fault trends, take maintenance measures in advance, and effectively deal with the situation. Optimizing the three-dimensional topological structure model for dynamic mapping processing can combine the state trend information with the actual structure, providing a more accurate basis for fault prediction. Building a dynamic trend evolution twin model can help monitor the state changes of the power adapter in real time and make corresponding adjustments and predictions. Performing super-normal parameter operation simulations and performance degradation calculations can help evaluate the long-term operation of circuit component nodes. Generating super-normal performance degradation data can help determine components that may fail and improve the accuracy of fault prediction. Fault probability prediction and visualization based on super-normal performance degradation data can help determine the probability distribution of potential faults. Building a fault probability situation visualization model can help users intuitively understand the failure probability of each component and guide maintenance decisions and the formulation of preventive measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 A schematic flow chart of the steps of an intelligent fault prediction method for a power adapter according to the present invention;

[0071] Figure 2 Detailed implementation flow chart of step S1;

[0072] Figure 3 Detailed implementation flow chart of step S2;

[0073] Figure 4 Detailed implementation flow chart of step S3. DETAILED DESCRIPTION

[0074] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0075] The present application example provides a method and system for intelligent fault prediction of a power adapter. The execution subject of the method and system for intelligent fault prediction of a power adapter includes but is not limited to: mechanical equipment, data processing platform, cloud server node, network upload device, etc. equipped with the system can be regarded as the general computing node of the present application, and the data processing platform includes but is not limited to: at least one of an audio and image management system, an information management system, and a cloud data management system.

[0076] See also Figures 1 to 4 The present invention provides an intelligent fault prediction method for a power adapter, and the intelligent fault prediction method for a power adapter comprises the following steps:

[0077] Step S1: obtaining a circuit board structure diagram inside a power adapter; marking circuit component nodes based on the circuit board structure diagram; performing spatial topological structure analysis on the circuit component nodes, thereby obtaining a component spatial topological structure diagram;

[0078] Step S2: reconstructing the three-dimensional topology of the circuit component nodes and iteratively optimizing the local details, thereby constructing a local optimized three-dimensional topological structure model;

[0079] Step S3: acquiring multi-dimensional operating status data of different parts of the power adapter based on the multi-sensor; performing state characteristic analysis on the multi-dimensional operating status data of different parts, and performing state trend evolution, thereby obtaining state trend evolution data of different parts;

[0080] Step S4: performing topological positioning matching on the local optimized three-dimensional topological structure model based on the state trend evolution data of different parts, and then performing dynamic mapping processing to construct a dynamic trend evolution twin model;

[0081] Step S5: performing supernormal parameter operation simulation on the dynamic trend evolution twin model, and performing component performance attenuation calculations one by one to generate supernormal performance attenuation data for each component;

[0082] Step S6: Based on the abnormal performance attenuation data of each component, the fault probability of the circuit component node is predicted, and the fault distribution is visualized to build a fault probability situation visualization model.

[0083] By acquiring the circuit board structure diagram, the present invention can deeply understand the internal composition of the power adapter and provide basic data for subsequent analysis. Marking the circuit component nodes and performing spatial topological structure analysis are helpful to establish the overall structural model of the circuit board and provide a basis for fault prediction. Three-dimensional topological reconstruction and local detail iterative optimization can more clearly present the spatial layout and connection relationship of the components inside the circuit board and provide more detailed structural information for subsequent analysis. Constructing a local optimized three-dimensional topological structure model is helpful to deeply understand the relationship between the components inside the circuit board and provide visual support for fault prediction. The acquisition of multi-dimensional operating status data can comprehensively record the operating conditions of different parts of the power adapter and provide comprehensive data support for fault prediction. State feature analysis and state trend evolution are helpful to identify changes in the operating conditions of various parts inside the power adapter and can detect potential fault signs in advance. Topological positioning matching and dynamic mapping processing can match status trends with specific components, helping to understand the relationship between status data and circuit board structure. Building a dynamic trend evolution twin model can help to associate status data with actual components, providing a more accurate basis for subsequent fault prediction. Abnormal parameter operation simulation and performance degradation calculation can help simulate the operation of the power adapter under abnormal conditions, providing more scenario analysis for fault prediction. Generating abnormal performance degradation data of components can help evaluate the health status and life of components, and provide data support for fault probability prediction. Fault probability prediction and fault distribution visualization can help identify potential fault sources and risk locations, take preventive measures in advance, and reduce fault risks. Building a fault probability situation visualization model can help to intuitively display the failure probability of each component of the power adapter, providing guidance and decision support for maintenance personnel.

