A method for detecting and evaluating the insulation state of GIL equipment of a hydropower station based on an industrial internet platform

Abstract

A kind of GIL equipment insulation state detection and evaluation method based on industrial internet platform, through online monitoring system, live detection device or power-off high voltage test, characteristic parameters such as voltage, current, temperature are collected to generate original feature set. After photoelectric conversion, filtering and other processing, key features are extracted to generate structured feature set, and are serialized and uploaded to data center. Using partial discharge analysis, knowledge graph and other technologies, phase resolution map and dielectric spectrum map are calculated, fault diagnosis is carried out, and information such as fault location and fault type is determined. Based on the diagnosis result, the insulation state risk is evaluated by fault mode and effect analysis, fault tree analysis and other methods, the defects, safety risk and economic risk are quantified, and the alarm level and processing suggestion are output. If the risk is triggered, convolutional neural network analysis is performed on specific signals to update the fault location, and finally a diagnosis report is generated according to the fault influence range, diagnosis result and alarm level, realizing comprehensive monitoring and accurate evaluation of GIL equipment insulation state.

Description

Technical Field

[0001] This invention belongs to the field of hydropower equipment maintenance and management technology, and specifically relates to a method for detecting and evaluating the insulation status of GIL equipment in hydropower stations based on an industrial internet platform. Background Technology

[0002] In hydropower stations, gas-insulated transmission pipelines (GILs) are widely used as a highly efficient power transmission method due to their compact structure, high transmission efficiency, and low maintenance costs. However, the insulation condition of GIL equipment directly affects the safe and stable operation of the power system. Monitoring and assessing the insulation condition is crucial for preventing equipment failures, reducing unexpected power outages, and ensuring power supply. Traditional methods for detecting the insulation condition of GIL equipment have limitations. For example, because GIL equipment uses a fully enclosed structure, internal condition information is difficult to detect, and equipment failure will seriously affect the normal operation of the system. In addition, existing conventional technologies such as ultra-high frequency and ultrasonic waves are insufficient to accurately detect internal defects in insulation components. Statistics show that GIS / GIL equipment failures are mainly related to insulation and mechanical defects generated during production, transportation, and operation, and mechanical defects are also a significant factor leading to insulation degradation.

[0003] The application of Industrial Internet platforms has placed new demands on the monitoring and evaluation methods of GIL (Gas Insulation Line) equipment. Industrial Internet platforms enable remote monitoring, data analysis, and intelligent decision-making, providing new technical means for the insulation condition detection and evaluation of GIL equipment. However, effectively combining Industrial Internet platforms with the insulation condition detection and evaluation of GIL equipment to achieve accurate monitoring and intelligent evaluation remains a technical challenge. Currently, research on insulation condition detection and evaluation methods for GIL equipment mainly focuses on partial discharge detection, ultrasonic guided wave detection, and vibration signal monitoring. These methods can reflect the insulation condition of GIL equipment to a certain extent, but still suffer from limited detection range, insufficient accuracy, and inability to comprehensively evaluate the insulation condition. Summary of the Invention

[0004] The technical problem to be solved by this invention is to provide a method for detecting and evaluating the insulation status of GIL equipment in hydropower stations based on an industrial internet platform. This method can make full use of the advantages of the industrial internet platform and combine multiple detection technologies to achieve comprehensive and accurate monitoring and evaluation of the insulation status of GIL equipment, thereby improving the operational safety and reliability of the power system.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0006] A method for detecting and evaluating the insulation status of GIL equipment in hydropower stations based on an industrial internet platform, comprising the following steps:

[0007] S1: Collect characteristic parameters of the insulation state of GIL equipment through online monitoring system, live detection device or high voltage test during power outage, including voltage, current, temperature, vibration, UHF signal and ultrasonic signal, and generate original feature set A, where A represents the initial dataset containing all collected signals;

[0008] S2: Perform photoelectric conversion, filtering and signal amplification on the original feature set A, and combine time domain analysis and frequency domain analysis to extract vibration amplitude, UHF signal feature value and ultrasonic signal feature value to generate structured feature set B, where B represents a standardized dataset containing the extracted features;

[0009] S3: Serialize the structured feature set B into a JSON data packet according to a predefined format, upload it to the plant data center through an encrypted transmission channel, and synchronize it to the company's big data center in real time to generate a stored feature set C, where C represents the stored serialized dataset;

[0010] S4: Perform partial discharge analysis on the feature values ​​in the stored feature set C, calculate the phase-resolved partial discharge map PRPD and the phase-resolved pulse sequence map PRPS, combine with dielectric response analysis to generate a dielectric spectrum, and generate a visualization spectrum set D, where D represents a spectrum set containing PRPD, PRPS and dielectric spectrum.

[0011] S5: Based on the knowledge graph, perform fault diagnosis on the visualization graph set D. The knowledge graph uses equipment components as nodes E={e_1,e_2,…,e_m} and electrical connections and physical adjacencies as edges F. Combine PRPD and PRPS feature values ​​to determine the fault location, fault type, gas chamber status and decomposition product concentration, and generate a diagnostic result set G, where G represents a structured set containing fault information.

[0012] S6: Assess insulation condition risk based on diagnostic results, analyze defect trends, safety risks and economic risks, calculate risk values ​​using FMEA or FTA methods, and output alarm levels and handling recommendations.

[0013] Preferably, the sub-step of S1 is as follows:

[0014] S1.1: Through online monitoring and live testing or power outage testing, voltage signals, current signals, temperature data, vibration characteristics, ultra-high frequency signals and ultrasonic signals are collected from the GIL device to generate an initial dataset;

[0015] S1.2: Preprocessing techniques are used to denoise and standardize the initial dataset to obtain dataset Q. Denoising is accomplished through wavelet transform, and standardization is achieved using mean normalization.

[0016] S1.3: If the outlier values ​​of voltage and current signals in dataset Q exceed the preset threshold, the Pearson correlation coefficient between the voltage and current signals is calculated through correlation analysis to determine whether there are potential defects in the insulation condition.

[0017] S1.4: Based on the obtained dataset Q, extract temperature data and vibration features, and use principal component analysis algorithm to reduce the dimensionality of temperature data and vibration features to obtain dataset R;

[0018] S1.5: If the variance of the eigenvectors of UHF signals and ultrasonic signals in dataset R is lower than a preset threshold, then cluster analysis is used to classify the UHF signals and ultrasonic signals to determine the category of insulation state.

