Electromagnetic compatibility intelligent test and diagnosis method and system based on knowledge graph

Abstract

The invention provides an electromagnetic compatibility intelligent test and diagnosis method and system based on a knowledge graph. The method comprises the following steps: acquiring multi-source data; generating a knowledge graph in the electromagnetic compatibility field according to the multi-source data, and generating a fault chain based on the knowledge graph; reasoning an interference propagation path based on the knowledge graph, and predicting interference intensity through a time domain finite difference method and a moment method based on the propagation path; and calculating the influence of parameter loss in the fault chain on the fault probability, and generating a rectification scheme in combination with the knowledge graph, the fault chain, the propagation path and the interference intensity. The technical problems that complex electromagnetic interference propagation paths are difficult to deal with and interference intensity prediction depends on a single simulation method in the prior art are solved.

Description

Technical Field

[0001] This invention relates to the field of electromagnetic diagnostic technology, and more specifically, to an electromagnetic compatibility intelligent testing and diagnostic method and system based on knowledge graphs. Background Technology

[0002] With the widespread application of information technology equipment and the increasing level of electronic and information technology in various application systems, the electromagnetic environment is becoming increasingly complex, and electromagnetic interference and electromagnetic compatibility issues are constantly emerging. Electromagnetic compatibility (EMC) is a key indicator for the normal operation of electronic equipment in the electromagnetic environment.

[0003] Existing technologies often rely on empirical rules to determine the causal relationship between interference sources and fault phenomena, which makes it difficult to cope with multi-path interference propagation in complex systems. This leads to fault chain generation relying on manual logic, which is prone to missing potential interference paths. At the same time, single simulation methods are difficult to predict the interference intensity in complex coupled scenarios.

[0004] Therefore, it is necessary to provide a new knowledge graph-based intelligent testing and diagnostic method for electromagnetic compatibility to solve the above-mentioned technical problems.

[0005] Chinese patent application CN202110273764.8 discloses an intelligent electromagnetic compatibility (EMC) testing and diagnosis method based on a knowledge graph. This patent's diagnostic process is based on an EMC knowledge graph and deep learning. It uses an RNN to combine five-tuple paths to generate relation vectors and achieves fault diagnosis through multi-layered reasoning. Simultaneously, it employs Stacking ensemble learning to fuse the reasoning results of various subclasses, forming a comprehensive reasoning model through weighted voting, thus improving diagnostic accuracy and intelligence. In contrast, this application constructs a knowledge graph, combines causal reasoning with electromagnetic simulation, identifies the causal chain between interference sources and faults, infers interference propagation paths, and assesses interference intensity, achieving intelligent diagnosis and localization of EMC problems. The technical approaches are significantly different. Summary of the Invention

[0006] In view of the shortcomings of the existing technology, the purpose of this invention is to provide an electromagnetic compatibility intelligent testing and diagnosis method and system based on knowledge graph.

[0007] According to one aspect of the present invention, an electromagnetic compatibility intelligent testing and diagnosis method based on a knowledge graph is provided, comprising the following steps: Step 1: acquiring multi-source data; Step 2: generating a knowledge graph in the field of electromagnetic compatibility based on the multi-source data, and generating a fault chain based on the knowledge graph; Step 3: inferring the interference propagation path based on the knowledge graph and the fault chain, and predicting the interference intensity based on the interference propagation path using the finite-difference time-domain method and the method of moments; Step 4: calculating the impact of missing parameters on the fault probability, and generating a rectification plan by combining the knowledge graph, the fault chain, the propagation path, and the interference intensity.

[0008] Preferably, the multi-source data includes structured data, unstructured data, and real-time data. The structured data includes equipment parameters, test standards, and historical test reports; the unstructured data includes technical documents, fault case descriptions, and expert experience; and the real-time data includes electromagnetic interference signals and environmental parameters collected by sensors.

[0009] Preferably, step one further includes: performing time-frequency analysis on the multi-source data through wavelet transform to extract the transient features of the interference signal, and then preprocessing the multi-source data.