[0084] In the embodiment of the present invention, refer to Figure 1 , is a schematic flow chart of the steps of an intelligent fault prediction method for a power adapter of the present invention. In this example, the steps of the intelligent fault prediction method for a power adapter include:

[0085] Step S1: obtaining a circuit board structure diagram inside a power adapter; marking circuit component nodes based on the circuit board structure diagram; performing spatial topological structure analysis on the circuit component nodes, thereby obtaining a component spatial topological structure diagram;

[0086] In this embodiment, a structural diagram of the internal circuit board of the power adapter is obtained by disassembly or scanning, and the structural diagram includes the shape, size and layout information of each electronic component of the circuit board. Various electronic components, such as inductors, capacitors, transformers, etc., are identified on the circuit board structural diagram, and these key components are marked and numbered to form a component node list and the marked component node position information. The spatial layout relationship between the components is analyzed to construct a three-dimensional spatial topological structure model of the power adapter circuit board.

[0087] Step S2: reconstructing the three-dimensional topology of the circuit component nodes and iteratively optimizing the local details, thereby constructing a local optimized three-dimensional topological structure model;

[0088] In this embodiment, based on the spatial distribution information of component nodes, three-dimensional modeling software such as CAD, 3DS Max, etc. is used to construct a corresponding three-dimensional model entity for each component node, and these three-dimensional models are connected according to the topological relationship between the components to form a three-dimensional topological model that describes the overall structure of the circuit board. The geometric shape and layout of each component in the three-dimensional topological model are checked, and the local details in the model are adjusted and optimized according to the actual circuit board structure diagram to ensure that the three-dimensional topological structure model is highly consistent with the actual circuit board structure, thereby forming a more refined and practical three-dimensional topological structure model, which not only includes the overall structure of the circuit board, but also optimizes the local details.

[0089] Step S3: acquiring multi-dimensional operating status data of different parts of the power adapter based on the multi-sensor; performing state characteristic analysis on the multi-dimensional operating status data of different parts, and performing state trend evolution, thereby obtaining state trend evolution data of different parts;

[0090] In this embodiment, various types of sensors are arranged at key parts of the power adapter, such as the input end, the output end, the circuit board, etc., including physical quantity sensors such as voltage, current, temperature, and vibration, to collect multi-dimensional operating status data of these parts in real time, and the collected multi-dimensional status data is analyzed to extract statistical characteristics of the operating status of each part, such as the mean, variance, extreme value, etc., to identify the significant characteristics of the operating status of each part, analyze the changing trend of the operating status of each part over time, and establish a mathematical model to describe the law of state evolution.

[0091] Step S4: performing topological positioning matching on the local optimized three-dimensional topological structure model based on the state trend evolution data of different parts, and then performing dynamic mapping processing to construct a dynamic trend evolution twin model;

[0092] In this embodiment, the operating status trend evolution data of each part are associated with the constructed local optimized three-dimensional topological model, the operating status data corresponding to each component node in the three-dimensional model are identified, the status data are mapped to each component node of the three-dimensional topological model, and the operating status of different parts is represented by visual elements such as color, shape, and size, so that the three-dimensional model can intuitively present the operating status of each part of the power adapter. The generated dynamic mapping three-dimensional model is associated with the obtained status trend evolution data to form a twin model that can reflect the dynamic operating status of the power adapter in real time.

[0093] Step S5: performing supernormal parameter operation simulation on the dynamic trend evolution twin model, and performing component performance attenuation calculations one by one to generate supernormal performance attenuation data for each component;

[0094] In this embodiment, some abnormal operating conditions beyond the normal range are set for the normal operating parameters of the power adapter, including abnormal input voltage, output load, ambient temperature, etc., and the operating status under these extraordinary conditions is simulated in the twin model. For the simulated extraordinary conditions, the performance degradation of each key component in the twin model, such as inductor, capacitor, transformer, etc., is calculated under extraordinary conditions. The calculated performance degradation results of each component are used to form a data set describing the performance degradation of each key component of the power adapter under extraordinary conditions, so as to provide support for subsequent fault prediction and health management.

