Method for dynamic optimization of insulation performance of new fluorocarbon insulating gas in high and low voltage equipment
By constructing a digital twin model and combining it with multi-source heterogeneous data, real-time equipment data is collected for liquefaction risk verification and pressure correction. Combined with the coupled simulation of single-physics field model and digital twin model, the insulation performance of fluorocarbon insulating gas is dynamically optimized. This solves the problems of low accuracy in insulation performance assessment and insufficient fault prediction throughout the equipment's life cycle, and achieves safe and reliable operation of the equipment.
Patent Information
- Application Number
- CN202611225032.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-13
- Publication Date
- 2026-11-17
AI Technical Summary
In existing technologies, fluorocarbon insulating gases cannot adapt to a wide range of operating condition fluctuations throughout the entire life cycle of high and low voltage equipment. They pose problems such as liquefaction risk, insulation performance degradation, and low assessment accuracy. They also lack real-time identification and prediction of insulation deterioration, leading to the risk of fault expansion.
By constructing a digital twin model and combining it with multi-source heterogeneous data, real-time equipment data is collected for liquefaction risk verification and voltage division correction. By combining the coupled simulation of single-physics field model and digital twin model, insulation performance is dynamically optimized, insulation degradation trend is predicted, and multi-objective coordinated control is carried out.
It enables precise assessment and dynamic optimization of the overall insulation performance of equipment, avoids liquefaction risks in advance, accurately predicts changes in insulation performance, reduces the risk of failure, and ensures safe and reliable operation of equipment.
Smart Images

Figure CN123116786A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of electrical equipment insulation technology, specifically a method for dynamically optimizing the insulation performance of a novel fluorocarbon insulating gas in high and low voltage equipment. Background Technology
[0002] Sulfur hexafluoride (SF6) is widely used in high and low voltage electrical equipment, from 0.4kV low-voltage switchgear to 1100kV ultra-high voltage GIS and GIL, due to its excellent insulation and arc-quenching properties. However, its global warming potential (GWP) is as high as 23,500 times that of CO2, and its atmospheric lifetime is as long as 3,200 years. It has been listed as a strong greenhouse gas by the Kyoto Protocol, and the use of SF6 is increasingly restricted globally. Developing environmentally friendly alternative insulating gases has become a core development trend in the power industry.
[0003] New types of fluorocarbon gases, such as perfluoroisobutyronitrile, perfluoropentanone, perfluorohexanone, and trifluoromethanesulfonyl fluoride, have the characteristics of high insulation strength, low GWP value, and zero ODP, making them the most promising alternative medium to sulfur hexafluoride. They have already been demonstrated and applied in medium and low voltage ring main units, high voltage GIS / GIL and other equipment.
[0004] However, the engineering applications of fluorocarbon insulating gases still suffer from the following core drawbacks in existing technologies: 1. Most existing technologies adopt static, fixed-ratio mixed gas designs, which can only meet the insulation requirements under rated operating conditions and cannot adapt to the wide range of operating condition fluctuations throughout the entire life cycle of high and low voltage equipment. Factors such as temperature changes, load fluctuations, altitude differences, and component leakage during equipment operation will cause dynamic changes in the insulation strength of the mixed gas. Static ratios cannot simultaneously address the risk of liquefaction at extreme low temperatures and the problem of insufficient insulation margin at high temperatures / high altitudes.
[0005] 2. Fluorocarbon insulating gases decompose under discharge or overheating faults, producing byproducts such as CO, cyanide, and fluorocarbons. At the same time, the main insulating components are continuously consumed, leading to a continuous decline in insulation performance. Current technology can only detect gas components and decomposition products through periodic maintenance, and cannot achieve online identification and real-time suppression of insulation degradation, which poses a risk of fault expansion.
[0006] 3. Existing insulation optimization methods mostly involve single-dimensional parameter adjustments, such as optimizing only the gas ratio or improving only the equipment insulation structure. They do not consider the coupled influence of multiple factors such as gas composition, pressure and temperature, electric field distribution, and decomposition products on insulation performance, and therefore cannot achieve global optimal control of insulation performance, resulting in limited optimization effects.
[0007] 4. Existing technologies lack the ability to predict and actively control insulation performance. Most of them are passive remedies after insulation deterioration occurs. They cannot identify the critical boundary of insulation deterioration in advance, making it difficult to achieve predictive maintenance of equipment and full life cycle management of insulation performance.
