Transformer health monitoring method based on importance degree and fault influence degree analysis

Abstract

The invention discloses a transformer health monitoring method based on importance and fault influence degree analysis, and the method comprises the following steps: 1, splitting each part of a transformer, and carrying out the importance evaluation of each part; 2, obtaining the fault type of each component and main risk factors causing the fault, calculating the risk priority of the fault, and carrying out the quantitative evaluation of the influence degree of the risk on each component according to the risk priority of the fault; 3, combining the importance degree evaluation result of each component with the influence degree quantitative evaluation result to obtain a comprehensive risk index; and 4, preferentially deploying the sensors of the high-priority monitoring points in the limited space according to the comprehensive risk index of each component. The invention provides a multi-dimensional comprehensive risk index calculation method combining component importance evaluation based on fuzzy comprehensive evaluation and fault influence degree evaluation of quantitative risk, and accurate configuration of the sensor is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of transformer health monitoring, and in particular to a transformer health monitoring method based on importance and fault impact analysis. Background Art

[0002] As a core component of a ship's power system, marine transformers meet the power needs of various electronic devices onboard, improving energy efficiency. The complex and harsh operating environment of ships at sea increases the likelihood of transformer failure, which can easily cause significant economic losses and even threaten personnel safety. Marine transformers have complex internal structures, resulting in numerous potential fault points, each with a wide range of potential fault types. To ensure ship safety during navigation, health monitoring of marine transformers is necessary to improve their reliability, ensure stable power system operation, enhance operational efficiency, and extend their service life.

[0003] Traditional condition monitoring methods rely on deploying sensors at multiple locations to collect transformer operating data. However, marine transformers are smaller and more compact than their land-based counterparts, so selecting data collection points requires focusing on critical components and the locations associated with the causes of component failures. Summary of the Invention

[0004] In view of the deficiencies in the prior art, the present invention provides a transformer health monitoring method based on importance and fault impact analysis to solve the technical problem in the prior art that sensors cannot be well arranged when available space is limited.

[0005] The present invention provides a transformer health monitoring method based on importance and fault impact analysis, which is characterized by comprising the following steps:

[0006] Step 1: Disassemble the transformer components and evaluate the importance of each component;

[0007] Step 2: Obtain the fault type of each component and the main risk factors that cause the fault, calculate the risk priority level of the fault, and quantitatively evaluate the impact of the risk on each component based on the risk priority level of the fault;

[0008] Step 3: Combine the importance assessment results of each component with the quantitative impact assessment results to obtain a comprehensive risk index;

[0009] Step 4: Deploy sensors at high-priority monitoring points in a limited space based on the comprehensive risk index of each component.

[0010] Furthermore, in step 1, the specific steps of the importance evaluation method are:

[0011] Step 11: Obtain the factors affecting the importance of each component and construct a factor set using the fuzzy comprehensive evaluation method;

[0012] Step 12: Divide the importance assessment levels and establish an evaluation set;

[0013] Step 13: Perform single-factor fuzzy evaluation on each component to obtain the evaluation matrix of each component;

[0014] Step 14: Set the membership between the function importance and the importance evaluation level in the factor set to obtain a membership set;

[0015] Step 15: Construct the fuzzy set of factor weights in the factor set;

[0016] Step 16: Obtain a fuzzy vector based on the fuzzy set of factor weights and the membership set, and use the importance evaluation level corresponding to the maximum value in the fuzzy vector as the importance of each component.

[0017] Furthermore, in step 2, the specific method for calculating the risk priority level of the fault is:

[0018] A fault tree consisting of fault types and the main risk factors that cause faults is constructed, and the fault tree is analyzed to determine the risk priority of the fault by calculating the fault severity score, fault probability score, and fault detectability score. Among them, the fault severity is the negative impact of the fault on the operation of the ship's overall power system; the fault probability is the number of times the fault occurs per unit time or the possibility of occurrence under normal operating conditions; the fault detectability is the possibility of discovering the fault before or in the early stage of the fault; the score range is: 1 to N.

