Method, system, medium and product for risk assessment based on failure semantics drive

Abstract

The application discloses a risk assessment method and system based on fault semantic driving, a medium and a product, and relates to the field of power new energy equipment detection. According to the method, after a power new energy equipment detection system obtains a power equipment fault description text input by an operation and maintenance personnel, the power new energy equipment detection system matches the fault description text with fault semantic elements corresponding to nodes in a fault evolution track graph, and calculates the current risk intensity of the target node. At the same time, the power new energy equipment detection system calculates the current risk transmission value of the target node through a target directed edge to connect a downstream node according to an initiation probability corresponding to the target directed edge, and determines a current fault risk vector of the power equipment at a current time by combining an initial fault risk vector of the power equipment at an initial time, so as to determine a current risk assessment value of the power equipment. The fault diagnosis method based on risk propagation and accumulation can dynamically track and evaluate the fault diffusion process, and improves the accuracy and timeliness of fault detection.

Description

Technical Field

[0001] This application relates to the field of power and new energy equipment testing, and in particular to risk assessment methods, systems, media and products based on fault semantics. Background Technology

[0002] With the continuous expansion of the power grid and the increasing number of devices, the safe and reliable operation of power equipment faces severe challenges. Power equipment failures can lead to power outages and significant economic losses, thus requiring timely detection and handling of these faults. In actual operation, power equipment often experiences various types of faults, and these faults exhibit complex correlations and evolutionary patterns, posing significant challenges to fault detection and handling.

[0003] Currently, power equipment fault detection methods receive fault description information from operation and maintenance personnel, extract key fault entities and state feature vectors using a pre-trained semantic model, and construct a semantic retrieval topology network in conjunction with diagnostic prompt templates to perform fault diagnosis analysis, ultimately generating diagnostic conclusions and handling suggestions.

[0004] However, this semantic analysis-based power equipment fault detection method mainly relies on static semantic models and diagnostic prompt templates, failing to fully consider the propagation mechanism and risk accumulation between faults, making it prone to misjudgment or omission when facing complex fault chains. Summary of the Invention

[0005] This application provides a risk assessment method, system, medium, and product based on fault semantics to improve the accuracy and timeliness of fault detection.

[0006] Firstly, this application provides a risk assessment method based on fault semantics, applied to a power and new energy equipment detection system. The method includes: acquiring fault description text of power equipment input by maintenance personnel; determining whether the fault description text of power equipment matches the fault semantic elements corresponding to nodes in a fault evolution trajectory diagram, wherein the fault evolution trajectory diagram includes multiple nodes and multiple directed edges, nodes are connected by directed edges, different nodes correspond to different fault semantic elements, different directed edges correspond to different initiation probabilities, the fault semantic elements include equipment component state features, physical phenomenon state features, and attribute quantification state features, and the initiation probability is used to represent the probability of fault propagation between adjacent nodes; if so, based on a preset risk activation energy, the method is applied to the power... Energy is injected into the target node that matches the equipment fault description text to obtain the current risk intensity of the target node. Based on the trigger probability corresponding to the directed edge in the fault evolution trajectory graph, the current risk transmission value of the target node connected to the downstream node through the target directed edge is calculated. Based on the initial fault risk vector, current risk intensity, and current risk transmission value of the power equipment at the initial moment, the current fault risk vector of the power equipment at the current moment is obtained. Each component in the current fault risk vector is multiplied by its corresponding risk weight coefficient and summed to obtain the current risk assessment value of the power equipment. If the current risk assessment value is greater than the risk threshold, the fault propagation link is determined based on the connection relationship of the fault evolution trajectory graph, the target node, and the downstream node.

[0007] By adopting the above technical solution, the power and new energy equipment detection system, after obtaining the fault description text of the power equipment input by maintenance personnel, matches it with the fault semantic elements corresponding to the nodes in the fault evolution trajectory diagram and calculates the current risk intensity of the matched target node. Simultaneously, the system calculates the current risk transmission value of the target node connected to downstream nodes through the target directed edge by using the trigger probability corresponding to the directed edge in the fault evolution trajectory diagram. Combined with the initial fault risk vector of the power equipment at the initial moment, it determines the current fault risk vector of the power equipment at the current moment, thus determining the current risk assessment value of the power equipment. This fault diagnosis method based on risk propagation and accumulation can dynamically track and evaluate the fault propagation process. When the current risk assessment value exceeds the risk threshold, it can promptly identify the fault propagation link, thereby effectively preventing further fault spread. Compared with traditional static semantic analysis methods, it better reflects the dynamic evolution characteristics of faults and improves the accuracy and timeliness of fault detection.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, after determining whether the power equipment fault description text matches the fault semantic elements corresponding to the nodes in the fault evolution trajectory diagram, the method further includes: if not, performing semantic analysis on the power equipment fault description text based on a preset natural language processing model to obtain fault semantic features; calculating the similarity between the fault semantic features and the fault semantic elements corresponding to each node in the fault evolution trajectory diagram to determine the target node whose similarity exceeds a similarity threshold; and performing an energy injection step on the target node based on the target node.

[0009] By adopting the above technical solution, when the description text of power equipment faults cannot be directly matched with existing fault semantic elements, the power and new energy equipment detection system uses a natural language processing model to perform semantic analysis on the description text and determines the closest target node through similarity calculation. This fuzzy matching mechanism significantly improves the fault tolerance and adaptability of fault identification. Even if the description by maintenance personnel is not standardized or is ambiguous, it can accurately locate the relevant fault nodes, greatly enhancing the practicality and robustness of the fault detection method.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, after calculating the similarity between the fault semantic features and the fault semantic elements corresponding to each node in the fault evolution trajectory diagram to determine the target node whose similarity exceeds the similarity threshold, the method further includes: if the similarity between the fault semantic features and the fault semantic elements corresponding to each node in the fault evolution trajectory diagram is lower than the similarity threshold, then a temporary fault node is generated based on the power equipment fault description text, and the temporary fault semantic elements corresponding to the temporary fault node are determined; the semantic correlation degree between the temporary fault semantic elements and the fault semantic elements corresponding to each node in the fault evolution trajectory diagram is calculated, and the existing nodes whose semantic correlation degree exceeds the semantic correlation degree threshold are determined as potential upstream nodes and potential downstream nodes of the temporary fault node; temporary directed edges are established from potential upstream nodes to temporary fault nodes and from temporary fault nodes to potential downstream nodes, and temporary triggering probabilities are set for the temporary directed edges according to the semantic correlation degree; the temporary fault nodes, temporary directed edges, and temporary triggering probabilities are output as fault evolution trajectory diagram expansion suggestions for operation and maintenance personnel to optimize the fault evolution trajectory diagram.