[0008] Therefore, developing a dynamic optimization method for the insulation performance of fluorocarbon insulating gases that can adapt to all operating conditions and achieve proactive prediction, coordinated control, and closed-loop optimization has become a key issue that urgently needs to be addressed in the engineering application of insulating gases. Summary of the Invention
[0009] To address the aforementioned technical problems, the present invention aims to provide a novel method for dynamically optimizing the insulation performance of fluorocarbon insulating gases in high and low voltage equipment, comprising the following steps: Step s1: For high and low voltage electrical equipment filled with new fluorocarbon insulating mixed gas, collect multi-source heterogeneous data in high and low voltage electrical equipment in real time, and perform liquefaction risk pre-verification and voltage division correction on the multi-source heterogeneous data; Step s2: Construct digital twin models of high and low voltage electrical equipment and several single-physics field models, attach and bind the several single-physics field models to the core computing domain in the digital twin model, dynamically assign material parameters of each single-physics field model based on real-time multi-source heterogeneous data, determine the coupling coefficient solution logic between each single-physics field model, attach and bind the coupling coefficient solution logic to the corresponding core computing domain to generate a digital twin; Step s3: Quantify the insulation performance of the multi-source heterogeneous data, obtain the intrinsic insulation strength base value under the current working condition, and correct the intrinsic insulation strength base value according to the global coupling simulation results output by the digital twin to obtain the real-time insulation strength, real-time insulation margin and insulation state of the new fluorocarbon insulating gas under the current working condition. Step s4: When the insulation state is in the warning state, based on the global coupling simulation results of the digital twin, the real-time insulation strength and the real-time insulation margin, and combined with the pre-trained insulation degradation mechanism model, obtain the insulation margin time-series decay curve within the future set time window, and construct the control queue based on the insulation margin time-series decay curve. Step s5: Starting from the first position in the control queue, perform multi-target coordinated dynamic control of insulation performance in sequence, obtain the optimal control parameter scheme to be executed in sequence, and execute the corresponding control action according to the optimal control parameter scheme to be executed in sequence.
[0010] Furthermore, the multi-source heterogeneous data includes multi-dimensional state parameters of novel fluorocarbon insulating gases, equipment operating electrical parameters, and environmental parameters.
[0011] Furthermore, the process of pre-verifying liquefaction risk and correcting partial pressure for multi-source heterogeneous data includes: The main insulating gas, gas composition, pressure, and temperature parameters in the environmental parameters are obtained from the new fluorocarbon insulating gas. The saturated vapor pressure of the main insulating gas at the current temperature parameter is obtained. The pressure corresponding to the main insulating gas in the multi-dimensional state parameters is compared with the saturated vapor pressure. If the pressure of the main insulating gas is greater than the saturated vapor pressure, the pressure of the main insulating gas in the multi-dimensional state parameters is corrected to the saturated vapor pressure, and the gas composition of the main insulating gas is corrected simultaneously.
[0012] Furthermore, the process of building a digital twin includes: Extract 3D design drawings, system material parameters, equipment rating and boundary parameters, and test database of high and low voltage electrical equipment. Construct 3D geometric models of high and low voltage electrical equipment based on the 3D design drawings. Divide the 3D geometric model into multiple physical domains and split it into several core computational domains. Set API interfaces for each core computational domain. Assign attribute values to each core computational domain based on system material parameters and equipment rating and boundary parameters. Determine the material parameters and equipment rating and boundary parameters of each core computational domain. Match multi-source heterogeneous data with the core computational domains of the 3D geometric model to generate a digital twin model. A single-physics-field basic model is constructed, and the core computing domains of the single-physics-field basic model and the digital twin model are bound together through an API interface. A correlation function between multi-source heterogeneous data and material parameters of each core computing domain is constructed. The material parameters of each core computing domain are dynamically updated based on the correlation function and multi-source heterogeneous data. The material parameters of the single-physics-field basic model bound to each core computing domain are dynamically assigned based on the dynamically updated material parameters of each core computing domain. Boundary conditions are set for the single-physics-field basic model bound to each core computing domain based on the device rating and boundary parameters of each core computing domain. Based on the physical characteristics of high and low voltage electrical equipment filled with a new type of fluorocarbon insulating mixed gas, the coupling type between each single physical field basic model is obtained. According to different coupling types, the coupling coefficient solution logic between each single physical field basic model is set. The coupling coefficient solution logic between each single physical field basic model is mounted and bound to the core computing domain corresponding to each single physical field basic model through the API interface to generate a digital twin. The digital twin is subjected to secondary calibration and accuracy verification based on the experimental database to obtain a digital twin that has completed accuracy verification.