[0019] Furthermore, the formula for calculating the risk priority level of the fault is:

[0020] RPN=S×O×D

[0021] Where S is the fault severity score; O is the fault occurrence probability score; and D is the fault detectability score.

[0022] Furthermore, when the sensor monitoring data changes more than the standard deviation of the historical data, the current fault severity score is updated. The fault severity score update formula is:

[0023]

[0024] Where S base Indicates the current fault severity score, α1~α nIndicates the sensitivity coefficient of the fault to the 1st-nth monitoring data, ΔX1~ΔX n Indicates the change in the parameters of the 1st to nth monitoring data, σ1~σ n Indicates the standard deviation of the historical data of the 1st to nth types of monitoring data.

[0025] Furthermore, when the ship's navigation environment changes, the current fault probability score is updated. The formula for updating the fault probability score is:

[0026] O=O base ×exp(β1T+β2H+…+β n E n )

[0027] Where, O base Indicates the current fault probability score, β1~β n Indicates the influence coefficient of various environmental factors in the equipment operating environment, T indicates the ambient temperature factor in the equipment operating environment, H indicates the ambient humidity factor in the equipment operating environment, E n Indicates the nth environmental factor in the device operating environment.

[0028] Furthermore, when the sensor performance degrades by one level, the fault detectability score is updated. The fault detectability score update formula is:

[0029]

[0030] Where D base Indicates the initially determined fault detectability level, γ1~γ n Indicates the speed at which the performance of the first to nth sensors involved in the fault decays over time, Δt1~Δt n Indicates the usage time of the 1st to nth type of sensors involved in the fault, T life1 ~T lifen Indicates the design life of the 1st to nth sensors involved in the fault.

[0031] Furthermore, in step 3, the method for obtaining the comprehensive risk index is:

[0032] Construct an importance vector matrix, where the importance of the matrix's behavioral components is listed as each component; construct a fault risk priority vector matrix, where the risk priority corresponding to each fault of the matrix's behavioral components is listed as each component; multiply the importance vector matrix and the risk priority vector matrix to obtain a comprehensive risk index. The specific formula is:

[0033] IR=(I / M)*(R / N 3 )

[0034] Where I is the importance vector matrix; R is the risk priority vector matrix; M is the highest importance level value; and N is the maximum value of the evaluation range in the risk priority level.

[0035] Furthermore, after step 4, the method further includes: when any one of the fault severity score, the fault probability score or the fault detectability score is updated, adjusting the activation state of the sensor according to the updated comprehensive risk index.

[0036] Furthermore, after step 4, the method further includes: constructing a digital twin system to implement monitoring, and the method of constructing the digital twin system is:

[0037] Build three-dimensional models of each component based on the physical entities of each transformer; assemble the three-dimensional models of each component in the development engine to build a complete twin model of the transformer; deploy sensor models of high-priority monitoring points in the twin system based on the comprehensive risk index of each component; build a health monitoring interface to achieve interactivity, and conduct data monitoring of high-priority points on the marine transformer.

[0038] Beneficial effects of the present invention:

[0039] The present invention proposes a multi-dimensional comprehensive risk index calculation method that combines component importance assessment based on fuzzy comprehensive evaluation with fault impact assessment based on quantified risk to achieve accurate configuration of sensors.

[0040] The present invention constructs a multi-dimensional factor set through a fuzzy comprehensive evaluation method to achieve the evaluation of component importance, and combines weight distribution to avoid the limitations of traditional subjective judgment of component importance.

[0041] The present invention performs single-factor fuzzy evaluation on each component to obtain an evaluation matrix for each component, making importance assessment more accurate.

[0042] The present invention establishes a three-dimensional fault impact assessment model based on the fault tree of component failure types and their causes, realizes dynamic quantification of failure risks, and provides a targeted basis for health monitoring.