[0010] Preferably, the wavelet transform involves multi-resolution analysis of the signal using mother wavelet functions at different scales, and includes the following steps: Continuous wavelet transform: The wavelet coefficients of signal S are calculated using the cwt function, where the cwt function is as follows: COEFS

[0011] in, It is a scale sequence (1:1:64). For Daubechies 4 wavelet bases; Scale to Frequency Conversion: The scal2frq function converts the scale into the actual frequency. The scal2frq function is as follows:

[0012] in, The sampling period; Time-frequency plots are drawn to locate the transient characteristics of interference signals.

[0013] Preferably, the preprocessing includes: text cleaning and standardization of terminology for unstructured data; filtering, noise reduction, and spectrum analysis for electromagnetic interference signals.

[0014] Preferably, step two includes: extracting entities, relationships, and attributes from multi-source data based on a deep learning model to generate a knowledge graph in the field of electromagnetic compatibility; then analyzing and obtaining the causal relationship between interference sources and fault phenomena based on a causal reasoning logic mechanism; and generating a fault chain in combination with logical rules. Entities include interference source type, sensitive equipment model, and test standard number; relationships include causal relationships; and attributes include quantitative parameters.

[0015] Preferably, the knowledge graph-based reasoning interference propagation path includes: Construct the knowledge graph nodes and edges; use Dijkstra's algorithm to calculate the shortest path, where Dijkstra's algorithm is as follows: , in, This represents the edge weight (e.g., propagation strength in dB).

[0016] The propagation of electromagnetic fields in the time domain is simulated using the finite-difference time-domain method. By iteratively updating the electric and magnetic field components, the coupling interference between devices is calculated. Then, the electromagnetic response of the shielding material is calculated using the method of moments. Finally, the current distribution is expanded using basis functions to solve for the scattered field.

[0017] Preferably, the calculation of the impact of missing parameters on the failure probability in the fault chain includes: constructing a fuzzy model using a membership function, quantifying the impact of missing parameters on the interference results, constructing a Bayesian network model, and weighting the conflict data with confidence.

[0018] Preferably, the method for generating the rectification plan includes: obtaining the global optimal solution through crossover and mutation operations based on a genetic algorithm, and obtaining the local optimal solution by simulating the cooperative behavior of a particle swarm based on a particle swarm optimization algorithm.

[0019] According to another aspect of the present invention, an electromagnetic compatibility intelligent testing and diagnostic system based on a knowledge graph includes: Module M1: Acquires multi-source data; Module M2: Generates a knowledge graph in the field of electromagnetic compatibility based on the multi-source data, and generates a fault chain based on the knowledge graph; Module M3: Based on the knowledge graph, infers the interference propagation path, and predicts the interference intensity based on the interference propagation path using the finite difference method in the time domain and the method of moments; Module M4: Calculates the impact of missing parameters on the fault probability, and generates a rectification plan by combining the knowledge graph, fault chain, propagation path, and interference intensity.

[0020] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention constructs a knowledge graph in the field of electromagnetic compatibility (EMC), analyzes the causal relationship between interference sources and fault phenomena based on causal reasoning logic, and generates fault chains by combining graph rules. This enables high-priority localization of interference paths, effectively reducing manual intervention in the diagnostic process and improving the automation and accuracy of problem localization. Compared to traditional troubleshooting methods that rely on human experience, this invention can still achieve rapid convergence and interpretable fault chain output in complex environments with multiple interference sources and multiple fault factors.

[0021] 2. This invention combines the finite-difference time-domain method and the method of moments to perform accurate simulations of interference propagation paths across multiple physical scales. In complex coupled scenarios where high-frequency interference and low-frequency noise coexist, it significantly improves the accuracy of interference intensity prediction. Experiments demonstrate that, compared to a single simulation model, this invention has a lower error tolerance in coupled interference modeling accuracy, making it particularly suitable for identifying and analyzing concealed interference paths in complex structures, thus providing a quantitative basis for remediation decisions. Attached Figure Description

[0022] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a flowchart illustrating a knowledge graph-based intelligent testing and diagnostic method for electromagnetic compatibility. Figure 2 This is a schematic diagram of the propagation path of inference interference in step three; Figure 3 This is a schematic diagram illustrating the composition of multi-source data. Detailed Implementation

[0023] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the protection scope of the present invention.