[0013] Furthermore, the process of quantifying the insulation performance of multi-source heterogeneous data to obtain the intrinsic insulation strength baseline value under the current operating conditions includes: A basic calculation model for the insulation strength of a novel fluorocarbon insulating gas is constructed. The input parameters of the basic calculation model are gas composition, pressure and temperature parameters, and the output parameter is the intrinsic insulation strength of the novel fluorocarbon insulating gas under standard operating conditions. The gas composition, pressure, and temperature parameters, after completing the pre-verification of liquefaction risk and partial pressure correction, are input into the basic calculation model of insulation strength. Based on the basic calculation model of insulation strength, the intrinsic insulation strength of the new fluorocarbon insulating gas under standard working conditions is output. Based on the real-time collected electrical parameters of the equipment operation, the current operating condition type is identified, and an impact coefficient reference table is constructed. The impact coefficient reference table includes the impact coefficients corresponding to different operating condition types. The impact coefficient is determined based on the current operating condition type and the impact coefficient reference table. The intrinsic insulation strength base value under the standard operating condition is corrected based on the impact coefficient to obtain the intrinsic insulation strength base value under the current operating condition.
[0014] Furthermore, based on the global coupled simulation results output by the digital twin, the intrinsic insulation strength baseline value is corrected, and the process of obtaining the real-time insulation strength, real-time insulation margin, and insulation state of the novel fluorocarbon insulating gas under the current operating conditions includes: Based on the global coupled simulation results output by the digital twin, the dynamic correction coefficient matrix is obtained; The intrinsic insulation strength base value under the current operating condition is corrected according to the dynamic correction coefficient matrix to obtain the real-time insulation strength under the current operating condition. At the same time, the minimum allowable insulation strength under the current operating condition is obtained. If the real-time insulation strength is less than the minimum allowable insulation strength, an insulation abnormality alarm is immediately triggered. If the real-time insulation strength is greater than or equal to the minimum allowable insulation strength, the real-time insulation margin is obtained according to the real-time insulation strength and the minimum allowable insulation strength, and the insulation status is obtained according to the insulation margin.
[0015] Furthermore, the insulation degradation mechanism model includes a bottom-level mechanism sub-model, a data-driven self-correcting model, and a total rate solution model. The input of the bottom-level mechanism sub-model is the global coupled simulation result output by the digital twin, and the output is the insulation performance degradation rate. The data-driven self-correcting model takes as input the insulation performance degradation rate sequence of different degradation paths within the historical period, the global coupled simulation result sequence, and the measured insulation margin sequence, and outputs the correction coefficients for the insulation performance degradation rate of different degradation paths.
[0016] The overall rate solution model is used to correct the insulation performance degradation rate output by the underlying mechanism sub-model corresponding to different degradation paths based on the correction coefficient of the insulation performance degradation rate of different degradation paths. The maximum insulation performance degradation rate is selected as the overall insulation degradation rate from the corrected insulation performance degradation rates.
[0017] Furthermore, the process of obtaining the insulation margin timing decay curve within a future set time window includes: The global coupled simulation results of the digital twin at the current moment are input into the insulation degradation mechanism model. The total insulation degradation rate is obtained according to the insulation degradation mechanism model. The global coupled simulation results of the digital twin at the current moment, the real-time insulation strength, the insulation margin, and the total insulation degradation rate are used as the initial state for prediction. The prediction time window is set according to the current working conditions. Obtain the operating condition boundary condition sequence of high and low voltage electrical equipment within a future time window, input the operating condition boundary condition sequence within the future time window into a digital twin, and output the global coupled simulation result sequence for each time step within the future time window based on the digital twin; The sequence of global coupled simulation results for each time step is input into the insulation degradation mechanism model. Based on the insulation degradation mechanism model, the total insulation degradation rate and the degradation path corresponding to the total insulation degradation rate for each time step are output. Based on the total insulation degradation rate for each time step and the real-time insulation strength in the initial state, the cumulative insulation strength attenuation is obtained. Based on the cumulative insulation strength attenuation and the real-time insulation strength in the initial state, the real-time insulation strength for each future time step is obtained. Based on the real-time insulation strength for each future time step, the insulation margin for each future time step is obtained. Based on the insulation margin at each future time step and the insulation margin in the initial state, construct the insulation margin time-series decay curve within the future time window.