[0095] Step S6: Based on the abnormal performance attenuation data of each component, the fault probability of the circuit component node is predicted, and the fault distribution is visualized to build a fault probability situation visualization model.

[0096] In this embodiment, the abnormal performance attenuation data of each component is used to establish a mathematical model that describes the component failure probability and the degree of performance attenuation. According to the model, the probability of failure of each component is predicted, and the predicted failure probability of each component is mapped back to the constructed three-dimensional topological structure model, so that the corresponding failure probability data is presented on each component node. The component failure probability is represented by visual encoding such as color and size, and the failure probability distribution is dynamically presented on the three-dimensional topological structure model, forming a visualization model that intuitively reflects the failure risk situation of the entire machine.

[0097] In this embodiment, refer to Figure 2 , is a flowchart of detailed implementation steps of step S1. In this embodiment, the detailed implementation steps of step S1 include:

[0098] Step S11: obtaining a circuit board structure diagram inside the power adapter;

[0099] Step S12: performing high-frequency filtering and noise reduction optimization on the circuit board structure diagram to generate a filtering optimization structure diagram;

[0100] Step S13: visually identify components of the filter optimization structure diagram and mark circuit component nodes;

[0101] Step S14: calculating the spatial position of the circuit component node in the circuit board to obtain the spatial position coordinates of the component node;

[0102] Step S15: performing component layout distribution analysis on the component node spatial position coordinates to obtain component layout distribution data;

[0103] Step S16: Performing a spatial topological structure analysis on the filtering optimization structure diagram based on the component layout distribution data, thereby obtaining a component spatial topological structure diagram.

[0104] In this embodiment, the power adapter is disassembled to expose the internal circuit board. A high-resolution camera can be used to photograph the circuit board to obtain its structural image. The structural image should contain the specific position and connection information of each electronic component on the circuit board. The original circuit board structural diagram is filtered using digital signal processing technology, with a focus on filtering out high-frequency noise components to improve image clarity and detail discernibility. The structural diagram after filtering optimization should highlight the outline and layout of each component on the circuit board. Computer vision and pattern recognition algorithms are used to automatically identify various electronic components on the circuit board. For each identified component, its key nodes, such as pins, solder joints, etc., are marked on the structural diagram. The completed structural diagram should show clear component positions and nodes. Point information, according to the scale information of the circuit board structure diagram, calculate the two-dimensional coordinates of each component node on the circuit board plane, and use mathematical transformation to convert these plane coordinates into coordinates describing the three-dimensional spatial position of the component in the circuit board, forming a complete component node spatial position data set, statistically analyze the component node coordinate data obtained, and mine the layout characteristics of the components on the circuit board, such as clustered distribution, sparse distribution, etc., and extract statistical indicators and data describing the overall layout distribution law of the components. Combined with the obtained component layout distribution information, the spatial topological analysis of the circuit board structure diagram after filtering optimization is performed, and a structural diagram describing the spatial layout topological relationship between circuit board components is constructed. The structural diagram can intuitively reflect the physical layout of the components and their mutual relationship.

[0105] In this embodiment, refer to Figure 3 , is a flowchart of detailed implementation steps of step S2. In this embodiment, the detailed implementation steps of step S2 include:

[0106] Step S21: Identify the circuit connections between components of the circuit component nodes to obtain component connection relationships;

[0107] Step S22: performing logical association mining on component connection relationships based on the spatial position coordinates of component nodes to obtain logical association data between components;

[0108] Step S23: reconstructing the component space topology structure diagram in three dimensions based on the logical association data between components to generate a component three-dimensional topology structure model;

[0109] Step S24: Iteratively optimize the local details of the component three-dimensional topological structure model, so as to construct a local optimized three-dimensional topological structure model.