[0011] By adopting the above technical solution, a dynamic expansion mechanism for the fault knowledge base is introduced to address the novel phenomenon of power equipment fault description. When the similarity between the fault semantic features and the fault semantic elements corresponding to each node in the fault evolution trajectory diagram is lower than the similarity threshold, the power and new energy equipment detection system generates a temporary fault node. It then determines the upstream and downstream relationship between the temporary fault node and existing nodes by calculating the semantic correlation between the temporary fault semantic elements corresponding to the temporary fault node and the fault semantic elements corresponding to each node in the fault evolution trajectory diagram. This adaptive learning mechanism enables the power and new energy equipment detection system to continuously generate fault evolution trajectory diagram expansion suggestions, helping operation and maintenance personnel improve the fault evolution trajectory diagram, enhancing the identification and prediction capabilities of novel power equipment fault phenomena, and strengthening the scalability and practical value of the power and new energy equipment detection system.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, based on a preset risk activation energy, energy injection is performed on target nodes that match the fault description text of power equipment. Specifically, this includes: statistically analyzing the historical fault description frequency and historical fault level corresponding to the target node; determining the risk activation energy adjustment coefficient of the target node based on the historical fault description frequency and historical fault level; multiplying the risk activation energy by the risk activation energy adjustment coefficient to obtain the adjusted risk activation energy; and performing energy injection on the target node based on the adjusted risk activation energy.

[0013] By adopting the above technical solution, the power new energy equipment detection system obtains the historical fault description frequency and historical fault level of the target node, and calculates the risk activation energy adjustment coefficient accordingly, thereby obtaining a more reasonable energy injection value. This adaptive adjustment mechanism based on historical data can distinguish the severity and frequency of different faults, avoiding the problems of over-warning or under-warning that may be caused by using fixed energy values. This makes the risk assessment more in line with the actual operating conditions and improves the accuracy and reliability of fault detection.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, before the step of obtaining the power equipment fault description text input by the maintenance personnel, the method further includes: if the power equipment fault description text input by the maintenance personnel is not received within a preset time period, based on a preset time decay factor, performing attenuation processing on each component in the initial fault risk vector of the power equipment at the initial moment to obtain the attenuated fault risk vector.

[0015] By adopting the above technical solution, even without obvious fault symptoms or reports, power equipment will naturally age and degrade in performance over time. If the power and new energy equipment detection system does not receive a fault description text input by maintenance personnel within a preset time period, it will automatically trigger a time decay mechanism to attenuate the initial fault risk vector of the power equipment, reflecting the deterioration risk of the power equipment over time. This proactive risk assessment mechanism breaks through the limitations of traditional methods that rely solely on fault descriptions for risk assessment. It is more in line with the actual aging patterns of equipment. By introducing a time decay factor, it can simulate the potential risk growth of power equipment during periods without fault reports, providing an important basis for preventive maintenance. It effectively avoids potential safety hazards caused by ignoring the natural aging of equipment, making risk assessment more comprehensive and objective, and improving the foresight and scientific nature of equipment management.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, based on a preset time decay factor, the components in the initial fault risk vector of the power equipment at the initial moment are attenuated. Specifically, this includes: obtaining the component operating load level of the node corresponding to each component in the initial fault risk vector of the power equipment at the initial moment; determining the time decay coefficient corresponding to each component based on the component operating load level; obtaining the adjusted time decay factor based on the time decay factor and the corresponding time decay coefficient; and attenuating each component in the initial fault risk vector of the power equipment at the initial moment based on the adjusted time decay factor.

[0017] By adopting the above technical solution, since different component operating load levels lead to significant differences in aging rates, the power new energy equipment detection system determines the time decay coefficient corresponding to each component based on the component's operating load level. For example, components operating under high load conditions may age faster, and their time decay coefficient is relatively high; while components operating under low load conditions may age more slowly, and a lower time decay coefficient can be used. This adaptive decay mechanism based on actual load overcomes the shortcomings of traditional fixed decay coefficients, which cannot reflect the differences in equipment aging under different operating conditions. This makes the risk assessment results more consistent with engineering reality and significantly improves the accuracy and reliability of risk assessment.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, after the step of determining the fault propagation link based on the connection relationship, target node, and downstream node of the fault evolution trajectory diagram if the current risk assessment value is greater than the risk threshold, the method further includes: generating an equipment monitoring plan according to the fault propagation link, the equipment monitoring plan including the monitoring cycle and monitoring method; acquiring equipment operation data collected according to the equipment monitoring plan, inputting the equipment operation data into a preset equipment status assessment model to obtain an equipment health status score; if the equipment health status score is lower than the equipment health status score threshold, extracting the changing trend of abnormal parameters in the equipment operation data, and predicting the development trend of abnormal parameters based on a preset abnormal trend evolution model; matching the corresponding fault handling solution from a preset fault handling knowledge base according to the development trend of abnormal parameters, and generating operation and maintenance guidance suggestions based on the fault handling solution.

[0019] By adopting the above technical solution, when the current risk assessment value exceeds the risk threshold, the power and new energy equipment detection system generates a targeted equipment monitoring plan, collects equipment operation data, and assesses the equipment health status score. When the equipment health status score is found to be below the equipment health status score threshold, the power and new energy equipment detection system uses an anomaly trend evolution model to predict the development trend of abnormal parameters and matches corresponding fault handling solutions from the fault handling knowledge base. This closed-loop mechanism not only achieves timely risk detection and handling but also provides specific solutions through operation and maintenance guidance suggestions, forming a complete chain from risk detection to problem resolution. This significantly improves the efficiency and effectiveness of equipment maintenance and reduces losses caused by equipment failures.

[0020] In a second aspect, embodiments of this application provide a power new energy equipment testing system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, and the memory is used to store computer program code, which includes computer instructions, and the one or more processors call the computer instructions to cause the power new energy equipment testing system to perform the method described in the first aspect and any possible implementation thereof.

[0021] Thirdly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a power and new energy equipment testing system, cause the power and new energy equipment testing system to execute the method described in the first aspect and any possible implementation thereof.

[0022] Fourthly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a power and new energy equipment testing system, cause the power and new energy equipment testing system to perform the method described in the first aspect and any possible implementation thereof.

