RCM-based electrical equipment operation reliability analysis method

Abstract

The invention discloses an electrical equipment operation reliability analysis method based on RCM, and belongs to the technical field of electrical equipment operation and maintaining.The method comprises the steps that electrical equipment is decomposed into main function nodes and auxiliary function nodes through a function tree analysis method, a function performance standard library is constructed, and quantitative indexes of all the nodes and an incidence relation graph of the quantitative indexes are stored; constructing a failure mode library based on the function performance standard library, fusing a physical failure mechanism with historical fault data to generate a failure mode-function node mapping table, calculating a system-level function failure probability by adopting a Monte Carlo simulation method, and outputting a risk level matrix; using the risk level matrix to match the risk level to the node logic of the RCM decision tree, and generating a hybrid strategy set; and according to an execution result of the mixed strategy set, calculating a comprehensive health degree through an equipment health index model, and iteratively updating a failure mode library and a decision tree threshold parameter in combination with a reinforcement learning algorithm.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of power equipment operation and maintenance, and particularly relates to an electrical equipment operation reliability analysis method based on RCM. BACKGROUND

[0002] The traditional power equipment maintenance mode mainly adopts the strategy of combining time-based maintenance (TBM) and corrective maintenance (CM). However, with the expansion of the smart grid and the high proportion of new energy access, this mode exposes significant defects: first, the problem of excessive maintenance is prominent. According to statistics, about 35% of the components in the planned maintenance of the electrical equipment of the thermal power plant do not need to be maintained; second, insufficient maintenance leads to chain failures. The root cause is that the traditional method lacks quantitative analysis of the degradation mechanism of the equipment function and the risk propagation of the system level, and cannot achieve precise maintenance.

[0003] In the field of reliability-centered maintenance (RCM), the existing technology has partially solved the above problems, but there are still key bottlenecks. For example, patent CN201910235678A proposes an RCM optimization method based on FMEA, but its limitations are: (1) lack of system-level failure correlation: only single equipment failure analysis is performed, ignoring the cascading failure risk caused by the functional coupling between devices. In the actual power grid, transformer winding overheating may cause relay protection misoperation, and then expand the power outage range, which is not modeled in this cross-device failure path; (2) insufficient dynamic adaptability: strategy optimization relies on static historical data, without introducing real-time state monitoring information and self-learning mechanism, making it difficult to cope with the dynamic changes in device load caused by new energy output fluctuations. Experiments show that the method has a false detection rate of up to 28% in the photovoltaic power station scenario. In addition, existing RCM methods have defects in multi-objective collaborative optimization, often considering safety and economy in isolation, without establishing a comprehensive decision-making model of cost-risk-life.

[0004] With the evolution of the smart grid towards high flexibility and digitization, the operating environment of electrical equipment is becoming increasingly complex. On the one hand, the interconnection of distributed power sources leads to bidirectional power flow, and the fluctuation range of device load rate increases by more than 60% compared to traditional power grids, accelerating the process of hidden failures such as insulation aging; on the other hand, the deployment of the Internet of Things (IoT) makes the device state data grow exponentially, but the data value has not been fully tapped. New RCM methods are needed to break through three technical barriers: 1. System-level functional failure modeling to quantify the global impact of key device failures on the power grid topology; 2. Multi-source heterogeneous data fusion to achieve decision-making upgrade from "experience-driven" to "data-mechanism hybrid-driven"; 3. Dynamic closed-loop optimization to adjust the maintenance strategy in real time through online learning.

[0005] In summary, the existing RCM technology is difficult to meet the fine and adaptive requirements of device reliability management of new power systems, and developing a new method supporting system-level failure correlation analysis and embedded dynamic optimization mechanism has become a common technical challenge in the industry. SUMMARY

[0006] The purpose of the present application is to overcome the problem that the existing RCM method does not consider system-level failure correlation, and a method for analyzing the operational reliability of electrical equipment based on RCM is proposed.