[0110] In this embodiment, the physical distance and spatial distribution relationship between the nodes of each component are analyzed, and the logical connection relationship between the components is mined according to the component type and connection rules to form a data set describing the logical association between the components. According to the logical association data, the topological structure diagram between the components is designed to determine the nodes (components) and edges (connection relationships). Use three-dimensional modeling software (such as Blender, SketchUp or CAD software) to build a three-dimensional topological structure model according to the topological structure diagram. Determine the three-dimensional position, size and shape of the component. Optimize the details of the three-dimensional model to ensure its visual and functional rationality and adjust the connection relationship between the components. Analyze the three-dimensional topological structure model to identify local details that need to be optimized, such as connection points, distances between components or layouts. Apply iterative optimization methods (such as genetic algorithms, simulated annealing or gradient descent) to adjust and optimize local details and improve the overall performance of the model. Perform multiple rounds of iterations, evaluate the effect of each optimization and select the optimal result. Generate an optimized local three-dimensional topological structure model and visualize it for easy understanding and application.

[0111] In this embodiment, the specific steps of step S24 are:

[0112] Detect the topological connections of the component's three-dimensional topological structure model and extract the topological connection curve;

[0113] Perform multi-point sampling on the topological connection curve to extract multiple connection curve sampling points;

[0114] Calculate the geometric curvature of multiple connection curve sampling points to obtain the curvature of each sampling point;

[0115] According to the preset standard parameters of the curve, the curvature of each sampling point is identified for deviation, and the deviation sampling points are extracted and marked as curve sawtooth points;

[0116] Correct the sawtooth parameters of the curve sawtooth points to obtain the correction curve sampling points;

[0117] Smoothing the connection port according to the sampling points of the correction curve to obtain a smooth topological connection curve;

[0118] Iteratively smooth all the curve jagged points and extract all the smooth topological connection curves;

[0119] Based on all the smooth topological connection curves, the local details of the component three-dimensional topological structure model are optimized to construct a local optimized three-dimensional topological structure model.

[0120] In this embodiment, the physical connections between components in the three-dimensional topological structure model are analyzed, and the topological connection curve corresponding to each physical connection is extracted to form a set of curve data describing the three-dimensional topological connection relationship of the circuit board. The sampling interval and number are determined to ensure that the sampling points are evenly distributed on the connection curve. Multi-point sampling is performed on each topological connection curve to extract multiple connection curve sampling points. The sampled connection curve sampling points are sorted into a list or a data set, and a suitable curvature calculation method is selected, such as Gaussian curvature, average curvature, etc. For each connection curve sampling point, its geometric curvature is calculated using the selected method, and the calculated curvature value and the corresponding sampling point are sorted into a data set. The preset curve standard parameters, including the curvature threshold, etc., are determined. Deviation recognition is performed: Evaluate the curvature of each sampling point, identify points where the curvature deviates from the standard parameters, mark the identified sawtooth points, and record their positions and curvature values, select an appropriate correction algorithm, such as a curve smoothing algorithm or an interpolation algorithm, perform parameter correction on the sampling points marked as sawtooth points, generate corrected curve sampling points, save the corrected sampling points as a new data set, select an appropriate smoothing method, such as Bezier curve smoothing or spline curve smoothing, smooth the corrected curve sampling points, generate a smooth topological connection curve, save the smoothed connection curve for subsequent analysis, design an iterative process to ensure that all sawtooth points are smoothed, iteratively smooth each connection curve, continuously process all sawtooth points, formulate a local detail optimization strategy based on the smooth topological connection curve, implement local detail optimization in the three-dimensional topological structure model, and adjust the connection and curvature characteristics of the model.

[0121] In this embodiment, refer to Figure 4 , is a flowchart of detailed implementation steps of step S3. In this embodiment, the detailed implementation steps of step S3 include:

[0122] Step S31: acquiring multi-dimensional operating status data of different parts of the power adapter based on multiple sensors;

[0123] Step S32: performing state characteristic analysis on the multi-dimensional operation state data of the different parts to obtain state characteristic data of the different parts;

[0124] Step S33: performing time series change analysis on the state characteristic data of different parts to obtain time series state change data of different parts;

[0125] Step S34: performing state association mining between different parts on the state feature data of different parts to obtain state association features of different parts;

[0126] Step S35: performing state trend evolution on the time series state change data of different parts according to the state association characteristics of different parts, so as to obtain the state trend evolution data of different parts.