[0023] Understandably, the power new energy equipment testing system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the methods provided in the embodiments of this application. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods, and will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages:

[0025] 1. By adopting the above technical solution, the power and new energy equipment detection system, after obtaining the fault description text of the power equipment input by maintenance personnel, matches it with the fault semantic elements corresponding to the nodes in the fault evolution trajectory diagram and calculates the current risk intensity of the matched target node. Simultaneously, the system calculates the current risk transmission value of the target node connected to downstream nodes through the target directed edge by using the trigger probability corresponding to the directed edge in the fault evolution trajectory diagram. Combined with the initial fault risk vector of the power equipment at the initial moment, it determines the current fault risk vector of the power equipment at the current moment, thus determining the current risk assessment value of the power equipment. This fault diagnosis method based on risk propagation and accumulation can dynamically track and evaluate the fault propagation process. When the current risk assessment value exceeds the risk threshold, it can promptly identify the fault propagation link, thereby effectively preventing further fault spread. Compared with traditional static semantic analysis methods, it better reflects the dynamic evolution characteristics of faults and improves the accuracy and timeliness of fault detection.

[0026] 2. By adopting the above technical solution, a dynamic expansion mechanism for the fault knowledge base is introduced to address the novel phenomenon of power equipment fault description. When the similarity between the fault semantic features and the fault semantic elements corresponding to each node in the fault evolution trajectory diagram is lower than the similarity threshold, the power and new energy equipment detection system generates a temporary fault node. It then calculates the semantic correlation between the temporary fault semantic elements corresponding to the temporary fault node and the fault semantic elements corresponding to each node in the fault evolution trajectory diagram to determine the upstream and downstream relationship between the temporary fault node and existing nodes. This adaptive learning mechanism enables the power and new energy equipment detection system to continuously generate fault evolution trajectory diagram expansion suggestions, helping maintenance personnel improve the fault evolution trajectory diagram, enhancing the identification and prediction capabilities of novel power equipment fault phenomena, and strengthening the scalability and practical value of the power and new energy equipment detection system.

[0027] 3. By adopting the above technical solution, even without obvious fault symptoms or reports, power equipment will naturally age and degrade in performance over time. If the power and new energy equipment detection system does not receive a fault description text input by maintenance personnel within a preset time period, it will automatically trigger a time decay mechanism to attenuate the initial fault risk vector of the power equipment, reflecting the deterioration risk of the power equipment over time. This proactive risk assessment mechanism breaks through the limitations of traditional methods that rely solely on fault descriptions for risk assessment. It is more in line with the actual aging patterns of equipment. By introducing a time decay factor, it can simulate the potential risk growth of power equipment during periods without fault reports, providing an important basis for preventive maintenance. It effectively avoids potential safety hazards caused by ignoring the natural aging of equipment, making risk assessment more comprehensive and objective, and improving the foresight and scientific nature of equipment management. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating a fault semantic-driven risk assessment method in an embodiment of this application.

[0029] Figure 2 This is another flowchart illustrating the fault semantic-driven risk assessment method in this application embodiment;

[0030] Figure 3 This is a schematic diagram of the physical device structure of a power new energy equipment testing system in the embodiments of this application. Detailed Implementation

[0031] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification of this application, the singular expressions “a,” “an,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0032] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0033] The following describes the process of the method provided in this implementation, based on the above scenario. Please refer to... Figure 1 This is a flowchart illustrating a risk assessment method based on fault semantics in this application.

[0034] S101. Obtain the power equipment fault description text input by the maintenance personnel;

[0035] Maintenance and operation personnel (O&M) are technical personnel responsible for the operation and maintenance of power equipment, including on-site inspection personnel and equipment repair personnel. Power equipment refers to various devices used in power systems for power generation, transmission, transformation, and distribution, such as transformers, switchgear, and generator sets. Power equipment fault description text refers to the written description by O&M personnel of abnormal phenomena or fault states of power equipment, typically including information such as the time of fault occurrence, location of fault occurrence, and fault symptoms.

[0036] When abnormalities or faults occur during the operation of power equipment, maintenance personnel need to promptly record the observed phenomena. Specifically, the power and new energy equipment detection system receives fault description text input by maintenance personnel through a human-machine interface. This fault description text can be a structured fault report or an unstructured natural language description. The power and new energy equipment detection system saves the input timestamp of the fault description text for subsequent time-series analysis.

[0037] The following is a typical example of a fault description text for electrical equipment:

[0038] Fault report number: F202509130921;

[0039] Report time: September 13, 2025, 09:21:35;

[0040] Equipment type: 220kV main transformer;

[0041] Equipment number: MT-B12-103;

[0042] Installation location: Bilibili's No. 12 main transformer;

[0043] Fault Description: During night shift inspection, a faint "buzzing" sound was detected at the high-voltage bushing of the main transformer using an ultrasonic detector, with a frequency in the range of approximately 40-50kHz. Further inspection revealed that the ultrasonic waveform exhibited typical partial discharge characteristics, with a noise intensity of 35dB (normal value should be <30dB), and the discharge duration was intermittent, lasting approximately 0.5-1 second each time.

[0044] Oil chromatography analysis results:

[0045] Acetylene content: 2.8 ppm (normal value should be <2 ppm);

[0046] Total hydrocarbons: 15 ppm;

[0047] Oil temperature: 42℃;

[0048] Presenter: Mr. Li;

[0049] Contact information: xxx-xxxx-xxxx;

[0050] On-site handling measures: Monitoring frequency has been increased to once every 2 hours.

[0051] S102. Determine whether the description text of the power equipment fault matches the fault semantic elements corresponding to the nodes in the fault evolution trajectory diagram. The fault evolution trajectory diagram includes multiple nodes and multiple directed edges. The nodes are connected by directed edges. Different nodes correspond to different fault semantic elements, and different directed edges correspond to different initiation probabilities. The fault semantic elements include equipment component state features, physical phenomenon state features, and attribute quantification state features. The initiation probability is used to represent the probability of the fault propagating between adjacent nodes.

[0052] The fault evolution trajectory graph is a directed graph model describing the fault evolution process of power equipment. Nodes represent different states in the fault evolution process of power equipment. Directed edges represent the transition relationships between different states in the fault evolution process of power equipment. Fault semantic elements refer to the semantic description of fault states, including equipment component state characteristics (such as "bearing overheating"), physical phenomenon state characteristics (such as "abnormal vibration"), and attribute-quantified state characteristics (such as "temperature exceeds 80℃"). Trigger probability is used to represent the probability that one fault state will lead to another fault state.

[0053] Taking the fault evolution trajectory diagram of a certain type of transformer as an example:

[0054] Fault semantic elements (nodes):

[0055] M1: (High-voltage bushing, partial discharge, abnormal acoustic signature - slight);

[0056] M2: (Insulating oil, gas chromatography, acetylene content - slightly above standard);

[0057] M3: (Insulating paper, decreased degree of polymerization, moderate deterioration);

[0058] M4: (Winding, inter-turn short circuit, fault - severe);

[0059] Evolutionary relationships and their probabilities (directed edges):

[0060] M1→M2: The probability of triggering this is 0.7. The reason is that long-term slight partial discharge of the high-voltage bushing will cause the insulating oil to decompose and produce acetylene gas, which is highly correlated.