[0007] To achieve the above purpose, the technical scheme adopted by the present application is as follows: In a first aspect, the present application provides a method for analyzing the operational reliability of electrical equipment based on RCM, comprising the following steps: The electrical equipment is decomposed into main function nodes and auxiliary function nodes by the function tree analysis method, a function performance standard library is constructed, and the quantitative indicators of each node and their correlation graph are stored; Based on the function performance standard library, a failure mode library is constructed, the physical failure mechanism and historical failure data are fused to generate a failure mode-function node mapping table, the Monte Carlo simulation method is used to calculate the system-level function failure probability, and a risk level matrix is output; The risk level is matched to the node logic of the RCM decision tree using the risk level matrix, and a mixed strategy set is generated; According to the execution result of the mixed strategy set, the comprehensive health degree is calculated through the device health index model, and the failure mode library and the threshold parameters of the decision tree are iteratively updated combined with the reinforcement learning algorithm.

[0008] Further, the function tree analysis method includes: The main function node is defined as a function that directly affects the stability of the power grid, including the generator excitation control accuracy and the transformer winding temperature limit; The auxiliary function node is defined as a subsystem that supports the operation of the main function, including the cooling system flow and the circuit breaker operating mechanism action time; Each function node is associated with a dynamic performance threshold.

[0009] Further, the construction of the failure mode library includes: Based on the Arrhenius model, the thermal aging rate of insulating materials is calculated at the physical mechanism layer, and the Archard wear model is used to calculate the service life of mechanical contact components; In the data-driven layer, LSTM neural network is used to train the historical data of the SCADA system to predict hidden failure modes; In the risk quantification layer, the risk value is defined.

[0010] Further, the Monte Carlo simulation method specifically comprises: establishing a device failure probability distribution model based on historical failure data; and calculating a system-level function failure risk value in combination with experience to correct model parameters; The model parameters comprise a rotor imbalance amount probability distribution and an insulation breakdown voltage degradation rate. The system-level failure criterion is that the functions of equal to or greater than 2 main devices in the same substation fail and the standby capacity is less than a load demand threshold.

[0011] Further, the system-level function failure risk value is equal to a failure probability multiplied by a safety loss coefficient, an environmental penalty factor and a repair cost.

[0012] Further, the risk level matrix is matched to the node logic of the RCM decision tree by using a risk level matrix and the RCM decision tree adopts a hierarchical mapping rule for strategy mapping, and the hierarchical mapping rule is as follows: The risk levels comprise a level I, a level II and a level III. The RCM node type of the level I is a predictive maintenance node, and the strategy generation rule of the level I is to enable online monitoring and analyze trends and increase threshold sensitivity. The RCM node type of the level II is a preventive maintenance node, and the strategy generation rule of the level II is to shorten a reference interval. The RCM node type of the level III is a post-failure maintenance node, and the strategy generation rule of the level III is to increase a safety stock amount.

[0013] Further, the decision logic of the RCM decision tree is as follows: First-level judgment: if the device health index is less than a first set value and the risk level is the level III, predictive maintenance is triggered, and an infrared imager and ultrasonic partial discharge detection are used to locate defects. Second-level judgment: if the first set value is less than or equal to the device health index and less than a second set value or the risk level is the level II, preventive maintenance is performed, and a lubrication cycle and a fastener torque value are updated according to a deterioration curve. Third-level judgment: if the device health index is greater than or equal to the second set value and the risk level is the level I, post-failure maintenance is enabled, and a spare parts inventory system is automatically replenished in linkage.