[0127] In this embodiment, a variety of sensors are arranged at key parts of the power adapter, including temperature sensors, current sensors, vibration sensors, etc., to collect multi-dimensional operation status data of each part in real time, determine a suitable feature extraction method, such as statistical features (mean, variance, etc.), frequency domain features (Fourier transform), etc., perform feature calculation on the operation status data of different parts, extract state feature data, select a suitable time series analysis method, such as sliding window method, time series analysis, etc., perform time series change analysis on the state feature data of different parts, identify change trends and patterns, organize the analyzed time series state change data into a data set, record the state change of each part, study the mutual relationship between the state features of each part, analyze the correlation and coupling of the state features of different parts, extract feature data describing the state association of different parts, select a suitable trend analysis method, such as linear regression, exponential smoothing method, etc., predict and deduce the state change trend of each part based on the state association feature, and form a set of data describing the future evolution trend of the state of different parts.

[0128] In this embodiment, step S4 includes the following steps:

[0129] Step S41: performing multi-part subdivision processing on the local optimized three-dimensional topological structure model, thereby obtaining a plurality of three-dimensional topological units;

[0130] Step S42: performing topological positioning matching on multiple three-dimensional topological units based on the state trend evolution data of different parts to obtain state trend positioning unit data;

[0131] Step S43: dynamically mapping the state trend evolution data of different parts to the local optimized three-dimensional topological structure model according to the state trend positioning unit data to construct a dynamic trend evolution twin model.

[0132] In this embodiment, the locally optimized three-dimensional topological structure model is imported into the corresponding modeling software or tool, and the subdivision strategy is determined, including the subdivision method (such as uniform subdivision or adaptive subdivision) and subdivision degree of each part. The locally optimized three-dimensional topological structure model is subdivided to generate multiple three-dimensional topological units. The grid subdivision algorithm or related functions in the CAD software can be used to verify the subdivided three-dimensional topological units to ensure that their geometric shapes and connection relationships meet expectations, organize the state trend evolution data of different parts to ensure that the data format is consistent, select a suitable topological positioning matching algorithm, such as nearest neighbor matching, shape matching, etc., match the state trend evolution data of different parts with the subdivided three-dimensional topological units, identify the state trend positioning units, determine the rules and methods of dynamic mapping, ensure that the state trend data can be accurately mapped to the corresponding three-dimensional topological units, and dynamically map the state trend evolution data of different parts to the locally optimized three-dimensional topological structure model according to the state trend positioning unit data. Data interpolation or mapping functions can be used to construct a dynamic trend evolution twin model to ensure that the model can reflect the state changes of different parts in real time.

[0133] In this embodiment, step S5 includes the following steps:

[0134] Step S51: performing normalized mean calculation on the multi-dimensional operating state data of the different parts to obtain a normal range of the multi-dimensional operating state;

[0135] Step S52: defining super-normal state parameters based on the normal range of the multi-dimensional operating state;

[0136] Step S53: performing an ultra-normal parameter operation simulation on the dynamic trend evolution twin model based on the ultra-normal state parameters to generate ultra-normal operation simulation data;

[0137] Step S54: performing multi-component operation response identification on the super-normal operation simulation data, and extracting multi-component operation response data;

[0138] Step S55: Calculate the performance attenuation of each component on the multi-component operation response data to generate abnormal performance attenuation data for each component.