[0019] S1.6: Obtain Pearson correlation coefficient and category information, and use the support vector machine algorithm to classify and train the dataset Q to obtain a prediction model of insulation state;

[0020] S1.7: Test the dataset R with the prediction model to obtain the classification results of the insulation status of GIL equipment.

[0021] Preferably, the sub-step of S2 is as follows:

[0022] S2.1: The original feature set A is processed by photoelectric conversion, and the initial signal data is obtained by using a photoelectric sensor to obtain the converted signal S;

[0023] S2.2: Filter the converted signal S by using a low-pass filter to remove high-frequency noise and obtain the filtered signal T;

[0024] S2.3: The filtered signal T is amplified by using an operational amplifier to increase the signal strength, resulting in the amplified signal U;

[0025] S2.4: Perform time-domain analysis on the amplified signal U, calculate the time-series statistics of the signal, and obtain the vibration amplitude characteristics V;

[0026] S2.5: Perform frequency domain analysis on the amplified signal U, extract the frequency components through fast Fourier transform, and obtain the ultra-high frequency signal characteristics W and ultrasonic signal characteristics X;

[0027] S2.6: If the integrity of the vibration amplitude feature V, the ultra-high frequency signal feature W, and the ultrasonic signal feature X meets the preset threshold, then a structured feature set B is generated through the feature fusion algorithm;

[0028] S2.7: If not satisfied, return to the signal filtering step for reprocessing and determine the integrity of the feature set;

[0029] S2.8: Standardize the structured feature set B using the z-score normalization method to obtain the standardized dataset Y.

[0030] Preferably, the sub-step of S3 is as follows:

[0031] S3.1: Obtain the raw data from the structured feature set, use a predefined serialization format to generate a JSON data packet, and obtain the serialized dataset;

[0032] S3.2: Encrypt the JSON data packet using an encryption algorithm to generate an encrypted data packet and ensure transmission security;

[0033] S3.3: If the integrity verification of the encrypted data packet passes, it will be transmitted to the plant data center through the encrypted channel to obtain the transmission completion status;

[0034] S3.4: Receive encrypted data packets from the plant data center, decrypt them to generate a storage feature set, and obtain the storage dataset;

[0035] S3.5: If the stored feature set is consistent with the serialized dataset, it is sent to the big data center through the synchronization transmission protocol to determine the synchronization completion status;

[0036] S3.6: Based on the storage feature set received by the big data center, a data verification algorithm is used to determine data integrity;

[0037] S3.7: If the data integrity verification passes, the feature set is stored in the big data center to obtain the final storage status.

[0038] Preferably, the sub-step of S4 is as follows:

[0039] S4.1: Obtain feature values ​​from the stored feature set C, calculate phase-resolved features through partial discharge analysis, and obtain the initial data for the PRPD and PRPS maps;

[0040] S4.2: The initial data of the PRPD map is processed by Fourier transform to extract the phase-resolved discharge characteristics and obtain the frequency domain distribution of the discharge map;

[0041] S4.3: Process the initial data of the PRPS diagram through time series analysis to determine the time distribution characteristics of the pulse sequence and obtain the time domain characteristics of the pulse sequence diagram;

[0042] S4.4: Extract dielectric response data from eigenvalues ​​and use dielectric spectrum analysis algorithm to generate the frequency domain response distribution of dielectric spectrum;

[0043] S4.5: If the frequency domain distribution of the PRPD map matches the preset discharge characteristic threshold, then combine the time domain characteristics of the PRPS map to generate a joint feature set of the discharge mode;

[0044] S4.6: Obtain the joint feature set and the frequency domain response distribution of the dielectric spectrum, and use the data fusion algorithm to obtain a visualization spectrum set D containing the PRPD map, PRPS map and dielectric spectrum;

[0045] S4.7: Through visualization processing, the atlas set D is converted into a two-dimensional image output to obtain the final visualization result.

[0046] Preferably, the sub-step of S5 is as follows:

[0047] S5.1: By parsing the visualization graph set D, the device component nodes E={e_1, e_2, ..., e_m} and electrical connections and physical adjacency edges F are obtained from the knowledge graph. The topology relation matrix Z is constructed to obtain the connection characteristics between device components.

[0048] S5.2: Based on the topological relationship matrix Z, the PRPD feature values ​​are fused, and the k-nearest neighbor algorithm is used to calculate the abnormal probability P(e_i) of each equipment component. If P(e_i) is greater than the preset threshold T_1, then component e_i is determined as a potential fault location, and the fault location set α is obtained.

[0049] S5.3: For each component e_i in the fault location set α, extract the corresponding PRPS feature value, use the support vector machine algorithm to classify the fault mode, determine the fault type as breakdown and discharge or overheating, and obtain the fault type set β;

[0050] S5.4: Obtain the air chamber state parameter S(e_i) of component e_i from the fault location set α. If S(e_i) deviates from the preset threshold range T_2, the air chamber state is judged to be abnormal, and the air chamber state set γ is obtained.

[0051] S5.5: Based on the fault type set β and the gas cell state set γ, analyze the characteristics of the decomposition products. If the fault type is discharge and the gas cell state is abnormal, calculate the concentration of the decomposition products using spectral analysis technology to obtain the set of decomposition product concentrations δ.

[0052] S5.6: By fusing the fault location set α, the fault type set β, the gas chamber state set γ, and the decomposition product concentration set δ, a structured fault diagnosis result set G is generated, where G={α, β, γ, δ};

[0053] S5.7: If there are multiple component fault associations in the fault diagnosis result set G, the fault propagation path is updated using a graph convolutional network based on the electrical connection edges of the knowledge graph to obtain the final diagnosis result set G.

[0054] Preferably, the sub-step of S6 is as follows:

[0055] S6.1: Based on the diagnostic result set G, perform fault mode and impact analysis to identify each fault mode and its impact on system performance and component operation. Combine the edges F of the knowledge graph to generate a component-related impact set H, where H represents the mapping set between fault modes and affected components.

[0056] S6.2: Based on the component-related impact set H, perform fault tree analysis, define insulation failure as the top event, take the fault modes in H as the basic events, construct the causal propagation path, and generate the fault propagation graph I, where I represents the set of logical relationships from component failure to system failure;

[0057] S6.3: Based on the diagnostic result set G and the fault propagation map I, quantify the defect trend, safety risk and economic risk of the insulation state, calculate the comprehensive risk value J, where J represents the risk quantification value, and generate alarm level K and handling suggestions L, where K represents the risk level and L represents the set of handling measures.