[0043] The multi-dimensional comprehensive risk index of the present invention normalizes and integrates component importance and fault impact, solves the problem of redundant or omitted monitoring points caused by a single evaluation method, and improves health monitoring efficiency.

[0044] The present invention dynamically adjusts the fault severity score, fault probability score or fault detectability score according to the real-time status of each transformer component, realizes dynamic adjustment of the comprehensive risk index, thereby dynamically adjusting the activation state of the sensor and improving the accuracy of monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings, which are schematic and should not be construed as limiting the present invention in any way. In the accompanying drawings:

[0046] Figure 1 is a flow chart of a specific embodiment of the present invention;

[0047] Figure 2 A diagram of components of a marine transformer according to a specific embodiment of the present invention;

[0048] Figure 3 A block diagram of a system for evaluating the importance of marine transformer components according to a specific embodiment of the present invention;

[0049] Figure 4 A fault tree showing the fault types and causes of various components of a marine transformer according to a specific embodiment of the present invention;

[0050] Figure 5 This is a flow chart of sensor installation point selection according to a specific embodiment of the present invention;

[0051] Figure 6 It is a system framework diagram of a specific embodiment of the present invention. DETAILED DESCRIPTION

[0052] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0053] The present invention will be further described below with reference to specific examples. Those skilled in the art will appreciate that these examples are intended only to illustrate the present invention and are not intended to limit the scope of the present invention, and that modifications to various equivalent forms of the present invention fall within the scope defined by the appended claims.

[0054] like Figure 1 、 5 As shown, the present invention provides a transformer health monitoring method based on importance and fault impact analysis, comprising the following steps:

[0055] Step 1: Before evaluating the importance of marine transformers, it is necessary to disassemble the various components of the transformer and obtain the main components of the transformer: winding, core, insulation bushing, oil tank, tap changer, protection device and cooling device, such as Figure 2As shown, the main parts of the iron core are: iron core column, iron yoke; the main parts of the winding are: primary winding, secondary winding; the main parts of the oil tank and oil pillow are: main container, end cover, oil level gauge, valve; the main parts of the cooling device are: radiator, fan, flow rate relay; the main parts of the protection device are: oil conservator, oil level gauge, gas relay, temperature measuring element, desiccant; the main parts of the insulating bushing are: high-voltage bushing, low-voltage bushing; the main parts of the tap changer are: main contact, main on-off contact;

[0056] After that, the importance of each component is evaluated and the components of the marine transformer are sorted according to their importance. The specific steps are as follows:

[0057] Step 11: Figure 3 As shown in the figure, the factors affecting the importance of each component are obtained, and a factor set is constructed through the fuzzy comprehensive evaluation method; the factor set mainly includes seven indicators: functional importance, reliability, maintainability, safety, economy, equipment manufacturing technology, and environmental adaptability. At the same time, the evaluation criteria of each indicator are as follows: functional importance: the role and contribution of the component, and the direct impact on the operation of the equipment; reliability: failure frequency, environmental adaptability; maintainability: maintenance difficulty, maintenance cost; safety: impact on personnel safety; economy: procurement, installation, transportation and other costs, economic losses during maintenance; equipment manufacturing technology: component materials, technical level; environmental adaptability: the impact of components on the environment after being scrapped;

[0058] The indicator set is:

[0059] U = (functional importance u1, reliability u2, maintainability u3, safety u4, economy u5, equipment manufacturing technology u6, environmental adaptability u7)

[0060] Step 12: Divide the component importance evaluation level into G = {Extremely High (EH), High (H), Moderately High (MH), Medium (M), Low (L)}, and establish the evaluation set V = (v1, v2, v3, v4, v5);

[0061] Step 13: Perform single-factor fuzzy evaluation on each component to obtain the evaluation matrix of each component;

[0062] Step 14: Assume that the highest membership of the functional importance in the factor set to the component importance evaluation level is r11, then the membership set of the i-th factor in the factor set to the evaluation level is R i =(r i1 , r i2 , r i3 , r i4 , r i5 ), with R i Establish fuzzy comprehensive evaluation matrix R for row 7*5 ;