[0139] In this embodiment, the multi-dimensional operation status data of different parts are sorted to ensure that the data is complete and the format is consistent. The multi-dimensional operation status data of each part are averaged to obtain the mean of different state parameters, and the normal range is calculated. The upper and lower limits of each state parameter can be determined by the mean ± standard deviation, and the standard of the super-normal state is determined, such as the threshold value exceeding the normal range (such as ±20% of the normal range). According to the normal range, the super-normal state parameters are defined, and which states are beyond the normal range are identified to ensure that the dynamic trend evolution twin model can be used for the operation simulation of super-normal parameters. The defined super-normal state parameters are input into the twin model, and the twin model is run to simulate the super-normal parameters. The operation effects under different states are observed, and the simulation results are recorded to generate super-normal operation simulation data, including the changes of each state parameter. A suitable response identification method is selected, such as a machine learning classification algorithm or a pattern recognition technology, and the super-normal operation simulation data is analyzed to identify the operation response of each component. The extracted multi-component operation response data is sorted into a data set, and the response characteristics of each component are recorded: Select a suitable performance decay calculation model, such as an exponential decay model or a linear decay model, perform performance decay calculations on the operating response data of each component, obtain the extraordinary performance decay data of each component, and organize the calculated extraordinary performance decay data into a data set to facilitate subsequent analysis and decision-making.

[0140] In this embodiment, step S6 includes the following steps:

[0141] Step S61: performing component life prediction on the abnormal performance attenuation data of each component to generate a component life prediction value;

[0142] Step S62: performing failure probability prediction on circuit component nodes based on component life prediction values, and generating failure probability prediction data for each component;

[0143] Step S63: performing failure probability situation evolution on the failure probability prediction data of each component to obtain a failure probability situation curve of each component;

[0144] Step S64: Visualize the fault distribution of the dynamic trend evolution twin model based on the fault probability situation curve of each component, and build a fault probability situation visualization model.

[0145] In this embodiment, the abnormal performance attenuation data of each component is sorted to ensure that the data format is consistent and suitable for analysis, and a suitable life prediction model is selected, such as a Weibull distribution model, an exponential distribution model, or a machine learning model. The abnormal performance attenuation data is input into the life prediction model to predict the component life, and a suitable failure probability prediction model is selected, such as a fault tree analysis (FTA) or a Markov model. According to the predicted life value of the component, the input parameters of the failure probability model are set, and the selected model is used to predict the failure probability of each component to generate the failure probability prediction data of each component. A suitable failure probability situation evolution method is selected, such as time series analysis or dynamic system modeling. According to the failure probability prediction data, the failure probability situation evolution is performed to obtain a curve of the failure probability of each component changing with time, and the failure probability situation curve of each component is saved as a graph or a data set for subsequent visualization. A suitable visualization tool or software is selected, such as Tableau, Matplotlib Or D3.js, design a fault probability situation visualization model, determine the display content and format (such as line chart, heat map, etc.), input the fault probability situation curve data into the visualization tool, and generate a visualization model of the fault distribution.

[0146] In this embodiment, a smart fault prediction system for a power adapter is provided, which is used to execute the smart fault prediction method for the power adapter as described above, including:

[0147] A topology module is used to obtain a circuit board structure diagram inside a power adapter; mark circuit component nodes based on the circuit board structure diagram; and perform spatial topology analysis on the circuit component nodes to obtain a component spatial topology diagram;

[0148] The 3D topology reconstruction module reconstructs the 3D topology of the circuit component nodes and performs iterative optimization of local details to build a local optimized 3D topology structure model.

[0149] A state trend evolution module, which obtains multi-dimensional operation state data of different parts of the power adapter based on multiple sensors; performs state feature analysis on the multi-dimensional operation state data of different parts, and performs state trend evolution, thereby obtaining state trend evolution data of different parts;

[0150] The dynamic mapping module performs topological positioning and matching on the locally optimized three-dimensional topological structure model based on the state trend evolution data of different parts, and then performs dynamic mapping processing to construct a dynamic trend evolution twin model;

[0151] The super-normal simulation module simulates the super-normal parameter operation of the dynamic trend evolution twin model, calculates the performance attenuation of each component, and generates the super-normal performance attenuation data of each component;

[0152] The fault probability prediction module predicts the fault probability of circuit component nodes based on the abnormal performance attenuation data of each component, visualizes the fault distribution, and builds a fault probability situation visualization model.