[0061] M2→M3: The probability of triggering this is 0.8. The reason is that excessive acetylene content in the insulating oil will accelerate the aging of the insulating paper, and the correlation is extremely strong.

[0062] M3→M4: The probability of triggering the circuit is 0.9. The reason is that the deterioration of the insulation paper is the core precursor to the short circuit between winding turns, and it will almost inevitably occur.

[0063] The fault evolution trajectory diagram shows the complete evolution path of a certain type of transformer from the initial fault symptoms (partial discharge of high-voltage bushing) to the final serious fault (inter-turn short circuit in winding). Each node corresponds to a standardized fault semantic element, and the directed edges between nodes quantify the probability relationship of fault propagation.

[0064] After obtaining the fault description text of the power equipment, the power and new energy equipment detection system needs to match it with known fault patterns (nodes in the fault evolution trajectory diagram). Specifically, firstly, the system performs semantic parsing on the fault description text to obtain the parsing results. Then, the system compares the parsing results with the fault semantic elements corresponding to each node in the fault evolution trajectory diagram to determine if there are any target nodes with consistent semantic expressions. The matching process considers semantic variations such as synonyms and near-synonyms, and fuzzy matching strategies can be used to improve the accuracy of the matching.

[0065] Following the previous example and the example in step S101, the matching analysis process is as follows:

[0066] (1) Semantic parsing results:

[0067] Equipment component: High-voltage bushing;

[0068] Physical phenomena: partial discharge, abnormal ultrasound;

[0069] Attribute quantification: Noise intensity 35dB, acetylene content 2.8ppm;

[0070] (2) Matching with nodes in the fault evolution trajectory diagram:

[0071] High matching degree with node M1 (high voltage bushing, partial discharge, acoustic abnormality - slight);

[0072] Equipment components are fully matched: high-pressure bushing;

[0073] The physical phenomena match perfectly: partial discharge;

[0074] Attribute characteristics match: noise anomaly is slight;

[0075] Partially matched with node M2 ​​(insulating oil, gas chromatography, acetylene content - trace excess);

[0076] Acetylene content of 2.8 ppm was detected, exceeding the normal value of 2 ppm;

[0077] It is in the early stages of failure evolution;

[0078] (3) Matching conclusion: The power equipment fault description text is highly matched with node M1, and early features of node M2 ​​appear.

[0079] S103. If so, based on the preset risk activation energy, energy is injected into the target node that matches the power equipment fault description text to obtain the current risk intensity of the target node. According to the triggering probability corresponding to the directed edge in the fault evolution trajectory diagram, the current risk transmission value of the target node connected to the downstream node through the target directed edge is calculated.

[0080] Here, risk activation energy refers to the energy value used to quantify the degree of fault risk. A target node is a node in the fault evolution trajectory diagram that matches the fault description text of the power equipment. Current risk intensity represents the current risk level of the target node. Current risk propagation value represents the energy value of risk propagation from the target node to downstream nodes. A downstream node is a successor node connected to the target node via a directed edge.

[0081] When the fault description text of the power equipment is confirmed to match the target node, the power and new energy equipment detection system needs to assess the risk level of the target node and its propagation impact. Specifically, the system injects energy into the target node according to a preset risk activation energy and calculates the current risk intensity of the target node. Then, based on the topology in the fault evolution trajectory graph and the triggering probability on the directed edges, the system calculates the current risk propagation value from the target node to downstream nodes. This risk propagation mechanism based on energy diffusion can simulate the chain reaction effect of a fault.

[0082] Taking the transformer failure example above, the process of risk energy injection and risk propagation calculation in step S103 is explained:

[0083] (1) Risk energy injection: Assuming the preset risk activation energy is 10, then the current risk intensity of M1 is 10;

[0084] (2) Risk propagation calculation: The probability of triggering from M1 to M2 is 0.7, so the current risk propagation value from M1 to M2 is 10 × 0.7 = 7. The probability of triggering from M2 to M3 is 0.8, so the current risk propagation value from M2 to M3 is 7 × 0.8 = 5.6, and so on. This will not be elaborated further here...

[0085] It should be noted that the preset risk activation energy can be divided into different levels. For example, the risk activation energy corresponding to minor risk is 50 energy units, that of medium risk is 75 energy units, and that of severe risk is 100 energy units. The formulas for calculating risk energy injection and risk propagation can be: Current risk intensity = Base risk energy × (Current parameter value / Threshold parameter value); Current risk propagation value = Current risk intensity × Trigger probability. The specific settings for risk energy injection and risk propagation calculations can be flexibly adjusted according to actual conditions and are not limited here.

[0086] Optionally, under normal circumstances, energy injection into target nodes that match the fault description text of power equipment, based on a preset risk activation energy, can be achieved in the following ways, without limitation: statistically analyze the frequency and level of historical fault descriptions corresponding to the target node; determine the risk activation energy adjustment coefficient of the target node based on the frequency and level of historical fault descriptions; multiply the risk activation energy by the risk activation energy adjustment coefficient to obtain the adjusted risk activation energy; and inject energy into the target node based on the adjusted risk activation energy.

[0087] S104. Based on the initial fault risk vector, current risk intensity, and current risk propagation value of the power equipment at the initial moment, obtain the current fault risk vector of the power equipment at the current moment.

[0088] The initial time refers to the point in time when risk assessment begins. The initial fault risk vector is a vector describing the risk level of each node of the power equipment at the initial time, and its dimension is equal to the number of nodes in the fault evolution trajectory diagram. The current time refers to the point in time when the power equipment fault description text is received. The current fault risk vector is a vector describing the risk level of each node of the power equipment at the current time, and its dimension is equal to the number of nodes in the fault evolution trajectory diagram, which can describe the overall fault risk status of the power equipment.

[0089] After calculating the risk intensity and assessing the risk propagation of the target node, the power and new energy equipment detection system needs to update the overall risk status of the power equipment. Specifically, the system obtains the initial fault risk vector of the power equipment. This initial fault risk vector records the basic risk level of each node of the power equipment, typically [M1, M2, ..., Mn], corresponding to [0, 0, ..., 0]. Then, the system updates the current risk intensity of the target node to the position of the corresponding component and adjusts the components of the relevant downstream nodes according to the current risk propagation value. In this way, the system obtains the current fault risk vector reflecting the overall risk status of the power equipment at the current moment, achieving dynamic updating of the risk status.