[0063] Step 15: The importance of each factor indicator in the factor set is also different. A fuzzy set of factor weights A = (a1, a2, a3, a4, a5, a6, a7) is established. Based on the suggestions of operation and maintenance personnel and experts, the weights are distributed as A = (0.3, 0.2, 0.2, 0.1, 0.1, 0.05, 0.05);

[0064] Step 16: Transform the fuzzy vector on the index set U into the fuzzy vector on the evaluation set V through fuzzy transformation, and use B=A 1*7 *R 7*5 Indicates that the fuzzy vector B = (b1, b2, b3, b4, b5) is obtained, and the importance evaluation level corresponding to the maximum value among b1 to b5 is taken as the importance of each component.

[0065] Step 2: If Figure 4 As shown in the figure, the fault types of each component and the main risk factors that cause the faults are obtained, and a fault tree consisting of the fault types and the main risk factors that cause the faults is constructed. Through the fault tree, the risk priority of the fault is determined by calculating the fault severity score, fault probability score, and fault detectability score. Among them, the fault severity score (Severity, S) refers to the degree of negative impact of the fault on the operation of the ship's overall power system. The scoring criteria are 1 to 5 points, and the score increases as the negative impact of the fault increases. The fault probability score (Dectection, D) refers to whether the system's monitoring capability can detect the fault in time before the fault occurs or in the early stage of the fault. The scoring criteria are 1 to 5 points, and the score increases as the fault monitoring capability decreases. The fault detectability score (Dectection, D) refers to whether the system's monitoring capability can detect the fault in time before the fault occurs or in the early stage of the fault. The scoring criteria are 1 to 5 points, and the score increases as the fault monitoring capability decreases.

[0066] By quantitatively evaluating the severity, probability, and detectability of a fault, the risk priority level (RPN) of the fault is calculated to determine the impact of the fault. The formula for calculating the risk priority level of the fault is:

[0067] RPN=S×O×D

[0068] Where S is the fault severity score; O is the fault occurrence probability score; and D is the fault detectability score.

[0069] The impact of faults is divided into low risk, medium risk, high risk, and extremely high risk, and their corresponding RPNs are less than or equal to 8, 9 to 14, 15 to 29, and greater than 30, respectively.

[0070] The fault severity score S, fault probability score O, and fault detectability score D in the quantitative assessment of fault impact can be dynamically updated and adjusted. By driving the update of the fault severity score S at the fault stage, the update of the fault probability score O by environmental perception, and the update of the fault detectability score D by life tracking, dynamic adjustment of the RPN is achieved, and further dynamic adjustment of the IR is achieved. The specific update method is:

[0071] When the sensor monitoring data changes more than the standard deviation of the historical data, the current fault severity score is updated. The fault severity score update formula is:

[0072]

[0073] Where S base Indicates the current fault severity score, α1~α n Indicates the sensitivity coefficient of the fault to the 1st-nth monitoring data, ΔX1~ΔX n Indicates the change in the parameters of the 1st to nth monitoring data, σ1~σ n Indicates the standard deviation of the historical data of the 1st to nth types of monitoring data.

[0074] When the ship's navigation environment changes, the current fault probability score is updated. The update formula for the fault probability score is:

[0075] O=O base ×exp(β1T+β2H+…+β n E n )

[0076] Where, O base Indicates the current fault probability score, β1~β n Indicates the influence coefficient of various environmental factors in the equipment operating environment, T indicates the ambient temperature factor in the equipment operating environment, H indicates the ambient humidity factor in the equipment operating environment, E n Indicates the nth environmental factor in the device operating environment.

[0077] When the sensor performance drops by one level, the fault detectability score is updated. The fault detectability score update formula is:

[0078]

[0079] Where D base Indicates the initially determined fault detectability level, γ1~γ n Indicates the speed at which the performance of the first to nth sensors involved in the fault decays over time, Δt1~Δt n Indicates the usage time of the 1st to nth type of sensors involved in the fault, Tlife1 ~T lifen Indicates the design life of the 1st to nth sensors involved in the fault.