[0058] S6.4: If the alarm level K is higher than the preset threshold M, where M represents the risk trigger threshold, then perform convolutional neural network analysis on the UHF signal and ultrasonic signal in the structured feature set B. The convolution kernel is based on the pre-trained model to generate the fault location classification probability N, where N represents the location verification probability, and update the fault location in the diagnostic result set G.

[0059] S6.5: Based on the fault location in the updated diagnostic result set G, query the associated equipment components through the knowledge graph, define the edge F as the electrical dependency, and generate the fault impact range O, where O represents the set of components {e_i,e_j,...} affected by the fault;

[0060] S6.6: Based on the fault impact range O, the diagnostic result set G, and the alarm level K, generate a diagnostic report P according to predefined alarm rules. P includes the fault type, fault location, impact range, and handling suggestions. P is then transmitted to the monitoring system to complete the insulation status monitoring.

[0061] Preferably, the specific method in S6.1 is as follows:

[0062] Fault mode data is obtained from the diagnostic result set, and feature information is extracted to form a fault mode set. For this set, combined with system performance parameters and component operating status, an impact analysis algorithm is used to determine the impact range of each fault mode on the system and components. With the help of a knowledge graph, the component association relationship is obtained based on edge F, generating a preliminary association impact set between fault modes and components. If there are multiple paths of influence in the preliminary set, a depth-first search algorithm is used to traverse edge F to determine the propagation path, resulting in an optimized association impact set. Based on the optimized set, a mapping relationship between fault modes and affected components is constructed to form a component mapping set. The identification information of affected components is extracted from the component mapping set, and the priority is determined based on the system performance change trend to obtain the final component association impact set H. The consistency between H and the diagnostic result set is verified by data comparison to determine the accuracy of H and output the final mapping set.

[0063] Preferably, the specific method in S6.2 is as follows:

[0064] Obtain the component-related impact set H, extract failure modes from it to form a basic event set ε; use fault tree analysis to set insulation failure as the top event ζ, and construct a fault tree structure η based on ε; analyze the causal propagation relationship between the basic events in η through Boolean algebra operations to generate causal propagation paths θ; calculate the logical relationship strength ι of each path based on θ to obtain the fault propagation map I; if a path in I exceeds a preset threshold, mark this path as a high-risk path κ; extract the key basic event λ with the help of κ to generate an optimized fault tree μ; finally, output the system failure logic relationship set ν based on μ, showing the complete causal chain from key component failure to insulation failure.

[0065] Preferably, the specific method in S6.3 is as follows:

[0066] Insulation status data is extracted from the diagnostic result set G, and statistical analysis is used to obtain defect characteristics and trends. Combining the defect trends and fault propagation map I, graph analysis algorithms are used to calculate the defect propagation path and identify the distribution of potential fault points. If the distribution of potential fault points exceeds a preset threshold, a safety risk value ξ is obtained through a safety risk model, thus quantifying the safety risk. An economic loss assessment model is used, combined with the defect trends, to calculate the economic risk value ο, achieving economic risk quantification. A weighted fusion method is employed to combine the safety risk value ξ and the economic risk value ο to obtain a comprehensive risk value J, determining the comprehensive risk quantification index. If the comprehensive risk value J exceeds a preset threshold, an alarm level K is generated according to risk grading rules. Finally, based on the alarm level K, a corresponding set of handling measures L is extracted from the preset measures library to generate suggested measures.

[0067] The present invention can achieve the following beneficial effects:

[0068] 1. By integrating multiple sensing technologies and the data processing capabilities of the industrial internet platform, this method can achieve real-time and comprehensive monitoring of the insulation status of GIL equipment, thereby improving the accuracy and efficiency of detection.

[0069] 2. By utilizing the big data analytics capabilities of the industrial internet platform, this method can perform in-depth analysis on the large amount of data collected, thereby more accurately predicting potential faults and insulation degradation trends.

[0070] 3. Through continuous monitoring and evaluation of GIL equipment, this method helps to detect and address early failures in a timely manner, avoid large-scale equipment damage, and thus significantly reduce maintenance costs and repair expenses.

[0071] 4. This method can assess the insulation status of GIL equipment in real time, promptly identify potential safety hazards, reduce unexpected power outages, and thus improve the reliability and stability of the entire power system.

[0072] 5. By combining the intelligent decision support system of the industrial internet platform, this method can automatically generate maintenance and repair suggestions, realizing intelligent management of GIL equipment.

[0073] 6. Through effective monitoring and timely maintenance, this method helps extend the service life of GIL equipment, reduce the frequency of equipment replacement, and lower long-term operating costs. Attached Figure Description

[0074] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0075] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0076] Preferred solutions include Figure 1As shown, a method for detecting and evaluating the insulation status of GIL (Gas Insulation Line) equipment in hydropower stations based on an Industrial Internet platform is a strategy for conducting insulation status detection and evaluation of GIL equipment in hydropower stations based on Industrial Internet platform technology. This method can utilize relevant components, algorithms, and data of the Industrial Internet platform to achieve more efficient equipment insulation status detection and evaluation. The structure includes a signal sensing layer, a front-end processing layer, a signal transmission layer, a data processing layer, a diagnostic model layer, and a status evaluation layer. The signal sensing layer is responsible for collecting characteristic parameters reflecting the insulation status of the GIL equipment, including but not limited to equipment voltage and current, temperature and humidity, vibration, UHF, ultrasound, gas decomposition products, gas density, capacitance, dielectric spectrum, insulation resistance, and DC resistance. Collection methods include online monitoring systems, live-line testing devices, and high-voltage power outage tests. The front-end processing layer performs preliminary processing on the collected signals, including photoelectric conversion, filtering, amplification, time-domain analysis, and frequency-domain analysis, to generate data such as vibration amplitude, UHF, and ultrasound. The signal transmission layer uploads the front-end processed data to the plant data center according to a specified standard format, and then from the plant data center to the company's big data center. The data processing layer... The uploaded data is processed to generate various graphs, such as PRPD, PRPS, FPRPD, original waveforms, dielectric spectra, and correlation comparison graphs. The diagnostic model layer uses knowledge graph technology to diagnose defects in the generated graphs, including the location of the discharge source, fault type, location in the gas chamber, gas chamber pressure change trend, discharge source development trend, and concentration of gas chamber decomposition products. The status assessment layer uses the diagnosed defect location, fault type, trend, cause, safety risk, and economic risk information to assess the insulation status and output handling suggestions. FMEA or FTA are used to calculate the insulation status risk value of the GIL equipment, output alarm levels, and provide handling suggestions for different levels. This invention provides a specific method for detecting and assessing the insulation status of GIL equipment in hydropower stations. Through the above technical solution, comprehensive detection and assessment of the insulation status of GIL equipment is achieved, effectively improving the safety and stability of the power system. The specific implementation steps are as follows:

[0077] Signal sensing layer: S1: Collect characteristic parameters of the insulation state of GIL equipment through online monitoring system, live detection device or high voltage test during power outage, including voltage, current, temperature, vibration, UHF signal and ultrasonic signal, and generate original feature set A, where A represents the initial dataset containing all collected signals.

[0078] S1.1: Through online monitoring and live testing or power outage testing, voltage signals, current signals, temperature data, vibration characteristics, ultra-high frequency signals and ultrasonic signals are collected from the GIL device to generate an initial dataset.

[0079] S1.2: Preprocessing techniques are used to denoise and standardize the initial dataset to obtain dataset Q. Denoising is accomplished through wavelet transform, and standardization is achieved using mean normalization.

[0080] S1.3: If the outlier values ​​of voltage and current signals in dataset Q exceed the preset threshold, the Pearson correlation coefficient between the voltage and current signals is calculated through correlation analysis to determine whether there are potential defects in the insulation status.

[0081] S1.4: Based on the obtained dataset Q, extract temperature data and vibration features, and use principal component analysis algorithm to reduce the dimensionality of temperature data and vibration features to obtain dataset R.

[0082] S1.5: If the variance of the eigenvectors of UHF signals and ultrasonic signals in dataset R is lower than a preset threshold, then cluster analysis is used to classify the UHF signals and ultrasonic signals to determine the category of insulation state.

[0083] S1.6: Obtain the Pearson correlation coefficient and category information, and use the support vector machine algorithm to classify and train the dataset Q to obtain a prediction model of the insulation state.

[0084] S1.7: Test the dataset R with the prediction model to obtain the classification results of the insulation status of GIL equipment.

[0085] Front-end processing layer: S2: Performs photoelectric conversion, filtering and signal amplification on the original feature set A, and extracts vibration amplitude, UHF signal feature value and ultrasonic signal feature value by combining time domain analysis and frequency domain analysis to generate structured feature set B, where B represents a standardized dataset containing the extracted features.

[0086] S2.1: The original feature set A is processed by photoelectric conversion, and the initial signal data is obtained by using a photoelectric sensor to obtain the converted signal S.

[0087] S2.2: Filter the converted signal S by using a low-pass filter to remove high-frequency noise and obtain the filtered signal T.

[0088] S2.3: Amplify the filtered signal T by using an operational amplifier to increase the signal strength, resulting in the amplified signal U.

[0089] S2.4: Perform time-domain analysis on the amplified signal U, calculate the time-series statistics of the signal, and obtain the vibration amplitude characteristic V.

[0090] S2.5: Perform frequency domain analysis on the amplified signal U, extract the frequency components through fast Fourier transform, and obtain the ultra-high frequency signal characteristics W and ultrasonic signal characteristics X.

[0091] S2.6: If the integrity of the vibration amplitude feature V, the ultra-high frequency signal feature W, and the ultrasonic signal feature X meets the preset threshold, then a structured feature set B is generated through the feature fusion algorithm.

[0092] S2.7: If not satisfied, return to the signal filtering step for reprocessing and determine the integrity of the feature set.

[0093] S2.8: Standardize the structured feature set B using the z-score normalization method to obtain the standardized dataset Y.

[0094] Signal transmission layer: S3: Serializes the structured feature set B into a JSON data packet according to a predefined format, uploads it to the plant data center through an encrypted transmission channel, and synchronizes it to the company's big data center in real time to generate a stored feature set C, where C represents the stored serialized dataset.

[0095] S3.1: Obtain the raw data from the structured feature set, use a predefined serialization format to generate a JSON data packet, and obtain the serialized dataset.

[0096] S3.2: Encrypt the JSON data packet using an encryption algorithm to generate an encrypted data packet and ensure transmission security.

[0097] S3.3: If the integrity verification of the encrypted data packet passes, it will be transmitted to the plant data center through the encrypted channel to obtain the transmission completion status.

[0098] S3.4: Receive encrypted data packets from the plant data center, decrypt them to generate a storage feature set, and obtain the storage dataset.

[0099] S3.5: If the stored feature set is consistent with the serialized dataset, it is sent to the big data center through the synchronization transmission protocol to determine the synchronization completion status.

[0100] S3.6: Based on the storage feature set received by the big data center, a data verification algorithm is used to determine data integrity.

[0101] S3.7: If the data integrity verification passes, the feature set is stored in the big data center to obtain the final storage status.

[0102] Data processing layer: S4: Perform partial discharge analysis on the feature values ​​in the stored feature set C, calculate the phase-resolved partial discharge map PRPD and the phase-resolved pulse sequence map PRPS, combine with dielectric response analysis to generate dielectric spectrum, and generate a visualization spectrum set D, where D represents the spectrum set containing PRPD, PRPS and dielectric spectrum.

[0103] S4.1: Obtain feature values ​​from the stored feature set C, calculate phase-resolved features through partial discharge analysis, and obtain the initial data for the PRPD and PRPS maps.

[0104] S4.2: The initial data of the PRPD map is processed by Fourier transform to extract the phase-resolved discharge features and obtain the frequency domain distribution of the discharge map.

[0105] S4.3: Process the initial data of the PRPS diagram through time series analysis to determine the time distribution characteristics of the pulse sequence and obtain the time domain characteristics of the pulse sequence diagram.

[0106] S4.4: Extract dielectric response data from eigenvalues ​​and use a dielectric spectrum analysis algorithm to generate the frequency domain response distribution of the dielectric spectrum.

[0107] S4.5: If the frequency domain distribution of the PRPD map matches the preset discharge characteristic threshold, then a joint feature set of the discharge mode is generated by combining the time domain characteristics of the PRPS map.