[0090] S105. Multiply each component in the current fault risk vector by its corresponding risk weight coefficient and sum them to obtain the current risk assessment value of the power equipment.

[0091] Among them, the risk weight coefficient represents the degree of impact of different nodes on the overall risk of power equipment, and is usually determined by expert experience or obtained through historical data analysis. The current risk assessment value is a scalar value that characterizes the overall risk level of power equipment at the current moment, and is used for risk warning and decision-making. Multiplication refers to multiplying the components in the current fault risk vector with their corresponding risk weight coefficients.

[0092] After obtaining the current fault risk vector, the multi-dimensional risk information needs to be integrated into a single evaluation index. Specifically, the power and new energy equipment detection system assigns a corresponding risk weight coefficient to each component of the current fault risk vector. This risk weight coefficient reflects the severity and importance of that type of fault. The power and new energy equipment detection system then sums the weighted components to obtain a comprehensive current risk assessment value. This weighted summation method can balance the impact of different types of faults, providing a more accurate risk assessment result.

[0093] S106. If the current risk assessment value is greater than the risk threshold, the fault propagation link is determined based on the connection relationship of the fault evolution trajectory diagram, the target node, and the downstream node.

[0094] Here, the risk threshold represents a pre-set risk warning level used to trigger a risk alert. The fault propagation link refers to the risk propagation path formed from the target node through a series of connected nodes. The connection relationship refers to the topological structure between nodes in the fault evolution trajectory diagram. There is a direct or indirect causal relationship between the target node and downstream nodes.

[0095] When the power and new energy equipment detection system detects that the current risk assessment value exceeds the risk threshold, it is necessary to analyze potential risk propagation paths. Specifically, starting from the target node, the system identifies potentially affected downstream node sequences along the directed edges of the fault evolution trajectory graph. Using a depth-first search algorithm, the system finds paths with a high probability of risk propagation, forming a complete fault propagation link. This link analysis helps maintenance personnel understand the fault propagation trend, take timely and targeted preventative measures, and effectively control the fault scope.

[0096] By adopting the above technical solution, the power and new energy equipment detection system, after obtaining the fault description text of the power equipment input by maintenance personnel, matches it with the fault semantic elements corresponding to the nodes in the fault evolution trajectory diagram and calculates the current risk intensity of the matched target node. Simultaneously, the system calculates the current risk transmission value of the target node connected to downstream nodes through the target directed edge by using the trigger probability corresponding to the directed edge in the fault evolution trajectory diagram. Combined with the initial fault risk vector of the power equipment at the initial moment, it determines the current fault risk vector of the power equipment at the current moment, thus determining the current risk assessment value of the power equipment. This fault diagnosis method based on risk propagation and accumulation can dynamically track and evaluate the fault propagation process. When the current risk assessment value exceeds the risk threshold, it can promptly identify the fault propagation link, thereby effectively preventing further fault spread. Compared with traditional static semantic analysis methods, it better reflects the dynamic evolution characteristics of faults and improves the accuracy and timeliness of fault detection.

[0097] The following provides a more detailed description of the process of the method provided in this implementation. Please refer to [link / reference]. Figure 2 This is another flowchart illustrating the risk assessment method based on fault semantics in this application.

[0098] S201. If no power equipment fault description text input by the maintenance personnel is received within the preset time period, based on the preset time decay factor, the components in the initial fault risk vector of the power equipment at the initial moment are attenuated to obtain the attenuated fault risk vector.

[0099] The preset duration refers to the maximum time interval for waiting to obtain the fault description text of the power equipment, typically several days. The time decay factor is a coefficient used to describe the natural degradation of power equipment performance over time; it is generally a positive number less than 1. Attenuation processing refers to the process of calculating the performance degradation of the power equipment. The attenuated fault risk vector refers to the new risk state vector of the power equipment obtained after time decay processing.

[0100] During normal operation of power equipment, even without obvious faults, its performance will gradually degrade over time. Specifically, the power and new energy equipment detection system continuously monitors the input of power equipment fault description texts. If no new power equipment fault description text is received after a preset time (e.g., 72 hours), the system will activate a mechanism to allow the power equipment's performance to naturally degrade over time. Assuming the initial fault risk vector of a power equipment is [0.2, 0.3, 0.1, 0.4] and the time decay factor is 0.95, the system will calculate the decay of each component in the initial fault risk vector: new component value = original component value × (1 / time decay factor), resulting in the decayed fault risk vector [0.21, 0.316, 0.105, 0.421]. This processing mechanism considers the natural aging process of power equipment, reflecting the cumulative effect of risk even during periods without fault reports, making risk assessment more realistic. The power and new energy equipment testing system will set different time decay factors according to the characteristics of different types of equipment. For example, a larger time decay factor can be used for key equipment operating under high load to reflect its faster degradation rate.

[0101] Optionally, under normal circumstances, the attenuation processing of each component in the initial fault risk vector of the power equipment at the initial moment can be achieved by the following method based on the preset time attenuation factor, which is not limited here: obtain the component operating load level of the node corresponding to each component in the initial fault risk vector of the power equipment at the initial moment; determine the time attenuation coefficient corresponding to each component according to the component operating load level; obtain the adjusted time attenuation factor based on the time attenuation factor and the corresponding time attenuation coefficient; and perform attenuation processing on each component in the initial fault risk vector of the power equipment at the initial moment based on the adjusted time attenuation factor.

[0102] S202. Obtain the power equipment fault description text input by the operation and maintenance personnel.

[0103] For details, please refer to step S101, which will not be repeated here.

[0104] S203. Determine whether the fault description text of the power equipment matches the fault semantic elements corresponding to the nodes in the fault evolution trajectory diagram. The fault evolution trajectory diagram includes multiple nodes and multiple directed edges. The nodes are connected by directed edges. Different nodes correspond to different fault semantic elements, and different directed edges correspond to different initiation probabilities. The fault semantic elements include equipment component state features, physical phenomenon state features, and attribute quantification state features. The initiation probability is used to represent the probability of the fault propagating between adjacent nodes.

[0105] For details, please refer to step S102, which will not be repeated here.

[0106] S204. If so, based on the preset risk activation energy, energy is injected into the target node that matches the power equipment fault description text to obtain the current risk intensity of the target node. According to the trigger probability corresponding to the directed edge in the fault evolution trajectory diagram, the current risk transmission value of the target node connected to the downstream node through the target directed edge is calculated.

[0107] For details, please refer to step S103, which will not be repeated here.

[0108] S205. Based on the initial fault risk vector, current risk intensity, and current risk propagation value of the power equipment at the initial moment, obtain the current fault risk vector of the power equipment at the current moment.