[0014] In a second aspect, the application provides an electrical equipment operation reliability analysis system based on RCM, comprising: A function performance standard library construction module is configured to decompose electrical equipment into main function nodes and auxiliary function nodes by a function tree analysis method, construct a function performance standard library, and store quantitative indexes of each node and an associated relationship graph. a risk level matrix output module, configured to construct a failure mode library based on the function performance standard library, fuse the physical failure mechanism and the historical fault data, generate a failure mode-function node mapping table, calculate a system-level function failure probability by using a Monte Carlo simulation method, and output the risk level matrix; a hybrid strategy set generation module, configured to match the risk level to the node logic of the RCM decision tree by using the risk level matrix, and generate the hybrid strategy set; a comprehensive health degree calculation module, configured to calculate the comprehensive health degree by using a device health index model according to an execution result of the hybrid strategy set, and iteratively update the failure mode library and the decision tree threshold parameter by using a reinforcement learning algorithm.

[0015] In a third aspect, the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the RCM-based electrical equipment operation reliability analysis method when executing the computer program.

[0016] In a fourth aspect, the present application provides a computer readable storage medium, which stores a computer program, wherein the computer program is executed by a processor to implement the RCM-based electrical equipment operation reliability analysis method.

[0017] Compared with the prior art, the present application has the following beneficial technical effects: The RCM-based electrical equipment operation reliability analysis method provided by the present application integrates function tree analysis, failure mode library construction, Monte Carlo simulation, RCM decision tree, and reinforcement learning, realizes systematic upgrading of electrical equipment reliability management, and realizes the transformation from a "black box" to a "transparent" state by deconstructing a complex device into a quantifiable function node network and establishing a standardized performance index system. The failure mode library is constructed by fusing the physical failure model and the historical fault data. This double-driving mechanism not only retains the universality of the theoretical model but also injects the specificity of the actual operation data, so that the failure mode prediction accuracy is improved. The Monte Carlo simulation introduces the probability dimension and converts the traditional deterministic analysis into a probabilistic risk assessment. The risk matrix is embedded into the RCM decision tree to form a "risk-strategy" dynamic mapping. Compared with the traditional RCM static decision table, the hybrid strategy set can automatically generate a multi-dimensional strategy including state monitoring frequency, maintenance threshold, and spare parts reserve. The device health index model serves as a feedback signal to drive the adaptive adjustment of the failure mode library and the decision tree threshold. From the construction of the function performance standard library in the design stage to the dynamic update of the failure mode in the operation and maintenance stage, a reliability management chain covering the whole cycle of the device is formed. This longitudinal through-connection capability enables the collaborative optimization of device selection, operation and maintenance strategy, and technical transformation. BRIEF DESCRIPTION OF DRAWINGS

[0018] The drawings described herein are for purposes of illustration only and are not intended to limit the scope of the present disclosure in any way. Additionally, those skilled in the art will recognize that the drawings described herein are not necessarily drawn to scale and that, unless otherwise specifically stated, the dimensions of the various features in the drawings are intended to be approximate rather than exact. Figure 1 A flow chart of an RCM-based electrical equipment operation reliability analysis method of the present application.

[0019] Figure 2 A structural diagram of an RCM-based electrical equipment operation reliability analysis system of the present application.

[0020] Figure 3 An electronic device diagram of an RCM-based electrical equipment operation reliability analysis method of the present application. DETAILED DESCRIPTION

[0021] In order to enable persons skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by persons skilled in the art without creative labor should fall within the scope of protection of the present application.

[0022] Embodiment one Referring to Figure 1 An RCM-based electrical equipment operation reliability analysis method, comprising the following steps: decomposing the electrical equipment into main function nodes and auxiliary function nodes through a function tree analysis method, constructing a function performance standard library, storing the quantitative indicators of each node and the associated relationship atlas, constructing a failure mode library based on the function performance standard library, fusing the physical failure mechanism and historical failure data to generate a failure mode-function node mapping table, calculating the system-level function failure probability by using a Monte Carlo simulation method, and outputting a risk level matrix; matching the risk level to the node logic of the RCM decision tree by using the risk level matrix to generate a mixed strategy set; calculating the comprehensive health degree through a device health index model according to the execution result of the mixed strategy set, and iteratively updating the failure mode library and the decision tree threshold parameters in combination with a reinforcement learning algorithm.