[0108] S4.6: Obtain the joint feature set and the frequency domain response distribution of the dielectric spectrum, and use the data fusion algorithm to obtain a visualization spectrum set D containing the PRPD map, PRPS map and dielectric spectrum.

[0109] S4.7: Through visualization processing, the atlas set D is converted into a two-dimensional image output to obtain the final visualization result.

[0110] Diagnostic Model Layer: S5: Based on the knowledge graph, fault diagnosis is performed on the visualization graph set D. The knowledge graph uses equipment components as nodes E={e_1,e_2,…,e_m} and electrical connections and physical adjacencies as edges F. Combining PRPD and PRPS feature values, the fault location, fault type, gas chamber state and decomposition product concentration are determined, and a diagnostic result set G is generated, where G represents a structured set containing fault information.

[0111] S5.1: By parsing the visualization graph set D, the device component nodes E={e_1, e_2, ..., e_m} and electrical connections and physical adjacency edges F are obtained from the knowledge graph. The topology relation matrix Z is constructed to obtain the connection characteristics between device components.

[0112] S5.2: Based on the topological relationship matrix Z, the PRPD feature values ​​are fused, and the k-nearest neighbor algorithm is used to calculate the abnormal probability P(e_i) of each equipment component. If P(e_i) is greater than the preset threshold T_1, then component e_i is determined as a potential fault location, and the fault location set α is obtained.

[0113] S5.3: For each component e_i in the fault location set α, extract the corresponding PRPS feature value, use the support vector machine algorithm to classify the fault mode, determine the fault type as breakdown and discharge or overheating, and obtain the fault type set β.

[0114] S5.4: Obtain the chamber state parameter S(e_i) of component e_i from the fault location set α. If S(e_i) deviates from the preset threshold range T_2, the chamber state is judged to be abnormal, and the chamber state set γ is obtained. S5.5: Analyze the characteristics of decomposition products based on the fault type set β and the chamber state set γ. If the fault type is discharge and the chamber state is abnormal, calculate the concentration of decomposition products using spectral analysis technology to obtain the decomposition product concentration set δ.

[0115] S5.6: By fusing the fault location set α, the fault type set β, the gas chamber state set γ, and the decomposition product concentration set δ, a structured fault diagnosis result set G is generated, where G={α, β, γ, δ}.

[0116] S5.7: If there are multiple component fault associations in the fault diagnosis result set G, the fault propagation path is updated using a graph convolutional network based on the electrical connection edges of the knowledge graph to obtain the final diagnosis result set G.

[0117] Condition assessment layer: S6: Based on the diagnostic results, assess the risk of insulation condition, analyze defect trends, safety risks and economic risks, calculate risk values ​​through methods such as FMEA or FTA, and output alarm levels and handling suggestions.

[0118] S6.1: Based on the diagnostic result set G, perform Failure Mode and Effects Analysis (FMEA) to identify each failure mode and its impact on system performance and component operation. Combine the edges F of the knowledge graph to generate a component-related impact set H, where H represents the mapping set between failure modes and affected components.

[0119] S6.1.1: Obtain fault mode data from the diagnostic result set, and extract the feature information of each fault mode by parsing to obtain a fault mode set.

[0120] S6.1.2: For the set of failure modes, an impact analysis algorithm is used to determine the impact range of each failure mode on system performance and component operation by combining system performance parameters and component operating status.

[0121] S6.1.3: Through the knowledge graph, obtain the inter-component relationships represented by graph edge F, and generate a preliminary set of association effects between failure modes and components.

[0122] S6.1.4: If there are multiple paths of influence in the initial set of associated influences, then the depth-first search algorithm is used to traverse the graph edges F, determine the propagation path of the failure mode, and obtain the optimized set of associated influences.

[0123] S6.1.5: Based on the optimized set of associated effects, construct the mapping relationship between failure modes and affected components, and generate a set of component mappings.

[0124] S6.1.6: Extract the identification information of the affected components from the component mapping set, combine it with the system performance change trend, determine the priority of the affected components, and obtain the final component association influence set H.

[0125] S6.1.7: By verifying the consistency between the component association influence set H and the diagnostic result set, the accuracy of H is determined using a data comparison method, and the final mapping set is output.

[0126] S6.2: Based on the component-related impact set H, perform fault tree analysis (FTA), define insulation failure as the top event, take the failure modes in H as the base events, construct the causal propagation path, and generate the fault propagation graph I, where I represents the set of logical relationships from component failure to system failure.

[0127] S6.2.1: Obtain the component-related impact set H, extract the failure modes from it, and generate the basic event set ε, where ε represents all initial failure points that may lead to system failure.

[0128] S6.2.2: Using the fault tree analysis method, insulation failure is set as the top event ζ, and a fault tree structure η is constructed based on ε, where η represents the logical connection from the basic event to the top event.

[0129] S6.2.3: Using Boolean algebra operations, analyze the causal propagation relationship between basic events in η and generate a causal propagation path θ, where θ represents the order in which the fault is transmitted from the component to the system.

[0130] S6.2.4: Based on θ, calculate the logical relationship strength ι for each path, where ι represents the probability weight of fault propagation, and obtain the fault propagation map I.

[0131] S6.2.5: If the value of ι in a certain path in I exceeds the preset threshold, then the path is marked as a high-risk path κ, where κ represents the critical propagation chain of system failure.

[0132] S6.2.6: Extract key basic events λ through κ, where λ represents the failure mode that has the greatest impact on system failure, and generate an optimized fault tree μ.

[0133] S6.2.7: Based on μ, output the set of logical relationships ν of system failure, where ν represents the complete causal chain from critical component failure to insulation failure.

[0134] S6.3: Based on the diagnostic result set G and the fault propagation map I, quantify the defect trend, safety risk and economic risk of the insulation state, calculate the comprehensive risk value J, where J represents the risk quantification value, and generate alarm level K and handling suggestions L, where K represents the risk level and L represents the set of handling measures.

[0135] S6.3.1: Obtain insulation status data from the diagnostic result set G, extract defect features using statistical analysis methods, and obtain the defect change trend.

[0136] S6.3.2: Based on the defect change trend and combined with the fault propagation map I, a graph analysis algorithm is used to calculate the defect propagation path and determine the distribution of potential fault points.

[0137] S6.3.3: If the distribution of potential fault points exceeds the preset threshold, the safety risk value ξ is calculated through the safety risk model to obtain the safety risk quantification result.