[0080] Although dynamic adjustment of the RPN will affect the subsequent comprehensive risk index, it does not mean that the sensor layout will change. When initially laying out sensors, it is necessary to consider monitoring points and related parameters that were previously unimportant but have increased in importance under specific circumstances, and install backup sensors accordingly. When a fault with an originally low IR value increases to a point where a sensor layout is required for health monitoring due to dynamic adjustment, the backup sensor installed at that point can be activated. When a fault with an originally high IR value increases to a higher health monitoring requirement due to dynamic adjustment, the sampling frequency of the sensor is increased to more accurately display the health status of the equipment.

[0081] Step 3: Combine the importance assessment results of each component with the quantitative impact assessment results to obtain a comprehensive risk index;

[0082] The component importance obtained by the fuzzy comprehensive evaluation method was converted into numerical weights to obtain the importance level value (I): very high = 5, high = 4, moderately high = 3, medium = 2, and low = 1;

[0083] Establish the importance vector matrix I: I1 represents the importance level value of the winding, I2 represents the importance level value of the iron core, I3 represents the importance level value of the oil tank, I4 represents the importance level value of the cooling device, I5 represents the importance level value of the protection device, I6 represents the importance level value of the insulating sleeve, and I7 represents the importance level value of the decomposition switch.

[0084]

[0085] Establish the risk priority vector matrix R of the fault: 11 Indicates the risk priority level of the first fault of the winding and winding deformation, R 12 Indicates the risk priority level of the second type of winding fault and short circuit fault, R 1n Indicates the risk priority level of the nth type of winding fault...R 7n Indicates the risk priority level of the nth fault of the tap changer. It should be noted that the present embodiment defines the number of faults for each component differently, some are more and some are less. If there is no such fault, its value is 0. However, the method of the present invention is applicable to more detailed fault type classification in the future.

[0086]

[0087] Calculate the comprehensive risk index (IR): IR 11Indicates the comprehensive risk index of the first fault of the winding and winding deformation, IR 12 Indicates the comprehensive risk index of the first fault and short circuit fault of the winding, IR 1n The comprehensive risk index of the nth type of fault in the winding... IR 7n represents the comprehensive risk index of the nth fault of the tap changer,

[0088]

[0089] Where I is the importance vector matrix; R is the risk priority level; M is the highest importance level value, which is 5 in this embodiment; N is the maximum value of the evaluation range in the risk priority level, which is 5 in this embodiment, and N is the maximum value of the evaluation range in the risk priority level. 3 is 125.

[0090] The values within the matrix range from 0 to 1. A higher value indicates a greater comprehensive risk index for the fault, and a higher priority for monitoring points related to the fault's cause. By sorting the comprehensive risk index, we can identify fault types with higher monitoring priorities and their associated monitoring points. It's important to note that this invention remains applicable even when performing more detailed component disassembly and establishing component fault trees.

[0091] Step 4: Deploy sensors at high-priority monitoring points in a limited space based on the comprehensive risk index of each component.

[0092] like Figure 6 As shown in the figure, the system mainly uses the digital twin method to realize monitoring data visualization and health monitoring with stronger human-computer interaction, including: twin model modeling module, component importance ranking module, fault impact analysis module, health monitoring data acquisition module, and twin health monitoring system implementation module.

[0093] SolidWorks software was used to create 3D models of the various components and parts of the marine transformer, including the geometric model, behavioral model, physical model, and rule model of the marine transformer entity. The model was imported into the Unity engine using 3DMAX software for rendering, forming a digital twin model of the marine transformer. Locations with high comprehensive risk indices were marked, and interactive logic was added to the components using the C# language. This interactive logic primarily included component hiding button logic and mouse click-to-view logic. The component hiding button logic provided hide and show buttons for each component, allowing users to hide and show components through buttons, allowing them to see the internal components through the external components. Mouse click-to-view logic: When the user moves the mouse over a component or location with interactive logic, the data for that monitoring point is displayed directly on the component or monitoring point.