[0109] For details, please refer to step S104, which will not be repeated here.

[0110] S206. Multiply each component in the current fault risk vector by its corresponding risk weight coefficient and sum them to obtain the current risk assessment value of the power equipment.

[0111] For details, please refer to step S105, which will not be repeated here.

[0112] S207. If the current risk assessment value is greater than the risk threshold, the fault propagation link is determined based on the connection relationship of the fault evolution trajectory diagram, the target node, and the downstream node.

[0113] For details, please refer to step S106, which will not be repeated here.

[0114] S208. Based on the fault propagation path, generate an equipment monitoring plan, which includes the monitoring cycle and monitoring method.

[0115] The equipment monitoring plan refers to a monitoring scheme developed for a specific fault propagation path. The monitoring cycle refers to the time interval for inspecting power equipment, which can be a fixed cycle or a dynamically adjusted cycle. The monitoring method refers to the specific detection means and methods, including various forms such as online monitoring, offline detection, and manual inspection.

[0116] After identifying the fault propagation path, the power and new energy equipment detection system needs to formulate corresponding monitoring strategies to track fault development. Specifically, the system analyzes the characteristics of each node in the fault propagation path and then determines the monitoring priority based on the risk level and detection difficulty of different nodes. For high-risk nodes (such as inter-turn short circuits in transformer windings), the system may set shorter monitoring cycles (e.g., every 4 hours) and more refined monitoring methods (e.g., online partial discharge monitoring); for low-risk nodes, the system may use longer monitoring cycles (e.g., every 24 hours) and conventional monitoring methods (e.g., infrared thermography). The system automatically generates a complete equipment monitoring plan that includes specific monitoring time points, monitoring items, and execution requirements.

[0117] S209. Obtain equipment operation data collected according to the equipment monitoring plan, input the equipment operation data into the preset equipment status assessment model, and obtain the equipment health status score.

[0118] Equipment operation data refers to equipment operating parameters collected through various monitoring methods, including electrical, mechanical, and environmental parameters. Equipment condition assessment models are mathematical models used to evaluate the health status of equipment, typically built based on machine learning or expert systems. Equipment health status scores are quantitative representations of the overall operating condition of the equipment, generally using a 0-100 scoring system.

[0119] During the execution of the equipment monitoring plan, the power and new energy equipment monitoring system needs to analyze and evaluate the collected equipment operation data. Specifically, the system collects equipment operation data, including key parameters such as temperature, vibration, and partial discharge, through various sensors and testing equipment. This multi-source heterogeneous data is then input into a pre-trained equipment condition assessment model. This model comprehensively considers the weights and correlations of various indicators to calculate an equipment health status score. For example, if a transformer's oil chromatography, partial discharge, and temperature indicators are all within the normal range, it will ultimately receive a health status score of 90, indicating that the equipment is in good condition.

[0120] S210. If the equipment health status score is lower than the equipment health status score threshold, extract the changing trend of abnormal parameters in the equipment operation data, and predict the development trend of abnormal parameters based on the preset abnormal trend evolution model.

[0121] Among these, the equipment health status scoring threshold refers to the scoring limit that triggers in-depth analysis, usually specified by the equipment management procedures. Abnormal parameters refer to operating indicators that exceed the normal range. The trend of change refers to the pattern of parameter change over time. The abnormal trend evolution model is a mathematical model used to predict the future trend of abnormal parameters, which can be constructed using time series analysis or machine learning methods.

[0122] When an abnormal equipment health status score is detected, the power and new energy equipment detection system needs to perform trend analysis and prediction. Specifically, the system identifies the abnormal parameters causing the decrease in the equipment health status score, such as an abnormally high dissolved gas content in transformer oil. Then, the system extracts recent historical data for this abnormal parameter and analyzes its changing patterns, such as an exponential growth trend in gas volume. This data is then input into an abnormal trend evolution model to predict the changing trend of the abnormal parameter over a future period, providing a basis for timely intervention.

[0123] S211. Based on the development trend of abnormal parameters, match the corresponding fault handling solution from the preset fault handling knowledge base, and generate operation and maintenance guidance suggestions based on the fault handling solution.

[0124] The fault handling knowledge base refers to a database storing various fault handling experiences and solutions. Fault handling solutions refer to specific handling measures and steps for specific fault types. Maintenance guidance and suggestions refer to specific maintenance operation instructions provided by the power and new energy equipment testing system, including handling steps, precautions, and required tools.

[0125] After predicting the development trend of abnormal parameters, the power and new energy equipment detection system needs to provide specific handling suggestions. Specifically, the system matches the development trend of abnormal parameters with cases in its fault handling knowledge base to identify the fault handling solution with the highest similarity. For example, for anomalies in dissolved gas in transformer oil, the system might match a "Handling Procedure for Abnormal Gas Content in Transformer Oil," which includes a detailed handling process: first, repeat sampling for confirmation; then, determining the fault type based on the gas component ratio; and finally, determining whether shutdown for maintenance is necessary. The system will then transform these professional fault handling solutions into clear operation and maintenance guidance suggestions to help on-site personnel accurately perform maintenance work.

[0126] S212. If not, then based on the preset natural language processing model, perform semantic analysis on the power equipment fault description text to obtain fault semantic features; calculate the similarity between the fault semantic features and the fault semantic elements corresponding to each node in the fault evolution trajectory diagram to determine the target node whose similarity exceeds the similarity threshold.

[0127] Natural language processing (NLP) models refer to computer models used to understand and process human natural language, such as BERT and GPT. Semantic analysis refers to the process of parsing and understanding the meaning of power equipment fault description text. Fault semantic features refer to the key semantic information extracted from power equipment fault description text, including fault location, fault phenomena, and parameter characteristics. Similarity calculation represents a numerical operation that measures the degree of similarity between two semantic expressions. The similarity threshold is the critical value for determining the degree of semantic matching, usually set between 0.6 and 0.8.

[0128] When the description text of a power equipment fault cannot be directly matched with existing fault semantic elements, the power and new energy equipment detection system needs to perform deep semantic analysis. Specifically, the system calls a pre-trained natural language processing model to perform word segmentation, part-of-speech tagging, and entity recognition on the fault description text to extract fault semantic features. Then, the system calculates the similarity between these fault semantic features and the fault semantic elements corresponding to each node in the fault evolution trajectory graph, and identifies nodes with similarity exceeding a similarity threshold (e.g., 0.7) as potential matching targets, i.e., target nodes. This deep learning-based semantic matching method can handle language variations such as synonyms and near-synonyms, improving the robustness of fault identification.