[0138] S6.3.4: By using the economic loss assessment model and combining the defect change trend, calculate the economic risk value ο to obtain the quantitative result of economic risk.

[0139] S6.3.5: Using a weighted fusion method, combining the safety risk value ξ and the economic risk value ο, calculate the comprehensive risk value J and determine the comprehensive risk quantification index.

[0140] S6.3.6: If the comprehensive risk value J exceeds the preset threshold, an alarm level K is generated according to the risk grading rules to obtain the risk level.

[0141] S6.3.7: Based on the alarm level K, extract the corresponding set of processing measures L from the preset measure library and generate suggested measures.

[0142] S6.4: If the alarm level K is higher than the preset threshold M, where M represents the risk trigger threshold, then perform convolutional neural network analysis on the UHF signal and ultrasonic signal in the structured feature set B. The convolution kernel is based on the pre-trained model to generate the fault location classification probability N, where N represents the location verification probability, and update the fault location in the diagnostic result set G.

[0143] S6.4.1: If the alarm level K exceeds the risk threshold M, where M represents the preset risk trigger threshold, then the UHF signal and ultrasonic signal are extracted from the structured feature set B to obtain the initial signal set π.

[0144] S6.4.2: By preprocessing the initial signal set π, a normalization operation is performed on the UHF signal and the ultrasonic signal using a normalization method to obtain the normalized signal set ρ.

[0145] S6.4.3: Use pre-trained convolutional kernels to perform convolutional neural network analysis on the standardized signal set ρ to generate fault location classification probability N, where N represents the location verification probability, and obtain the fault location distribution σ.

[0146] S6.4.4: If the maximum probability value in the fault location distribution σ is higher than the preset threshold T_loc, where T_loc represents the location confirmation threshold, then the location corresponding to the highest probability is extracted from the fault location distribution σ, and the fault location in the diagnosis result set G is updated to obtain the updated diagnosis set τ.

[0147] S6.4.5: Based on the updated diagnostic set τ, a clustering algorithm is used to group the fault locations to generate a fault region set υ, where υ represents the spatial distribution of the fault.

[0148] S6.4.6: By performing feature mapping on the fault region set υ, key features are extracted using principal component analysis to generate feature vector φ, where φ represents the dimensionality-reduced representation of the fault region.

[0149] S6.4.7: If the similarity between the feature vector φ and the historical fault database H_lib is higher than the preset threshold T_sim, where T_sim represents the fault matching threshold, then the corresponding fault mode is obtained from the historical fault database H_lib, the fault region set υ is updated, and the final fault distribution χ is obtained.

[0150] S6.5: Based on the fault location in the updated diagnostic result set G, query the associated equipment components through the knowledge graph. The edge F is defined as an electrical dependency relationship, and the fault impact range O is generated, where O represents the set of components {e_i,e_j,...} affected by the fault.

[0151] S6.5.1: Obtain the fault location identifier f_id from the diagnostic result set G.

[0152] S6.5.2: If f_id has associated nodes in the knowledge graph, then by querying the knowledge graph, extract the electrical dependency edges F that are directly connected to f_id, and obtain the initial set of associated components ψ={e_i, e_j, ...}.

[0153] S6.5.3: Based on the initial set of associated components ψ, a depth-first search algorithm is used to traverse the electrical dependency edges F in the knowledge graph related to each component e_i in ψ, and obtain the set of indirectly associated components ω={e_k, e_l, ...}.

[0154] S6.5.4: By merging the initial set of associated components ψ and the set of indirectly associated components ω, a merged component set ξ_merge={e_i, e_j, e_k, e_l, ...} is generated.

[0155] S6.5.5: For each component e_m in the merged component set ξ_merge, if the weight w_m of the electrical dependency edge F of e_m is greater than a preset threshold, then e_m is determined to be a component affected by the fault, and the candidate affected component set ζ_cand={e_m, e_n, ...} is obtained.

[0156] S6.5.6: Obtain the identifier and status information of each component e_n from the candidate affected component set ζ_cand, and generate the final affected component set O={e_i, e_j, ...} through set deduplication.

[0157] S6.5.7: Based on the set O of components affected by the fault, the topological sorting algorithm is used to arrange the components in O in a dependent order to obtain an ordered sequence of affected components η_seq={e_i->e_j->…}.

[0158] S6.5.8: Generate structured data θ_struct of the fault impact range by using the ordered sequence of affected components η_seq, which includes the identifier and status of each component and the order of dependencies, and determine the final output.

[0159] S6.6: Based on the fault impact range O, the diagnostic result set G, and the alarm level K, generate a diagnostic report P according to predefined alarm rules. P includes the fault type, fault location, impact range, and handling suggestions. P is then transmitted to the monitoring system to complete the insulation status monitoring.

[0160] S6.6.1: Obtain the abnormal data area from the fault impact range O, and determine the preliminary fault characteristics by comparing with predefined rules.

[0161] S6.6.2: Based on the preliminary fault characteristics, extract relevant fault modes from the diagnostic result set G, and use a classification algorithm to determine the fault type.

[0162] S6.6.3: If the fault type is confirmed, the location parameters are extracted from the abnormal data area to obtain the fault location.

[0163] S6.6.4: Based on the fault type and fault location, combined with the alarm level K, rule matching is used to obtain handling suggestions.

[0164] S6.6.5: Generate a diagnostic report P based on the handling recommendations, which includes the fault type, fault location, scope of impact, and handling recommendations.

[0165] S6.6.6: Extract structured data from the diagnostic report P and transmit it to the monitoring system via a data transmission protocol.

[0166] S6.6.7: Obtain feedback from the monitoring system and update the insulation status using a status monitoring algorithm.