[0094] Before implementing the mouse click and view logic, it is necessary to first bind monitoring points to health monitoring data. Health monitoring data collection includes calculating the comprehensive fault risk index, establishing high-priority monitoring points and indicators, designing sensor layouts, collecting marine transformer operating parameters, data management, and data transmission.

[0095] By calculating a comprehensive fault risk index, we determined that the priority monitoring points and indicators were: winding temperature, winding vibration, core temperature, core vibration, oil tank gas, humidity inside the tank, and winding discharge. Sensors were deployed at these locations to obtain marine transformer operating parameters such as phase voltage, phase current, current frequency, phase temperature, humidity, imbalance, and oil and gas data. Because Unity cannot directly access the data collected by sensors at these monitoring points, the transformer operating parameters were transferred to a MySQL database, and Unity's request API was used to bind the data to components.

[0096] Data visualization is achieved by building a health monitoring interface using the Xchart plug-in and the Unity engine's native GUI functionality. This interface uses bar charts, pie charts, direct data displays, and instrumentation to visually display health monitoring data. These intuitive methods still require data binding via the API. The constructed health monitoring interface, along with components incorporating interactive logic, is integrated with monitoring points to form a complete, interactive digital twin health monitoring system. Monitoring points are selected based on a comprehensive analysis of component importance and failure impact.

[0097] Compared with the existing technology, this embodiment disassembles the components and parts of the marine transformer, evaluates the importance of the components using the fuzzy comprehensive evaluation method, and combines the quantitative analysis of the impact of each component failure to achieve a comprehensive fault risk index ranking, determine high-priority monitoring points and their indicators, and prioritize monitoring of more important point data when the internal space of the marine transformer is limited. By building a health monitoring interface and a three-dimensional twin model of the marine transformer, and adding interactive logic to the components and monitoring points, a digital twin method is used to implement a health monitoring system based on importance and fault impact analysis, enhancing the system's intuitiveness and interactivity, and more comprehensively monitoring the health of the marine transformer, avoiding the investment of limited resources in components and faults that have less impact on the operation of the marine transformer. The intuitiveness and interactivity allow users to more conveniently obtain data representing the health information of the marine transformer, understand the health status of the equipment, reduce operation and maintenance costs, and improve the safety of the ship's power system.

[0098] Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention. Such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A transformer health monitoring method based on importance and fault impact analysis, characterized in that: The steps include: Step 1: Disassemble the transformer components and evaluate the importance of each component; Step 2: Obtain the fault type of each component and the main risk factors that cause the fault, calculate the risk priority level of the fault, and quantitatively evaluate the impact of the risk on each component based on the risk priority level of the fault; Step 3: Combine the importance assessment results of each component with the quantitative impact assessment results to obtain a comprehensive risk index; Step 4: Deploy sensors at high-priority monitoring points in a limited space based on the comprehensive risk index of each component.

2. The transformer health monitoring method based on importance and fault impact analysis according to claim 1, characterized in that: In step 1, the specific steps of the importance evaluation method are: Step 11: Obtain the factors affecting the importance of each component and construct a factor set using the fuzzy comprehensive evaluation method; Step 12: Divide the importance assessment levels and establish an evaluation set; Step 13: Perform single-factor fuzzy evaluation on each component to obtain the evaluation matrix of each component; Step 14: Set the membership between the function importance and the importance evaluation level in the factor set to obtain a membership set; Step 15: Construct the fuzzy set of factor weights in the factor set; Step 16: Obtain a fuzzy vector based on the fuzzy set of factor weights and the membership set, and use the importance evaluation level corresponding to the maximum value in the fuzzy vector as the importance of each component.