[0129] S213. If the similarity between the fault semantic features and the fault semantic elements corresponding to each node in the fault evolution trajectory diagram is lower than the similarity threshold, then a temporary fault node is generated based on the power equipment fault description text, and the temporary fault semantic elements corresponding to the temporary fault node are determined.

[0130] Temporary fault nodes refer to nodes temporarily created by the power and new energy equipment detection system to represent novel faults. Temporary fault semantic elements represent a standardized semantic description of the temporarily created fault features. A similarity score below the similarity threshold indicates that the fault semantic features differ significantly from known fault patterns.

[0131] When the power and new energy equipment detection system finds that the similarity between the fault semantic features and all existing nodes in the fault evolution trajectory diagram is below the similarity threshold, it indicates that a new type of fault may have occurred. Specifically, the power and new energy equipment detection system creates a temporary fault node and constructs a standardized temporary fault semantic element based on the original power equipment fault description text. This temporary fault semantic element also contains information in three dimensions: equipment component state features, physical phenomenon state features, and attribute quantification state features. For example, if a new fault description is "abnormal SF6 gas pressure fluctuation in switchgear, fluctuation amplitude exceeding 0.2MPa", the power and new energy equipment detection system will generate the temporary fault semantic element as (switchgear, SF6 pressure fluctuation, amplitude > 0.2MPa).

[0132] S214. Calculate the semantic correlation degree between the temporary fault semantic element and the fault semantic element corresponding to each node in the fault evolution trajectory diagram, and determine the existing nodes whose semantic correlation degree exceeds the semantic correlation degree threshold as potential upstream nodes and potential downstream nodes of the temporary fault node.

[0133] Semantic relevance refers to the degree to which there is a causal or evolutionary relationship at the semantic level between the semantic elements of a temporary fault and the corresponding fault semantic elements of each node in the fault evolution trajectory graph. Existing nodes represent existing fault nodes in the fault evolution trajectory graph. Potential upstream nodes are preceding nodes that may lead to the current fault. Potential downstream nodes are subsequent nodes that may be triggered by the current fault. The semantic relevance threshold is used to represent the critical value for determining whether there is a significant correlation between nodes.

[0134] After identifying temporary fault nodes, the power and new energy equipment detection system needs to integrate them into the existing fault evolution trajectory map. Specifically, the system identifies potential upstream and downstream relationships by calculating the semantic correlation between the semantic elements of the temporary fault and the corresponding fault semantic elements of each existing node. This semantic correlation calculation considers not only semantic similarity but also physical causal relationships. For example, "insulation aging" and "partial discharge" have a high semantic correlation because the former often leads to the latter. The system marks existing nodes with a semantic correlation exceeding a threshold (e.g., 0.5) as potential upstream or downstream nodes of the temporary fault node. This method can infer the position of the new fault in the entire fault evolution chain, providing a basis for subsequent risk propagation analysis.

[0135] S215. Establish temporary directed edges from potential upstream nodes to temporary fault nodes and from temporary fault nodes to potential downstream nodes, and set temporary trigger probabilities for the temporary directed edges based on semantic relevance.

[0136] Temporary directed edges refer to fault propagation paths temporarily established in the fault evolution trajectory graph. The direction indicates the directionality of fault propagation, reflecting the fault evolution sequence. The temporary trigger probability refers to the probability of fault propagation estimated by the power new energy equipment detection system based on semantic correlation, with a value ranging from 0 to 1.

[0137] After identifying the potential upstream and downstream nodes of the temporary fault node, the power and new energy equipment detection system needs to establish a complete propagation relationship. Specifically, the system establishes temporary directed edges from potential upstream nodes to the temporary fault node and from the temporary fault node to potential downstream nodes in the fault evolution trajectory graph. Then, using semantic correlation as a basis and combining expert experience rules, the system calculates the temporary trigger probability for each temporary directed edge. For example, if the semantic correlation between two nodes is 0.8 and they conform to a physical causal relationship, a higher temporary trigger probability, such as 0.7, may be set; if the semantic correlation between two nodes is 0.5, a lower temporary trigger probability, such as 0.3, may be set.

[0138] S216. Output temporary fault nodes, temporary directed edges, and temporary trigger probabilities as suggestions for expanding the fault evolution trajectory diagram, so that operation and maintenance personnel can optimize the fault evolution trajectory diagram.

[0139] Among them, the fault evolution trajectory diagram extension suggestion refers to the modification plan generated by the power and new energy equipment detection system to improve the existing fault evolution trajectory diagram. Optimization refers to the process by which operation and maintenance personnel review and adjust the system suggestions based on practical experience.

[0140] After constructing the temporary fault nodes and their relationships, the power and new energy equipment detection system needs to submit the analysis results to the operation and maintenance personnel for review. Specifically, the system organizes newly discovered fault modes and their evolutionary relationships into structured fault evolution trajectory diagram extension suggestions, including: detailed definitions of temporary fault nodes, their relationships with existing nodes, and the probability of triggering the suggestions. The system displays this information through a visual interface and marks high-risk propagation paths that require special attention. Operation and maintenance personnel can evaluate the fault evolution trajectory diagram extension suggestions based on their actual operation and maintenance experience, confirm or adjust relevant parameters, and ultimately decide whether to permanently integrate these new contents into the fault evolution trajectory diagram.

[0141] S217. Based on the target node, perform the step of injecting energy into the target node.

[0142] Here, the target node refers to the identified fault-matching node, which may be an existing fault node or a newly created temporary fault node. Energy injection refers to the process of applying preset risk activation energy to the target node.

[0143] After confirming the faulty node, the power and new energy equipment detection system needs to assess its risk level. Specifically, the power and new energy equipment detection system performs an energy injection operation on the identified target node, as described in steps S103 and subsequent steps.

[0144] The following describes the power and new energy equipment detection system in the embodiments of this invention from the perspective of hardware processing. Please refer to [link / reference]. Figure 3 This is a schematic diagram of the physical device structure of a power new energy equipment testing system in this application embodiment.

[0145] It should be noted that, Figure 3 The structure of the power new energy equipment testing system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0146] like Figure 3 As shown, the power new energy equipment detection system includes a CPU 301, which can perform various appropriate actions and processes according to a program stored in the read-only memory ROM 302 or a program loaded from the storage section 308 into the random access memory RAM 303, such as executing the methods described in the above embodiments. The RAM 303 also stores various programs and data required for system operation. The CPU 301, ROM 302, and RAM 303 are interconnected via a bus 304. An I / O interface 305 is also connected to the bus 304.