[0167] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A method for detecting and evaluating the insulation status of GIL equipment in hydropower stations based on an industrial internet platform, characterized in that... Includes the following steps: S1: Collect characteristic parameters of the insulation state of GIL equipment through online monitoring system, live detection device or high voltage test during power outage, including voltage, current, temperature, vibration, UHF signal and ultrasonic signal, and generate original feature set A, where A represents the initial dataset containing all collected signals; S2: Perform photoelectric conversion, filtering and signal amplification on the original feature set A, and combine time domain analysis and frequency domain analysis to extract vibration amplitude, UHF signal feature value and ultrasonic signal feature value to generate structured feature set B, where B represents a standardized dataset containing the extracted features; S3: Serialize the structured feature set B into a JSON data packet according to a predefined format, upload it to the plant data center through an encrypted transmission channel, and synchronize it to the company's big data center in real time to generate a stored feature set C, where C represents the stored serialized dataset; S4: Perform partial discharge analysis on the feature values ​​in the stored feature set C, calculate the phase-resolved partial discharge map PRPD and the phase-resolved pulse sequence map PRPS, combine with dielectric response analysis to generate a dielectric spectrum, and generate a visualization spectrum set D, where D represents a spectrum set containing PRPD, PRPS and dielectric spectrum. S5: Based on the knowledge graph, perform fault diagnosis on the visualization graph set D. The knowledge graph uses equipment components as nodes E={e_1,e_2,…,e_m} and electrical connections and physical adjacencies as edges F. Combine PRPD and PRPS feature values ​​to determine the fault location, fault type, gas chamber status and decomposition product concentration, and generate a diagnostic result set G, where G represents a structured set containing fault information. S6: Assess insulation condition risk based on diagnostic results, analyze defect trends, safety risks and economic risks, calculate risk values ​​using FMEA or FTA methods, and output alarm levels and handling recommendations.

2. The method for detecting and evaluating the insulation status of GIL equipment in hydropower stations based on an industrial internet platform according to claim 1, characterized in that: The sub-steps of S1 are: S1.1: Through online monitoring and live testing or power outage testing, voltage signals, current signals, temperature data, vibration characteristics, ultra-high frequency signals and ultrasonic signals are collected from the GIL device to generate an initial dataset; S1.2: Preprocessing techniques are used to denoise and standardize the initial dataset to obtain dataset Q. Denoising is accomplished through wavelet transform, and standardization is achieved using mean normalization. S1.3: If the outlier values ​​of voltage and current signals in dataset Q exceed the preset threshold, the Pearson correlation coefficient between the voltage and current signals is calculated through correlation analysis to determine whether there are potential defects in the insulation condition. S1.4: Based on the obtained dataset Q, extract temperature data and vibration features, and use principal component analysis algorithm to reduce the dimensionality of temperature data and vibration features to obtain dataset R; S1.5: If the variance of the eigenvectors of UHF signals and ultrasonic signals in dataset R is lower than a preset threshold, then cluster analysis is used to classify the UHF signals and ultrasonic signals to determine the category of insulation state. S1.6: Obtain Pearson correlation coefficient and category information, and use the support vector machine algorithm to classify and train the dataset Q to obtain a prediction model of insulation state; S1.7: Test the dataset R with the prediction model to obtain the classification results of the insulation status of GIL equipment.

3. The method for detecting and evaluating the insulation status of GIL equipment in hydropower stations based on an industrial internet platform according to claim 1, characterized in that: The sub-steps of S2 are: S2.1: The original feature set A is processed by photoelectric conversion, and the initial signal data is obtained by using a photoelectric sensor to obtain the converted signal S; S2.2: Filter the converted signal S by using a low-pass filter to remove high-frequency noise and obtain the filtered signal T; S2.3: The filtered signal T is amplified by using an operational amplifier to increase the signal strength, resulting in the amplified signal U; S2.4: Perform time-domain analysis on the amplified signal U, calculate the time-series statistics of the signal, and obtain the vibration amplitude characteristics V; S2.5: Perform frequency domain analysis on the amplified signal U, extract the frequency components through fast Fourier transform, and obtain the ultra-high frequency signal characteristics W and ultrasonic signal characteristics X; S2.6: If the integrity of the vibration amplitude feature V, the ultra-high frequency signal feature W, and the ultrasonic signal feature X meets the preset threshold, then a structured feature set B is generated through the feature fusion algorithm; S2.7: If not satisfied, return to the signal filtering step for reprocessing and determine the integrity of the feature set; S2.8: Standardize the structured feature set B using the z-score normalization method to obtain the standardized dataset Y.

4. The method for detecting and evaluating the insulation status of GIL equipment in hydropower stations based on an industrial internet platform according to claim 1, characterized in that: The sub-steps of S3 are: S3.1: Obtain the raw data from the structured feature set, use a predefined serialization format to generate a JSON data packet, and obtain the serialized dataset; S3.2: Encrypt the JSON data packet using an encryption algorithm to generate an encrypted data packet and ensure transmission security; S3.3: If the integrity verification of the encrypted data packet passes, it will be transmitted to the plant data center through the encrypted channel to obtain the transmission completion status; S3.4: Receive encrypted data packets from the plant data center, decrypt them to generate a storage feature set, and obtain the storage dataset; S3.5: If the stored feature set is consistent with the serialized dataset, it is sent to the big data center through the synchronization transmission protocol to determine the synchronization completion status; S3.6: Based on the storage feature set received by the big data center, a data verification algorithm is used to determine data integrity; S3.7: If the data integrity verification passes, the feature set is stored in the big data center to obtain the final storage status.

5. The method for detecting and evaluating the insulation status of GIL equipment in hydropower stations based on an industrial internet platform according to claim 1, characterized in that: The sub-steps of S4 are: S4.1: Obtain feature values ​​from the stored feature set C, calculate phase-resolved features through partial discharge analysis, and obtain the initial data for the PRPD and PRPS maps; S4.2: The initial data of the PRPD map is processed by Fourier transform to extract the phase-resolved discharge characteristics and obtain the frequency domain distribution of the discharge map; S4.3: Process the initial data of the PRPS diagram through time series analysis to determine the time distribution characteristics of the pulse sequence and obtain the time domain characteristics of the pulse sequence diagram; S4.4: Extract dielectric response data from eigenvalues ​​and use dielectric spectrum analysis algorithm to generate the frequency domain response distribution of dielectric spectrum; S4.5: If the frequency domain distribution of the PRPD map matches the preset discharge characteristic threshold, then combine the time domain characteristics of the PRPS map to generate a joint feature set of the discharge mode; S4.6: Obtain the joint feature set and the frequency domain response distribution of the dielectric spectrum, and use the data fusion algorithm to obtain a visualization spectrum set D containing the PRPD map, PRPS map and dielectric spectrum; S4.7: Through visualization processing, the atlas set D is converted into a two-dimensional image output to obtain the final visualization result.