[0018] Furthermore, the process of constructing the control queue based on the insulation margin time-series decay curve includes: Within the insulation margin range, threshold points are selected to divide sub-intervals into different warning levels. The insulation margin time-series decay curve within the future time window is obtained and compared with the sub-intervals of different warning levels one by one. The warning level at different times within the future time window is obtained, and the degradation path corresponding to the total insulation degradation rate at different times within the future time window is obtained. A control priority comparison table is preset, which includes the control priority corresponding to different warning levels and degradation paths. Based on the control priority comparison table and the warning level and degradation path at different times within the future time window, the control priorities at different times within the future time window are generated, and the control priorities at different times are sorted to generate a control queue.
[0019] Furthermore, the process of performing multi-objective coordinated dynamic control of insulation performance sequentially from the first position in the control queue to obtain the optimal control parameter scheme corresponding to each control priority includes: Several objective optimization functions are constructed, and the preset hard constraints corresponding to the current regulatory priority are obtained. Adaptive weights are set for different objective optimization functions according to the regulatory priority. A fitness function is constructed based on several objective optimization functions and adaptive weights. Several regulatory parameter schemes are preset. Chromosome encoding, hard constraint verification, and population initialization are performed on several regulatory parameter schemes to generate an initial population that meets the hard constraints. Based on the initial population and fitness function, the optimal regulatory parameter scheme corresponding to the current regulatory priority is obtained through a multi-objective genetic algorithm.
[0020] Compared with the prior art, the beneficial effects of the present invention are: 1. Existing technologies mostly rely on discrete data collected by single-point sensors, which can only achieve a rough assessment of insulation performance. They cannot capture local extreme operating conditions inside the equipment, and are prone to missed detections of "global parameters being qualified, but local insulation failure." Furthermore, they are difficult to adapt to the specific characteristics of fluorocarbon insulating gases, which are sensitive to temperature changes and easily liquefied. This invention avoids the interference of liquefaction factors on insulation assessment in advance by pre-checking the liquefaction risk of multi-source heterogeneous data and correcting for voltage division, thus ensuring the accuracy of the basic assessment data.
[0021] 2. Relying on the global coupled simulation results output by the digital twin, this method overcomes the limitations of single-point monitoring, accurately capturing the physical field distribution throughout the entire internal space of the equipment. This enables dynamic correction of the intrinsic insulation strength baseline value, ultimately yielding real-time insulation strength, real-time insulation margin, and insulation status that highly match the actual operating conditions of the equipment. This evaluation method considers both global performance and local details, completely solving the problems of low evaluation accuracy and weak anti-interference capability in existing technologies. It provides accurate and reliable decision-making basis for subsequent insulation control, fundamentally improving the scientific nature and accuracy of insulation evaluation.
[0022] 3. This invention binds the core computational domain of a single-physics field model to a digital twin model, constructs a correlation function between multi-source heterogeneous data and material parameters, and realizes dynamic assignment and updating of material parameters. At the same time, it sets the solution logic for the coupling coefficient between each single-physics field model according to the physical characteristics of the equipment and binds it to the core computational domain. The generated digital twin can receive multi-source heterogeneous data in real time and output full-domain coupled simulation results synchronously, realizing real-time linkage between the digital twin and the actual operating state of the equipment.
[0023] 4. The insulation degradation mechanism model constructed in this invention integrates a low-level mechanism sub-model, a data-driven self-correcting model, and a total rate solution model. It not only captures the microscopic mechanisms of degradation based on the global simulation results of a digital twin, but also improves prediction accuracy through self-correction using historical data. This allows for precise acquisition of the total insulation degradation rate and the dominant degradation path. Combined with the future operating condition boundary condition sequence and the rapid simulation capabilities of the digital twin, it can obtain the insulation margin time-series decay curve within a set future time window, clearly predicting the changing trend of insulation performance. Based on this, by classifying early warning levels, setting control priorities, and constructing control queues, it achieves advanced identification and hierarchical control of insulation degradation risks, upgrading insulation management from "post-event remediation" to "pre-event control," effectively avoiding the sudden risk of insulation failures and extending the safe operating cycle of equipment.