[0147] The following components are connected to I / O interface 305: input section 306 including audio input devices, push-button switches, etc.; output section 307 including a liquid crystal display (LCD) and audio output devices, indicator lights, etc.; storage section 308 including a hard disk, etc.; and communication section 309 including a network interface card such as a LAN (Local Area Network) card, modem, etc. Communication section 309 performs communication processing via a network such as the Internet. Drive 310 is also connected to I / O interface 305 as needed. Removable media 311, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., are installed on drive 310 as needed so that computer programs read from them can be installed into storage section 308 as needed.

[0148] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by CPU 301, it performs the various functions defined in the present invention.

[0149] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0150] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0151] Specifically, the power new energy equipment detection system of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, it implements the risk assessment method based on fault semantics provided in the above embodiment.

[0152] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the power and new energy equipment detection system described in the above embodiments; or it may exist independently and not assembled into the power and new energy equipment detection system. The storage medium carries one or more computer programs, which, when executed by a processor of the power and new energy equipment detection system, enable the power and new energy equipment detection system to implement the fault semantic-driven risk assessment method provided in the above embodiments.

[0153] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0154] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

[0155] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM or random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A risk assessment method based on fault semantics, characterized in that, The method, applied to a power and new energy equipment testing system, includes: Obtain the power equipment fault description text input by maintenance personnel; Determine whether the power equipment fault description text matches the fault semantic elements corresponding to the nodes in the fault evolution trajectory diagram. The fault evolution trajectory diagram includes multiple nodes and multiple directed edges. The nodes are connected by the directed edges. Different nodes correspond to different fault semantic elements, and different directed edges correspond to different initiation probabilities. The fault semantic elements include equipment component state features, physical phenomenon state features, and attribute quantification state features. The initiation probability is used to represent the probability of fault propagation between adjacent nodes. If so, based on the preset risk activation energy, energy is injected into the target node that matches the power equipment fault description text to obtain the current risk intensity of the target node. According to the trigger probability corresponding to the directed edge in the fault evolution trajectory diagram, the current risk transmission value of the target node connected to the downstream node through the target directed edge is calculated. Based on the initial fault risk vector of the power equipment at the initial moment, the current risk intensity, and the current risk propagation value, the current fault risk vector of the power equipment at the current moment is obtained. The current risk assessment value of the power equipment is obtained by multiplying each component of the current fault risk vector by its corresponding risk weight coefficient and summing the results. If the current risk assessment value is greater than the risk threshold, then the fault propagation link is determined based on the connection relationship of the fault evolution trajectory diagram, the target node, and the downstream node.

2. The method according to claim 1, characterized in that, After the step of determining whether the power equipment fault description text matches the fault semantic elements corresponding to the nodes in the fault evolution trajectory diagram, the method further includes: If not, then based on the preset natural language processing model, semantic analysis is performed on the power equipment fault description text to obtain fault semantic features; The similarity between the fault semantic features and the fault semantic elements corresponding to each node in the fault evolution trajectory diagram is calculated to determine the target node whose similarity exceeds the similarity threshold. Based on the target node, perform the step of injecting energy into the target node.

3. The method according to claim 2, characterized in that, After the step of calculating the similarity between the fault semantic features and the fault semantic elements corresponding to each node in the fault evolution trajectory graph to determine the target node whose similarity exceeds a similarity threshold, the method further includes: If the similarity between the fault semantic features and the fault semantic elements corresponding to each node in the fault evolution trajectory diagram is lower than the similarity threshold, then a temporary fault node is generated based on the power equipment fault description text, and the temporary fault semantic element corresponding to the temporary fault node is determined. Calculate the semantic correlation degree between the temporary fault semantic element and the fault semantic element corresponding to each node in the fault evolution trajectory diagram, and determine the existing nodes whose semantic correlation degree exceeds the semantic correlation degree threshold as potential upstream nodes and potential downstream nodes of the temporary fault node. Establish temporary directed edges from the potential upstream node to the temporary fault node and from the temporary fault node to the potential downstream node, and set temporary triggering probabilities for the temporary directed edges according to the semantic correlation degree; The temporary fault node, the temporary directed edge, and the temporary trigger probability are output as a fault evolution trajectory graph expansion suggestion, so that operation and maintenance personnel can optimize the fault evolution trajectory graph.

4. The method according to claim 1, characterized in that, The method of injecting energy into target nodes that match the power equipment fault description text, based on a preset risk activation energy, specifically includes: Statistically analyze the frequency and severity of historical fault descriptions corresponding to the target node; The risk activation energy adjustment coefficient of the target node is determined based on the frequency of the historical fault descriptions and the historical fault levels. Multiply the risk activation energy by the risk activation energy adjustment coefficient to obtain the adjusted risk activation energy; Based on the adjusted risk activation energy, energy is injected into the target node.

5. The method according to claim 1, characterized in that, Prior to the step of obtaining the power equipment fault description text input by the maintenance personnel, the method further includes: If no power equipment fault description text is received from maintenance personnel within a preset time period, the components of the initial fault risk vector of the power equipment at the initial moment are attenuated based on the preset time attenuation factor to obtain the attenuated fault risk vector.

6. The method according to claim 5, characterized in that, The attenuation process, based on a preset time decay factor, for each component of the initial fault risk vector of the power equipment at the initial moment, specifically includes: Obtain the component operating load level of each node corresponding to each component in the initial fault risk vector of the power equipment at the initial moment; Based on the operating load level of the component, determine the time decay coefficient corresponding to each component; Based on the time decay factor and the corresponding time decay coefficient, the adjusted time decay factor is obtained; Based on the adjusted time decay factor, the components in the initial fault risk vector of the power equipment at the initial moment are decayed.

7. The method according to claim 1, characterized in that, After the step of determining the fault propagation link based on the connection relationship of the fault evolution trajectory diagram, the target node, and the downstream node if the current risk assessment value is greater than the risk threshold, the method further includes: Based on the fault propagation path, a device monitoring plan is generated, which includes a monitoring cycle and a monitoring method. Acquire equipment operation data collected according to the equipment monitoring plan, input the equipment operation data into a preset equipment status assessment model, and obtain an equipment health status score; If the equipment health status score is lower than the equipment health status score threshold, the changing trend of abnormal parameters in the equipment operation data is extracted, and the development trend of the abnormal parameters is predicted based on the preset abnormal trend evolution model. Based on the development trend of the abnormal parameters, a corresponding fault handling solution is matched from the preset fault handling knowledge base, and operation and maintenance guidance suggestions are generated based on the fault handling solution.

8. A power new energy equipment testing system, characterized in that, The power new energy equipment testing system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to cause the power new energy equipment testing system to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is run on the power and new energy equipment testing system, the power and new energy equipment testing system performs the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, When the computer program product is run on the power and new energy equipment testing system, the power and new energy equipment testing system performs the method as described in any one of claims 1-7.

Images