[0024] For ease of understanding, the terms or concepts involved in this application are explained below; (1) Wavelet Transform A time-frequency joint analysis method is proposed, which performs multi-resolution decomposition of signals using a multi-scale mother wavelet function. This method can be used to extract transient features from non-stationary signals and is suitable for the analysis of short-term burst signals such as electromagnetic interference.

[0025] (2) Knowledge Graph A graph structure is used to represent domain knowledge models, where nodes represent entities (such as interference sources and devices), edges represent relationships (such as causality and propagation paths), and attributes represent quantified parameters, which are used to support semantic modeling and intelligent reasoning.

[0026] (3) The logical mechanism of causal reasoning Based on knowledge graphs, by analyzing the causal relationships between entities, the causal chain between the source of interference and the fault phenomenon can be identified to support fault location, source tracing and logical explanation.

[0027] (4) Fault chain A fault chain is a logical chain constructed from entities and their causal relationships in a knowledge graph. It describes the causal reasoning path from an electromagnetic interference source to a fault phenomenon. It emphasizes the semantic and logical relationship that "interference events lead to fault results," and typically includes nodes such as interference source, interference mechanism, equipment status, and fault mode, reflecting the semantic hierarchy and behavioral logic of the interference problem. For example, "high-frequency power supply → cable coupling → communication module → data interruption" is a typical fault chain.

[0028] (5) Interference Propagation Path The interference propagation path refers to the physical path by which an electromagnetic interference signal travels from the interference source through a propagation medium such as space, structure, or conductor to the target device. It emphasizes the propagation behavior of the electromagnetic field and its spatial coupling characteristics. This path is composed of physical connections and is often used for electromagnetic simulation and propagation intensity assessment, such as simulating the propagation of electric fields between equipment structures using the FDTD method. The propagation path focuses on "how the signal is transmitted".

[0029] (6) Dijkstra's algorithm A shortest path calculation algorithm is provided to find the shortest path from a source node to a target node in a weighted graph. In this invention, it is used to infer interference propagation paths.

[0030] (7) Finite-Difference Time-Domain Method (FDTD) A numerical simulation method is proposed to simulate the propagation process of electromagnetic waves by discretizing Maxwell's equations in time and space, and to evaluate the coupling and propagation characteristics of interference.

[0031] (8) Method of Moments (MoM) A boundary integral numerical method, based on Green's function, is used to solve electromagnetic boundary problems and calculate the electromagnetic response characteristics of shielding materials or structures.

[0032] (9) Membership function In fuzzy logic, it is a function used to describe the degree to which a variable belongs to a certain fuzzy set, and is used to model the influence intensity under conditions of parameter uncertainty or missing parameters.

[0033] (10) Bayesian Network A probabilistic graphical model with a directed acyclic graph structure is used to model the conditional dependencies between variables, enabling reasoning and confidence-weighted analysis under uncertain information.

[0034] (10) Genetic Algorithm An optimization algorithm based on natural selection and genetic mechanisms searches for the global optimal solution in the solution space through operations such as selection, crossover, and mutation, and is used to generate rectification strategies.

[0035] (11) Particle Swarm Optimization (PSO) Algorithm A swarm intelligence algorithm simulating bird flock foraging behavior, where each particle represents a solution, rapidly approximates the optimal solution through collaborative search, and is used to optimize the local convergence performance of rectification schemes.

[0036] In view of this, this application provides an electromagnetic compatibility intelligent testing and diagnostic method based on knowledge graphs, such as... Figure 1As shown, this method collects multi-source data, extracts transient features of interference signals using wavelet transform, constructs a knowledge graph containing entities such as interference sources, equipment, and testing standards, and generates fault chains based on causal reasoning. On this basis, it infers the interference propagation path and evaluates the interference intensity through electromagnetic simulation. To address missing parameters and data conflicts, it utilizes fuzzy logic and Bayesian network modeling and analysis, and finally generates rectification schemes through genetic algorithms and particle swarm optimization, achieving intelligent diagnosis and optimized rectification of electromagnetic interference. The following provides a detailed explanation of each step: Step 1: Acquire multi-source data, perform time-frequency analysis on the multi-source data through wavelet transform, extract transient features of interference signals, and then preprocess the multi-source data.