6. The method for detecting and evaluating the insulation status of GIL equipment in a hydropower station based on an industrial internet platform according to claim 1, characterized in that: The sub-steps of S5 are: S5.1: By parsing the visualization graph set D, the device component nodes E={e_1, e_2, ..., e_m} and electrical connections and physical adjacency edges F are obtained from the knowledge graph. The topology relation matrix Z is constructed to obtain the connection characteristics between device components. S5.2: Based on the topological relationship matrix Z, the PRPD feature values ​​are fused, and the k-nearest neighbor algorithm is used to calculate the abnormal probability P(e_i) of each equipment component. If P(e_i) is greater than the preset threshold T_1, then component e_i is determined as a potential fault location, and the fault location set α is obtained. S5.3: For each component e_i in the fault location set α, extract the corresponding PRPS feature value, use the support vector machine algorithm to classify the fault mode, determine the fault type as breakdown and discharge or overheating, and obtain the fault type set β; S5.4: Obtain the air chamber state parameter S(e_i) of component e_i from the fault location set α. If S(e_i) deviates from the preset threshold range T_2, the air chamber state is judged to be abnormal, and the air chamber state set γ is obtained. S5.5: Based on the fault type set β and the gas cell state set γ, analyze the characteristics of the decomposition products. If the fault type is discharge and the gas cell state is abnormal, calculate the concentration of the decomposition products using spectral analysis technology to obtain the set of decomposition product concentrations δ. S5.6: By fusing the fault location set α, the fault type set β, the gas chamber state set γ, and the decomposition product concentration set δ, a structured fault diagnosis result set G is generated, where G={α, β, γ, δ}; S5.7: If there are multiple component fault associations in the fault diagnosis result set G, the fault propagation path is updated using a graph convolutional network based on the electrical connection edges of the knowledge graph to obtain the final diagnosis result set G.

7. The method for detecting and evaluating the insulation status of GIL equipment in a hydropower station based on an industrial internet platform according to claim 1, characterized in that: The sub-steps of S6 are: S6.1: Based on the diagnostic result set G, perform fault mode and impact analysis to identify each fault mode and its impact on system performance and component operation. Combine the edges F of the knowledge graph to generate a component-related impact set H, where H represents the mapping set between fault modes and affected components. S6.2: Based on the component-related impact set H, perform fault tree analysis, define insulation failure as the top event, take the fault modes in H as the basic events, construct the causal propagation path, and generate the fault propagation graph I, where I represents the set of logical relationships from component failure to system failure; S6.3: Based on the diagnostic result set G and the fault propagation map I, quantify the defect trend, safety risk and economic risk of the insulation state, calculate the comprehensive risk value J, where J represents the risk quantification value, and generate alarm level K and handling suggestions L, where K represents the risk level and L represents the set of handling measures. S6.4: If the alarm level K is higher than the preset threshold M, where M represents the risk trigger threshold, then perform convolutional neural network analysis on the UHF signal and ultrasonic signal in the structured feature set B. The convolution kernel is based on the pre-trained model to generate the fault location classification probability N, where N represents the location verification probability, and update the fault location in the diagnostic result set G. S6.5: Based on the fault location in the updated diagnostic result set G, query the associated equipment components through the knowledge graph, define the edge F as the electrical dependency, and generate the fault impact range O, where O represents the set of components {e_i,e_j,...} affected by the fault; S6.6: Based on the fault impact range O, the diagnostic result set G, and the alarm level K, generate a diagnostic report P according to predefined alarm rules. P includes the fault type, fault location, impact range, and handling suggestions. P is then transmitted to the monitoring system to complete the insulation status monitoring.

8. The method for detecting and evaluating the insulation status of GIL equipment in a hydropower station based on an industrial internet platform according to claim 7, characterized in that: The specific method in S6.1 is as follows: Fault mode data is obtained from the diagnostic result set, and feature information is extracted to form a fault mode set. For this set, combined with system performance parameters and component operating status, an impact analysis algorithm is used to determine the impact range of each fault mode on the system and components. With the help of a knowledge graph, the component association relationship is obtained based on edge F, generating a preliminary association impact set between fault modes and components. If there are multiple paths of influence in the preliminary set, a depth-first search algorithm is used to traverse edge F to determine the propagation path, resulting in an optimized association impact set. Based on the optimized set, a mapping relationship between fault modes and affected components is constructed to form a component mapping set. The identification information of affected components is extracted from the component mapping set, and the priority is determined based on the system performance change trend to obtain the final component association impact set H. The consistency between H and the diagnostic result set is verified by data comparison to determine the accuracy of H and output the final mapping set.

9. The method for detecting and evaluating the insulation status of GIL equipment in a hydropower station based on an industrial internet platform, as described in claim 7, is characterized in that: The specific method in S6.2 is as follows: Obtain the component-related impact set H, extract the failure modes from it to form the basic event set ε; use the fault tree analysis method, set insulation failure as the top event ζ, and construct the fault tree structure η based on ε; analyze the causal propagation relationship between the basic events in η through Boolean algebra operations to generate causal propagation paths θ; calculate the logical relationship strength ι of each path based on θ to obtain the fault propagation map I; If a path in I exceeds a preset threshold, this path is marked as a high-risk path κ. By using κ to extract key basic events λ, an optimized fault tree μ is generated; finally, based on μ, a set of system failure logic relationships ν is output, which represents the complete causal chain from key component failure to insulation failure.

10. The method for detecting and evaluating the insulation status of GIL equipment in a hydropower station based on an industrial internet platform according to claim 7, characterized in that: The specific method in S6.3 is as follows: Insulation status data are extracted from the diagnostic result set G, and statistical analysis is used to obtain defect characteristics and trends. By combining the defect change trend and fault propagation map I, the defect propagation path is calculated using graph analysis algorithms to identify the distribution of potential fault points. If the distribution of potential failure points exceeds the preset threshold, the safety risk value ξ is obtained through the safety risk model to complete the quantification of safety risk; the economic risk value ο is calculated by using the economic loss assessment model in combination with the defect change trend to achieve the quantification of economic risk. A weighted fusion method is used to combine the safety risk value ξ and the economic risk value ο to obtain the comprehensive risk value J, and to determine the comprehensive risk quantification index; if the comprehensive risk value J exceeds the preset threshold, an alarm level K is generated according to the risk classification rules. Finally, based on the alarm level K, the corresponding set of handling measures L is extracted from the preset measures library to generate suggested measures.

Images