[0023] The function tree analysis method of this embodiment decomposes complex equipment into a quantifiable network of functional nodes, establishes a standardized performance index system, and realizes the transformation from "black box" to "transparent". For example, a transformer can be decomposed into main functional nodes such as insulation, cooling, and voltage regulation, each node is associated with quantifiable indexes such as dielectric loss value and temperature rise threshold, providing accurate data basis for subsequent analysis. Through the fusion of physical failure models (such as electric-thermal-mechanical stress coupling model) and historical failure data, a failure mode library is constructed. This double driving mechanism not only retains the universality of the theoretical model, but also injects the specificity of the actual operation data, making the failure mode prediction accuracy rate increase by 30%-50%. Typical applications such as switch cabinet contact failure mode analysis can consider both the electric-arc ablation physical process and historical contact resistance data. Monte Carlo simulation introduces the probability dimension, transforming traditional deterministic analysis into probabilistic risk assessment. Taking 100,000 simulations as an example, the system-level functional failure probability distribution curve can be obtained, which, combined with a risk matrix (such as a 5x5 risk level table), realizes the visualization of risk transmission from components to systems. After application in a certain power plant, the accuracy rate of high-risk equipment identification increased by 62%. The risk matrix is embedded into the RCM decision tree to form a "risk-strategy" dynamic mapping. Compared with the static decision table of traditional RCM, the mixed strategy set can automatically generate multi-dimensional strategies including state monitoring frequency, maintenance threshold, and spare parts reserve. For example, for high-risk main transformers, the system can recommend a monthly oil chromatographic analysis + intelligent lock phase discharge detection scheme. The reinforcement learning framework forms a closed loop of "strategy execution-health assessment-parameter update". The device health index model (integrating multi-source data such as vibration, temperature, and partial discharge) serves as a feedback signal to drive the adaptive adjustment of the failure mode library and the threshold of the decision tree. Practice in a certain wind farm shows that after three iterations, maintenance costs are reduced by 28% and equipment availability is increased by 19%. From the construction of the functional performance standard library in the design stage to the dynamic update of the failure mode in the operation and maintenance stage, a reliability management chain covering the whole cycle of equipment is formed. This vertical through-connection capability enables the optimization of equipment selection, maintenance strategy, and technical transformation to be coordinated, and typical cases show that the life cycle cost of equipment can be reduced by 15%-20%. By building a five-layer analysis framework including function decomposition, failure analysis, risk evaluation, strategy optimization, and dynamic feedback, reliability management is realized from the equipment level to the system level. Combined with online monitoring technology and big data analysis platform, an optimized maintenance system of prevention-prediction-correction is formed. Compared with the traditional planned maintenance mode, maintenance costs can be reduced by 20%-35% and equipment availability can be increased by 3-5 percentage points. A multi-dimensional maintenance strategy decision tree is developed to realize accurate matching of maintenance types; a data-driven closed-loop optimization mechanism is constructed to improve the adaptability of the strategy.

[0024] The function tree analysis method comprises: a main function node is defined as a function directly affecting power grid stability, including generator excitation control precision and transformer winding temperature rise limit value; an auxiliary function node is defined as a subsystem supporting the operation of the main function, including cooling system flow and circuit breaker operating mechanism action time; and each function node is associated with a dynamic performance threshold value.

[0025] The construction of the failure mode library comprises: calculating the thermal aging rate of insulating materials based on an Arrhenius model at a physical mechanism layer, and calculating the service life of mechanical contact components based on an Archard wear model; training historical SCADA system data by using an LSTM neural network at a data-driven layer to predict hidden failure modes; and defining a risk value at a risk quantification layer.