[0037] It should be noted that, as Figure 3 As shown, the multi-source data includes structured data, unstructured data, and real-time data. The structured data includes equipment parameters, test standards, and historical test reports; the unstructured data includes technical documents, fault case descriptions, and expert experience; and the real-time data includes electromagnetic interference signals and environmental parameters collected by sensors.

[0038] It is understood that the wavelet transform involves performing multi-resolution analysis of the signal using mother wavelet functions at different scales, and includes the following steps: Continuous wavelet transform: The wavelet coefficients of signal S are calculated using the cwt function, which is as follows:

[0039] in, It is a scale sequence (1:1:64). For Daubechies 4 wavelet bases; Scale to Frequency Conversion: The scal2frq function converts the scale into the actual frequency. The scal2frq function is as follows:

[0040] in, The sampling period; Time-frequency plots are drawn to locate the transient characteristics of interference signals.

[0041] Further preprocessing includes: text cleaning and standardization of terminology for unstructured data; filtering, noise reduction, and spectral analysis of electromagnetic interference signals.

[0042] Based on the above scheme, the core of this step lies in using wavelet transform to perform time-frequency domain analysis on multi-source data to extract the transient features of electromagnetic interference signals. Wavelet transform achieves multi-resolution decomposition of the signal by shifting and scaling the mother wavelet function at different scales, thereby accurately capturing the local features of non-stationary signals in both time and frequency dimensions. Continuous wavelet transform (CWT) is used to calculate the wavelet coefficients of signal S, and the signal energy distribution is analyzed using a scale sequence (1:1:64) and the Daubechies 4 wavelet basis. Subsequently, the scal2frq function is used to convert the scale information into actual frequencies, realizing the mapping between scale and frequency, so as to identify the transient features of the interference signal in the time-frequency diagram. In multi-source data processing, the system integrates structured data (such as equipment parameters, test standards, and historical reports), unstructured data (such as technical documents, case descriptions, and expert experience), and real-time data (such as electromagnetic signals collected by sensors) to form a multi-dimensional input feature set. The preprocessing stage improves the consistency of unstructured data through text cleaning and terminology standardization, and enhances the quality of electromagnetic signals through filtering, noise reduction, and spectrum analysis, providing clean and stable input data for subsequent modeling and knowledge extraction.

[0043] Step 2: Based on a deep learning model, extract entities, relationships, and attributes from multi-source data to generate a knowledge graph in the field of electromagnetic compatibility. Then, based on a causal reasoning logic mechanism, analyze and obtain the causal relationship between the interference source and the fault phenomenon, and generate a fault chain by combining logical rules. Among them, entities include interference source type, sensitive equipment model, and test standard number; relationships include causal relationships; and attributes include quantitative parameters.

[0044] Based on the above scheme, this step utilizes a deep learning model to achieve semantic extraction and knowledge association of multi-source data, constructing a knowledge graph in the field of electromagnetic compatibility. The system first uses entity recognition and relation extraction models to identify key entities (such as interference source type, equipment model, and test standard number) and their causal relationships and attribute parameters from different data sources. Then, based on the graph structure, it forms connections between nodes and edges, transforming the data from discrete text and signal forms into a structured semantic network. In the causal reasoning stage, by combining causal logic mechanisms and expert knowledge rules, the system analyzes the causal path between the interference source and the fault phenomenon, uncovering potential interference mechanisms. Furthermore, based on logical rules, it generates a "fault chain," i.e., a logical relationship chain of interference source—propagation path—sensitive equipment—fault result, providing knowledge support for subsequent propagation prediction and solution formulation.

[0045] Step 3: Based on the fault chain, infer the interference propagation path, and based on the interference propagation path, predict the interference intensity using the finite-difference time-domain method and the method of moments. For example, such as Figure 2 As shown, the knowledge graph-based reasoning interference propagation path includes: Construct knowledge graph nodes and edges; The shortest path is calculated using Dijkstra's algorithm, which is as follows: , in, This represents the edge weight (e.g., propagation strength in dB).

[0046] The propagation of electromagnetic fields in the time domain is simulated using the finite-difference time-domain method. By iteratively updating the electric and magnetic field components, the coupling interference between devices is calculated. Then, the electromagnetic response of the shielding material is calculated using the method of moments. Finally, the current distribution is expanded using basis functions to solve for the scattered field.