[0153] The present invention can help understand the composition and layout of the circuit board by obtaining the internal circuit board structure diagram of the power adapter, mark the circuit component nodes and perform spatial topological structure analysis to help establish the spatial structure model of the circuit board, and provide a basis for subsequent analysis. Three-dimensional topological reconstruction and local detail iterative optimization can more intuitively present the three-dimensional structure of the circuit board, and construct a local optimized three-dimensional topological structure model to help deeply understand the spatial relationship between the internal components of the circuit board, and provide better visualization support for subsequent analysis. Performing state feature analysis and trend evolution based on multi-dimensional operating state data obtained by multiple sensors can help understand the state changes of different parts of the power adapter, and obtaining state trend evolution data can help predict possible fault trends, take maintenance measures in advance, and effectively deal with the situation. Optimizing the three-dimensional topological structure model for dynamic mapping processing can combine the state trend information with the actual structure, providing a more accurate basis for fault prediction. Building a dynamic trend evolution twin model can help monitor the state changes of the power adapter in real time and make corresponding adjustments and predictions. Performing super-normal parameter operation simulations and performance degradation calculations can help evaluate the long-term operation of circuit component nodes. Generating super-normal performance degradation data can help determine components that may fail and improve the accuracy of fault prediction. Fault probability prediction and visualization based on super-normal performance degradation data can help determine the probability distribution of potential faults. Building a fault probability situation visualization model can help users intuitively understand the failure probability of each component and guide maintenance decisions and the formulation of preventive measures.

[0154] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0155] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. An intelligent fault prediction method for a power adapter, characterized in that: The following steps are involved: Step S1: obtaining a circuit board structure diagram inside a power adapter; marking circuit component nodes based on the circuit board structure diagram; Performing spatial topological structure analysis on circuit component nodes to obtain a component spatial topological structure diagram; Step S2: Reconstruct the three-dimensional topology of the circuit component nodes and perform iterative optimization of local details to construct a local optimized three-dimensional topological structure model; the specific steps of step S2 are: Step S21: Identify the circuit connections between components of the circuit component nodes to obtain component connection relationships; Step S22: performing logical association mining on component connection relationships based on the spatial position coordinates of component nodes to obtain logical association data between components; Step S23: reconstructing the component space topology structure diagram in three dimensions based on the logical association data between components to generate a component three-dimensional topology structure model; Step S24: Iteratively optimizing the local details of the component three-dimensional topological structure model, thereby constructing a local optimized three-dimensional topological structure model; Step S3: acquiring multi-dimensional operating status data of different parts of the power adapter based on the multi-sensor; performing state characteristic analysis on the multi-dimensional operating status data of different parts, and performing state trend evolution, thereby obtaining state trend evolution data of different parts; Step S4: Based on the state trend evolution data of different parts, the local optimized three-dimensional topological structure model is topologically positioned and matched, and then dynamically mapped to construct a dynamic trend evolution twin model; the specific steps of step S4 are: Step S41: performing multi-part subdivision processing on the local optimized three-dimensional topological structure model, thereby obtaining a plurality of three-dimensional topological units; Step S42: performing topological positioning matching on multiple three-dimensional topological units based on the state trend evolution data of different parts to obtain state trend positioning unit data; Step S43: dynamically mapping the state trend evolution data of different parts to the local optimized three-dimensional topological structure model according to the state trend positioning unit data to construct a dynamic trend evolution twin model; Step S5: simulate the supernormal parameter operation of the dynamic trend evolution twin model, and calculate the performance attenuation of each component one by one to generate supernormal performance attenuation data of each component; The specific steps of step S5 are: Step S51: performing normalized mean calculation on the multi-dimensional operating state data of the different parts to obtain a normal range of the multi-dimensional operating state; Step S52: defining super-normal state parameters based on the normal range of the multi-dimensional operating state; Step S53: performing an ultra-normal parameter operation simulation on the dynamic trend evolution twin model based on the ultra-normal state parameters to generate ultra-normal operation simulation data; Step S54: performing multi-component operation response identification on the super-normal operation simulation data, and extracting multi-component operation response data; Step S55: Calculating the performance attenuation of each component on the operation response data of multiple components, and generating abnormal performance attenuation data of each component; Step S6: Based on the abnormal performance attenuation data of each component, the fault probability of the circuit component node is predicted, and the fault distribution is visualized to build a fault probability situation visualization model.