[0026] The Monte Carlo simulation method specifically comprises: establishing a device failure probability distribution model based on historical failure data; combining experience to correct model parameters, and calculating a system-level function failure risk value; the model parameters comprise rotor imbalance quantity probability distribution and insulation breakdown voltage degradation rate; and the system-level failure criterion is that the functions of ≥2 main devices in the same substation fail and the standby capacity is < a load demand threshold value.

[0027] The system-level function failure risk value = failure probability × (safety loss coefficient + environmental penalty factor + repair cost).

[0028] The risk level matrix and the RCM decision tree adopt hierarchical mapping rules for strategy mapping, and the hierarchical mapping rules are as follows: the risk levels comprise level I, level II and level III; the RCM node type of level I is a predictive maintenance node, and the strategy generation rule of level I is to enable online monitoring and analyze trends and increase threshold sensitivity; the RCM node type of level II is a preventive maintenance node, and the strategy generation rule of level II is to shorten the reference interval; and the RCM node type of level III is a post-failure maintenance node, and the strategy generation rule of level III is to increase the safety stock amount.

[0029] The decision logic of the RCM decision tree specifically comprises: a first-level judgment: if the device health index < a first set value and the risk level is level III, triggering predictive maintenance, and positioning defects by using an infrared imager and ultrasonic partial discharge detection; a second-level judgment: if the first set value ≤ the device health index < a second set value or the risk level is level II, performing preventive maintenance, and updating the lubrication period and fastener torque value according to the deterioration curve; and a third-level judgment: if the device health index ≥ the second set value and the risk level is level I, enabling post-failure maintenance, and automatically replenishing the spare parts inventory system in linkage.

[0030] Embodiment two Referring to Figure 2 A RCM-based electrical equipment operation reliability analysis system comprises: a functional performance standard library construction module, configured to decompose the electrical equipment into main function nodes and auxiliary function nodes through a function tree analysis method, construct a functional performance standard library, and store quantified indexes of each node and a correlation graph atlas thereof; a risk level matrix output module, configured to construct a failure mode library based on the functional performance standard library, fuse a physical failure mechanism and historical fault data, generate a failure mode-function node mapping table, calculate a system-level function failure probability by using a Monte Carlo simulation method, and output a risk level matrix; a hybrid strategy set generation module, configured to match the risk level to a node logic of an RCM decision tree by using the risk level matrix, and generate a hybrid strategy set; a comprehensive health degree calculation module, configured to calculate a comprehensive health degree by using a device health index model according to an execution result of the hybrid strategy set, and iteratively update a failure mode library and threshold parameters of a decision tree in combination with a reinforcement learning algorithm.

[0031] Embodiment Three With reference to Figure 3 An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the RCM-based electrical equipment operation reliability analysis method when executing the computer program.

[0032] Embodiment Four A computer-readable storage medium stores a computer program, and the computer program implements the RCM-based electrical equipment operation reliability analysis method when executed by a processor.