[0047] Based on the above scheme, the process of reasoning about interference propagation paths using knowledge graphs includes three stages: constructing graph nodes and edges, path reasoning, and electromagnetic simulation evaluation. First, a knowledge graph in the field of electromagnetic compatibility is constructed, abstracting elements such as interference sources, propagation media, target devices, environmental parameters, and testing standards into nodes in the graph. Nodes represent specific entities. The physical connections between entities, the interference mechanisms, or propagation paths are abstracted into edges. Edges not only describe the connection direction between entities but can also include attributes such as propagation intensity, distance, and attenuation factor as weights.

[0048] In the constructed graph structure, a path search algorithm (such as Dijkstra's algorithm) is used to calculate the shortest path from the interference source to the sensitive device. The shortest path here represents the path with the strongest coupling effect or the lowest propagation cost during interference propagation, used to identify the main interference transmission channels, facilitating centralized simulation and diagnosis. The selection of this path is based on edge weight evaluation, reflecting factors such as propagation strength, impedance matching conditions, and transmission loss.

[0049] After obtaining the propagation path, the Finite-Difference Time-Domain (FDTD) method is further used to simulate the propagation process of the electromagnetic field along the path. FDTD, through the discretization of Maxwell's equations in both the time and space domains, enables iterative updates of the electric and magnetic field components, dynamically simulating the propagation behavior and coupling effects of interference signals within the structure. To further analyze the response of the terminal equipment or shielding structure to interference signals, the Method of Moments (MoM) is used to model the electromagnetic boundary problem. The MoM method, by constructing integral equations and expanding the current distribution using basis functions, can effectively solve for the scattered field characteristics and current response, thereby evaluating the protective effect of the shielding material.

[0050] Step 4: Calculate the impact of missing parameters in the fault chain on the fault probability, and generate a rectification plan by combining the knowledge graph, fault chain, propagation path, and interference intensity.

[0051] It is understood that the impact of missing calculation parameters on the probability of failure includes: constructing a fuzzy model using membership functions, quantifying the impact of missing parameters on interference results, constructing a Bayesian network model, and weighting the conflict data with confidence.

[0052] It should be noted that the generated rectification plan includes: obtaining the global optimal solution through crossover and mutation operations based on a genetic algorithm, and obtaining the local optimal solution by simulating the cooperative behavior of the particle swarm based on a particle swarm optimization algorithm.

[0053] Based on the above scheme, this step, using fuzzy inference and probabilistic models, calculates the impact of missing parameters on the failure probability and generates targeted remediation plans. The system first constructs a membership function to form a fuzzy model, which quantifies the uncertainty of interference results due to missing parameters. Then, a Bayesian network model is used to model the dependencies between data, and a confidence-weighted mechanism is used to fuse conflict information to obtain a dynamic assessment of the failure probability. In the remediation plan generation stage, the system combines multi-dimensional information such as knowledge graphs, failure chains, propagation paths, and interference intensity to establish an optimization objective function. A genetic algorithm is used for global optimization, exploring the optimal solution space using crossover and mutation operations; simultaneously, a particle swarm optimization algorithm is introduced for local optimization to simulate information sharing and convergence behavior among multiple particles. Through the combination of global and local optimization, the system can generate optimal remediation plans that balance cost and effectiveness, achieving intelligent decision support for electromagnetic compatibility issues.

[0054] The present invention also provides an electromagnetic compatibility intelligent testing and diagnostic system based on knowledge graphs. The electromagnetic compatibility intelligent testing and diagnostic system based on knowledge graphs can be implemented by executing the process steps of the electromagnetic compatibility intelligent testing and diagnostic method based on knowledge graphs. That is, those skilled in the art can understand the electromagnetic compatibility intelligent testing and diagnostic method based on knowledge graphs as a preferred embodiment of the electromagnetic compatibility intelligent testing and diagnostic system based on knowledge graphs.

[0055] Those skilled in the art will understand that, besides implementing the system and its various devices, modules, and units provided by this invention in the form of purely computer-readable program code, the same functions can be achieved entirely through logical programming of the method steps, enabling the system and its various devices, modules, and units to function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices, modules, and units provided by this invention can be considered a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered structures within the hardware component; alternatively, the devices, modules, and units for implementing various functions can be considered both software modules implementing the method and structures within the hardware component.