2. The intelligent fault prediction method for a power adapter according to claim 1, characterized in that: The specific steps of step S1 are: Step S11: obtaining a circuit board structure diagram inside the power adapter; Step S12: performing high-frequency filtering and noise reduction optimization on the circuit board structure diagram to generate a filtering optimization structure diagram; Step S13: visually identify components of the filter optimization structure diagram and mark circuit component nodes; Step S14: calculating the spatial position of the circuit component node in the circuit board to obtain the spatial position coordinates of the component node; Step S15: performing component layout distribution analysis on the component node spatial position coordinates to obtain component layout distribution data; Step S16: Performing a spatial topological structure analysis on the filtering optimization structure diagram based on the component layout distribution data, thereby obtaining a component spatial topological structure diagram.

3. The intelligent fault prediction method for a power adapter according to claim 1, characterized in that: The specific steps of step S24 are: Detect the topological connections of the component's three-dimensional topological structure model and extract the topological connection curve; Perform multi-point sampling on the topological connection curve to extract multiple connection curve sampling points; Calculate the geometric curvature of multiple connection curve sampling points to obtain the curvature of each sampling point; According to the preset standard parameters of the curve, the curvature of each sampling point is identified for deviation, and the deviation sampling points are extracted and marked as curve sawtooth points; Correct the sawtooth parameters of the curve sawtooth points to obtain the correction curve sampling points; Smoothing the connection port according to the sampling points of the correction curve to obtain a smooth topological connection curve; Iteratively smooth all the curve jagged points and extract all the smooth topological connection curves; Based on all the smooth topological connection curves, the local details of the component three-dimensional topological structure model are optimized to construct a local optimized three-dimensional topological structure model.

4. The intelligent fault prediction method for a power adapter according to claim 1, characterized in that: The specific steps of step S3 are: Step S31: acquiring multi-dimensional operating status data of different parts of the power adapter based on multiple sensors; Step S32: performing state characteristic analysis on the multi-dimensional operation state data of the different parts to obtain state characteristic data of the different parts; Step S33: performing time series change analysis on the state characteristic data of different parts to obtain time series state change data of different parts; Step S34: performing state association mining between different parts on the state feature data of different parts to obtain state association features of different parts; Step S35: performing state trend evolution on the time series state change data of different parts according to the state association characteristics of different parts, so as to obtain the state trend evolution data of different parts.

5. The intelligent fault prediction method for a power adapter according to claim 1, characterized in that: The specific steps of step S6 are: Step S61: performing component life prediction on the abnormal performance attenuation data of each component to generate a component life prediction value; Step S62: performing failure probability prediction on circuit component nodes based on component life prediction values, and generating failure probability prediction data for each component; Step S63: performing failure probability situation evolution on the failure probability prediction data of each component to obtain a failure probability situation curve of each component; Step S64: Visualize the fault distribution of the dynamic trend evolution twin model based on the fault probability situation curve of each component, and build a fault probability situation visualization model.

6. An intelligent fault prediction system for a power adapter, characterized in that: The method for executing the intelligent fault prediction method of the power adapter according to claim 1 comprises: A topology module is used to obtain a circuit board structure diagram inside a power adapter; mark circuit component nodes based on the circuit board structure diagram; and perform spatial topology analysis on the circuit component nodes to obtain a component spatial topology diagram; The 3D topology reconstruction module reconstructs the 3D topology of the circuit component nodes and performs iterative optimization of local details to build a local optimized 3D topology structure model. A state trend evolution module, which obtains multi-dimensional operation state data of different parts of the power adapter based on multiple sensors; performs state feature analysis on the multi-dimensional operation state data of different parts, and performs state trend evolution, thereby obtaining state trend evolution data of different parts; The dynamic mapping module performs topological positioning and matching on the locally optimized three-dimensional topological structure model based on the state trend evolution data of different parts, and then performs dynamic mapping processing to construct a dynamic trend evolution twin model; The super-normal simulation module simulates the super-normal parameter operation of the dynamic trend evolution twin model, calculates the performance attenuation of each component, and generates the super-normal performance attenuation data of each component; The fault probability prediction module predicts the fault probability of circuit component nodes based on the abnormal performance attenuation data of each component, visualizes the fault distribution, and builds a fault probability situation visualization model.

Images