3. The transformer health monitoring method based on importance and fault impact analysis according to claim 1 is characterized in that: In step 2, the specific method for calculating the risk priority level of the fault is: A fault tree consisting of fault types and the main risk factors that cause faults is constructed, and the fault tree is analyzed to determine the risk priority of the fault by calculating the fault severity score, fault probability score, and fault detectability score. Among them, the fault severity is the negative impact of the fault on the operation of the ship's overall power system; the fault probability is the number of times the fault occurs per unit time or the possibility of occurrence under normal operating conditions; the fault detectability is the possibility of discovering the fault before or in the early stage of the fault; the score range is: 1 to N.

4. The transformer health monitoring method based on importance and fault impact analysis according to claim 3 is characterized in that: The formula for calculating the risk priority level of the fault is: RPN=S×O×D Where S is the fault severity score; O is the fault occurrence probability score; and D is the fault detectability score.

5. The transformer health monitoring method based on importance and fault impact analysis according to claim 4 is characterized in that: When the sensor monitoring data changes more than the standard deviation of the historical data, the current fault severity score is updated. The fault severity score update formula is: Where S base Indicates the current fault severity score, α1~α n Indicates the sensitivity coefficient of the fault to the 1st-nth monitoring data, ΔX1~ΔX n Indicates the change in the parameters of the 1st to nth monitoring data, σ1~σ n Indicates the standard deviation of the historical data of the 1st to nth types of monitoring data.

6. The transformer health monitoring method based on importance and fault impact analysis according to claim 4 is characterized in that: When the ship's navigation environment changes, the current fault probability score is updated. The update formula for the fault probability score is: O=O base ×exp(β1T+β2H+…+β n E n ) Where, O base Indicates the current fault probability score, β1~β n Indicates the influence coefficient of various environmental factors in the equipment operating environment, T indicates the ambient temperature factor in the equipment operating environment, H indicates the ambient humidity factor in the equipment operating environment, E n Indicates the nth environmental factor in the device operating environment.

7. The transformer health monitoring method based on importance and fault impact analysis according to claim 4 is characterized in that: When the sensor performance drops by one level, the fault detectability score is updated. The fault detectability score update formula is: Where D base Indicates the initially determined fault detectability level, γ1~γ n Indicates the speed at which the performance of the first to nth sensors involved in the fault decays over time, Δt1~Δt n Indicates the usage time of the 1st to nth type of sensors involved in the fault, T life1 ~T lifen Indicates the design life of the 1st to nth sensors involved in the fault.

8. The transformer health monitoring method based on importance and fault impact analysis according to any one of claims 1 to 4, characterized in that: In step 3, the method for obtaining the comprehensive risk index is: Construct an importance vector matrix, where the importance of the matrix's behavioral components is listed as each component; construct a fault risk priority vector matrix, where the risk priority corresponding to each fault of the matrix's behavioral components is listed as each component; multiply the importance vector matrix and the risk priority vector matrix to obtain a comprehensive risk index. The specific formula is: IR=(I / M)*(R / N 3 ) Where I is the importance vector matrix; R is the risk priority vector matrix; M is the highest importance level value; and N is the maximum value of the evaluation range in the risk priority level.

9. The transformer health monitoring method based on importance and fault impact analysis according to any one of claims 5 to 7, characterized in that: After step 4, the method further includes: when any one of the fault severity score, the fault probability score, or the fault detectability score is updated, adjusting the activation state of the sensor according to the updated comprehensive risk index.

10. The transformer health monitoring method based on importance and fault impact analysis according to claim 3, characterized in that: After step 4, the method further includes: constructing a digital twin system to implement monitoring, and the method of constructing the digital twin system is: Build three-dimensional models of each component based on the physical entities of each transformer; assemble the three-dimensional models of each component in the development engine to build a complete twin model of the transformer; deploy sensor models of high-priority monitoring points in the twin system based on the comprehensive risk index of each component; build a health monitoring interface to achieve interactivity, and conduct data monitoring of high-priority points on the marine transformer.

Images