[0033] Embodiment Five An RCM-based electrical equipment operation reliability analysis method based on function-failure coupling analysis includes a closed-loop logic chain including the following steps: Step S1: Function hierarchical modeling and standard library construction decompose the electrical equipment into main function nodes and auxiliary function nodes through a function tree analysis method (FTA), wherein: The main function nodes include power transmission efficiency, insulation strength, and mechanical stability, and the dynamic performance threshold values thereof are defined , wherein is an initial threshold value, is an aging coefficient, is a time-varying load rate; A functional performance standard library is constructed, and quantified indexes of each node and a correlation graph atlas thereof are stored; Step S2: Failure-risk dynamic transmission chain generation is performed based on the standard library of step S1: A failure mode library is constructed, physical failure mechanisms (insulation thermal aging Arrhenius model, mechanical wear Archard model) are fused with historical failure data to generate a failure mode-function node mapping table; A Monte Carlo simulation is used to calculate system-level functional failure probability, the input being the failure transmission weight in the mapping table wij and real-time monitoring data, and the output being a three-dimensional risk level matrix , satisfying ; Step S3: Dynamic adaptation of strategy and optimization of weight Based on the risk matrix of step S2, the following is performed: Match the risk level to the node logic of the RCM decision tree: if then jump to the predictive maintenance branch, otherwise select the preventive or post-failure maintenance branch according to the preset rule; Generate a mixed strategy set , where the strategy weight wm is dynamically updated according to , the failure suppression rate of strategy m is , and the cost coefficient is ; Step S4: Closed-loop feedback driven parameter iteration Based on the strategy execution result of step S3, the following is performed: Calculate the comprehensive health degree through the device health index model , where is the index weight determined by the entropy weight method, and is the real-time monitoring value in the functional performance standard library; Use a reinforcement learning algorithm (deep Q network DQN) to iteratively update: State space ; Action space ; Reward function ; The steps S1→S4 form a closed-loop logic chain, which is specifically manifested as: The functional threshold of S1 directly limits the triggering condition of the failure mode library in S2; The risk matrix of S2 controls the strategy branch selection of S3 through the preset threshold value; The strategy execution effect of S3 is fed back to the DQN algorithm through the HI value and cost data of S4; The parameter update instruction output by DQN is fed back to the standard library of S1 and the failure mode library of S2 in real time, completing the dynamic closed loop.

[0034] In this embodiment, the threshold calculation formula of the functional performance standard library is The trigger logic embedded in the failure mode library allows functional degradation to directly drive failure probability updates (e.g., a 10% drop in insulation strength doubles the Arrhenius model aging rate). The dimensional design of the risk matrix (device × failure mode × severity) establishes a many-to-one relationship with the branch mapping of the RCM decision tree, ensuring that high-risk events trigger a multi-layer defense strategy. Strategy weight update formula The cost-benefit ratio is dynamically modified to prevent a single strategy from dominating for a long time. The DQN algorithm adjusts the function threshold, failure weight, and decision threshold at the same time to achieve coordinated optimization of the parameters of the entire chain (for example, when HI decreases, it will be reduced synchronously). , and improve ).

[0035] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, read-only optical disks, optical storage devices, etc.) containing computer-usable program code.

[0036] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0037] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0038] These computer program instructions can also be loaded into a computer or other programmable data processing devices, so that a series of operational steps are generated to realize the computer-implemented processes in the computer or other programmable devices, and the instructions executed in the computer or other programmable devices provide the processes for implementing the functions specified in the flowchart Figure 1 or multiple flows and / or the functions specified in one or more blocks of the flowchart Figure 1 or multiple blocks. Finally, it should be noted that the above-mentioned embodiments are merely used to illustrate the technical solutions of the present application, rather than limit the same. Although the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or equivalent replacements without departing from the spirit and scope of the present application, and any modifications or equivalent replacements made without departing from the spirit and scope of the present application should be covered within the protection scope of the present application.

Claims

1. A method for analyzing the operational reliability of electrical equipment based on RCM, characterized in that: The following steps are involved: By using the function tree analysis method, the electrical equipment is decomposed into main function nodes and auxiliary function nodes, and a functional performance standard library is constructed to store the quantitative indicators of each node and their correlation relationship maps; Build a failure mode library based on the functional performance standard library, integrate the physical failure mechanism with historical failure data, generate a failure mode-function node mapping table, use Monte Carlo simulation method to calculate the system-level functional failure probability, and output a risk level matrix; Use the risk level matrix to match the risk level to the node logic of the RCM decision tree to generate a hybrid strategy set; According to the execution results of the hybrid strategy set, the comprehensive health degree is calculated through the equipment health index model, and the failure mode library and decision tree threshold parameters are iteratively updated in combination with the reinforcement learning algorithm.