[0056] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A knowledge graph-based intelligent testing and diagnostic method for electromagnetic compatibility, characterized in that, Includes the following steps: Step 1: Obtain multi-source data; Step 2: Generate a knowledge graph in the field of electromagnetic compatibility based on multi-source data, and generate a fault chain based on the knowledge graph; Step 3: Based on the fault chain, infer the interference propagation path, and based on the interference propagation path, predict the interference intensity using the finite-difference time-domain method and the method of moments. Step 4: Calculate the impact of missing parameters in the fault chain on the fault probability, and generate a rectification plan by combining the knowledge graph, fault chain, propagation path, and interference intensity.

2. The method according to claim 1, characterized in that, The multi-source data includes structured data, unstructured data, and real-time data. Structured data includes equipment parameters, test standards, and historical test reports; unstructured data includes technical documents, fault case descriptions, and expert experience; and real-time data includes electromagnetic interference signals and environmental parameters collected by sensors.

3. The method according to claim 1 or 2, characterized in that, Step one further includes: performing time-frequency analysis on the multi-source data through wavelet transform to extract the transient features of the interference signal, and then preprocessing the multi-source data.

4. The method according to claim 3, characterized in that, The wavelet transform involves performing multi-resolution analysis of the signal using mother wavelet functions at different scales, and includes the following steps: Continuous wavelet transform: The wavelet coefficients of signal S are calculated using the cwt function, which is as follows: in, It is a scale sequence. For Daubechies 4 wavelet bases; Scale to Frequency Conversion: The scal2frq function converts the scale into the actual frequency. The scal2frq function is as follows: in, The sampling period; Time-frequency plots are drawn to locate the transient characteristics of interference signals.

5. The method according to claim 2, characterized in that, The preprocessing includes: text cleaning and standardization of terminology for unstructured data; filtering, noise reduction, and spectral analysis for electromagnetic interference signals.

6. The method according to claim 1, characterized in that, Step two includes: extracting entities, relationships, and attributes from multi-source data based on a deep learning model to generate a knowledge graph in the field of electromagnetic compatibility; then analyzing and obtaining the causal relationship between interference sources and fault phenomena based on a causal reasoning logic mechanism; and generating a fault chain by combining logical rules. Entities include interference source type, sensitive equipment model, and test standard number; relationships include causal relationships; and attributes include quantitative parameters.

7. The method according to claim 1, characterized in that, The knowledge graph-based reasoning interference propagation path includes: Construct knowledge graph nodes and edges; The shortest path is calculated using Dijkstra's algorithm, which is as follows: ， in, The edge weight; The propagation of electromagnetic fields in the time domain is simulated using the finite-difference time-domain method. By iteratively updating the electric and magnetic field components, the coupling interference between devices is calculated. Then, the electromagnetic response of the shielding material is calculated using the method of moments. Finally, the current distribution is expanded using basis functions to solve for the scattered field.

8. The method according to claim 1, characterized in that, The impact of missing calculation parameters on the failure probability in the fault chain includes: constructing a fuzzy model using membership functions, quantifying the impact of missing parameters on interference results, constructing a Bayesian network model, and weighting the conflict data with confidence.

9. The method according to claim 1, characterized in that, The generated rectification plan includes: obtaining the global optimal solution through crossover and mutation operations based on a genetic algorithm, and obtaining the local optimal solution by simulating the cooperative behavior of a particle swarm based on a particle swarm optimization algorithm.

10. A knowledge graph-based intelligent electromagnetic compatibility testing and diagnostic system, characterized in that, include: Module M1: Acquires data from multiple sources; Module M2: Generates a knowledge graph in the field of electromagnetic compatibility based on multi-source data, and generates fault chains based on the knowledge graph; Module M3: Based on knowledge graphs, it infers interference propagation paths and predicts interference intensity based on these paths using the finite difference time-domain method and the method of moments. Module M4: Calculates the impact of missing parameters on the probability of failure, and generates a rectification plan by combining knowledge graph, fault chain, propagation path, and interference intensity.

Images