2. The RCM-based electrical equipment operation reliability analysis method according to claim 1, characterized in that: Function tree analysis includes: The main function nodes are defined as functions that directly affect the stability of the power grid, including generator excitation control accuracy and transformer winding temperature rise limit; Auxiliary function nodes are defined as subsystems that support the operation of the main function, including cooling system flow and circuit breaker operating mechanism action time; Each functional node is associated with a dynamic performance threshold.

3. The RCM-based electrical equipment operation reliability analysis method according to claim 1, characterized in that: The construction of the failure mode library includes: At the physical mechanism level, the thermal aging rate of insulating materials is calculated based on the Arrhenius model, and the life of mechanical contact components is calculated based on the Archard wear model; In the data-driven layer, LSTM neural networks are used to train historical data of the SCADA system to predict hidden failure modes; Define risk value at the risk quantification level.

4. The RCM-based electrical equipment operation reliability analysis method according to claim 3 is characterized in that: The Monte Carlo simulation method specifically involves: establishing a device failure probability distribution model based on historical failure data; modifying model parameters based on experience to calculate the system-level functional failure risk value; The model parameters include the probability distribution of rotor unbalance and insulation breakdown voltage degradation rate; The system-level failure criterion is that ≥2 main equipment in the same substation fail and the spare capacity is less than the load demand threshold.

5. The RCM-based electrical equipment operation reliability analysis method according to claim 4 is characterized in that: System-level functional failure risk value = failure probability × (safety loss coefficient + environmental penalty factor + repair cost).

6. The RCM-based electrical equipment operation reliability analysis method according to claim 1, characterized in that: In the logic of using the risk level matrix to match the risk level to the node of the RCM decision tree, the risk level matrix and the RCM decision tree use a hierarchical mapping rule for strategy mapping. The hierarchical mapping rule is: Risk levels include Level I, Level II, and Level III; The RCM node type of level I is a predictive maintenance node, and the strategy generation rules of level I are to enable online monitoring and trend analysis and increase threshold sensitivity; The RCM node type of level II is a preventive maintenance node, and the strategy generation rule of level II is to shorten the benchmark interval; The RCM node type of level III is a post-fault repair node, and the strategy generation rule of level III is to increase the safety stock.

7. The RCM-based electrical equipment operation reliability analysis method according to claim 6, characterized in that: The decision logic of the RCM decision tree is as follows: Level 1 judgment: If the equipment health index is less than the first set value and the risk level is Level III, predictive maintenance is triggered, and infrared imaging and ultrasonic partial discharge detection are used to locate defects; Second-level judgment: If the first set value ≤ equipment health index < second set value or the risk level is Level II, perform preventive maintenance and update the lubrication cycle and fastener torque value according to the degradation curve; Level 3 judgment: If the equipment health index is ≥ the second set value and the risk level is Level I, post-fault maintenance is enabled and the spare parts inventory system is linked to automatically replenish stocks.

8. An electrical equipment operation reliability analysis system based on RCM, characterized in that: include: Functional performance standard library construction module, used to decompose electrical equipment into main function nodes and auxiliary function nodes through function tree analysis method, build functional performance standard library, and store quantitative indicators of each node and their correlation relationship map; The risk level matrix output module is used to build a failure mode library based on the functional performance standard library, integrate physical failure mechanisms with historical failure data, generate a failure mode-function node mapping table, use Monte Carlo simulation to calculate the system-level functional failure probability, and output a risk level matrix; A hybrid strategy set generation module is used to match risk levels to the node logic of the RCM decision tree using the risk level matrix to generate a hybrid strategy set; The comprehensive health calculation module is used to calculate the comprehensive health through the equipment health index model based on the execution results of the hybrid strategy set, and iteratively update the failure mode library and decision tree threshold parameters in combination with the reinforcement learning algorithm.

9. An electronic device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, an RCM-based electrical equipment operation reliability analysis method as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for analyzing the operational reliability of electrical equipment based on RCM as described in any one of claims 1 to 7 is implemented.

Images