[0024] 5. This invention, based on the priority sorting of the control queue, executes multi-objective collaborative control operations sequentially starting from the first position. By constructing a multi-dimensional objective optimization function, dynamically setting adaptive weights for each objective in conjunction with the control priority, and then solving for the optimal control parameter scheme through a multi-objective genetic algorithm, it achieves multi-objective global optimization of insulation performance, environmental protection, economy, and equipment lifespan. This control method strictly adheres to the core hard constraint of achieving insulation margin standards while adapting to differentiated optimization objectives according to different control priorities, avoiding the limitations of single control. Attached Figure Description
[0025] Figure 1 This is a flowchart illustrating the dynamic optimization method for the insulation performance of a novel fluorocarbon insulating gas in high and low voltage equipment, as described in this application. Detailed Implementation
[0026] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0027] like Figure 1 As shown, a method for dynamically optimizing the insulation performance of a novel fluorocarbon insulating gas in high and low voltage equipment includes the following steps: Step s1: For high and low voltage electrical equipment filled with new fluorocarbon insulating mixed gas, collect multi-source heterogeneous data in high and low voltage electrical equipment in real time, and perform liquefaction risk pre-verification and voltage division correction on the multi-source heterogeneous data; Step s2: Construct digital twin models of high and low voltage electrical equipment and several single-physics field models. Bind the single-physics field models to the core computing domains in the digital twin models. Construct correlation functions between multi-source heterogeneous data and material parameters for each core computing domain. Dynamically assign material parameters to each single-physics field model based on real-time multi-source heterogeneous data. Determine the coupling coefficient solution logic between each single-physics field model. Bind the coupling coefficient solution logic to the corresponding core computing domain to generate a digital twin. Step s3: Quantify the insulation performance of the multi-source heterogeneous data, obtain the intrinsic insulation strength base value under the current working condition, and correct the intrinsic insulation strength base value according to the global coupling simulation results output by the digital twin to obtain the real-time insulation strength, real-time insulation margin and insulation state of the new fluorocarbon insulating gas under the current working condition. Step s4: When the insulation state is in the warning state, based on the global coupling simulation results of the digital twin, the real-time insulation strength and the real-time insulation margin, and combined with the pre-trained insulation degradation mechanism model, obtain the insulation margin time-series decay curve within the future set time window, and construct the control queue based on the insulation margin time-series decay curve. Step s5: Starting from the first position in the control queue, perform multi-target coordinated dynamic control of insulation performance in sequence, obtain the optimal control parameter scheme to be executed in sequence, and execute the corresponding control action according to the optimal control parameter scheme to be executed in sequence.
[0028] It should be further explained that, in the specific implementation process, multi-source heterogeneous data includes multi-dimensional state parameters of the new fluorocarbon insulating gas, equipment operating electrical parameters, and environmental parameters.
[0029] The multi-dimensional status parameters, equipment operating electrical parameters, and environmental parameters are detailed in Table 1 below: Table 1 Parameter categories Specific data collection parameters core role Multidimensional state parameters of novel fluorocarbon insulating gases (core) <![CDATA[1. Volume concentration and partial pressure of main insulating fluorocarbon gas 2. Volume concentration and partial pressure of buffer gas 3. Absolute total pressure of the gas chamber 4. Multi-point average temperature of gas in the gas chamber 5. Trace water content (volume fraction) 6. Types and concentrations of characteristic decomposition products (CF3CN, C2F6, CO, COF2, CNCN, etc.) 7. Gas leakage rate 8. Partial discharge capacity, discharge repetition rate, discharge phase]]> It provides core input for insulation strength calculation and a basis for degradation correction. Equipment operating electrical parameters 1. Equipment terminal voltage and load current; 2. Harmonic content and DC bias component; 3. Lightning / switching overvoltage amplitude, steepness, and frequency; 4. Number of switch operations; 5. Grounding current and surface leakage current of basin insulators. Identifying the current operating condition type provides a basis for the weighting of insulation strength under different operating conditions, and also provides input for electric field distortion correction. Environmental parameters 1. Real-time ambient temperature, minimum / maximum ambient temperature; 2. Ambient humidity; 3. Altitude and atmospheric pressure of the installation location. Provides input for liquefaction risk correction, altitude correction, and environmental adaptability correction. It should be further explained that, in the specific implementation process, the process of pre-verifying the liquefaction risk and correcting the partial pressure of multi-source heterogeneous data includes: To obtain the primary insulating gas in a novel fluorocarbon insulating gas (artificially designed, comprising 5%–30% (volume fraction) in a mixed gas, and responsible for 70%–90% of the insulation strength, for example...) The gas composition (volume percentage of each individual gas in the mixed insulating gas), pressure, and temperature parameters from the multi-dimensional state parameters are used to obtain the saturated vapor pressure of the main insulating gas at the current temperature (according to Antoine's equation). The values are obtained, where A, B, and C are the Antoine constants obtained from preliminary experiments for different fluorocarbon gases under the current temperature parameters. For temperature, (This refers to the saturated vapor pressure). The pressure corresponding to the main insulating gas in the multi-dimensional state parameters is compared with the saturated vapor pressure. If the pressure of the main insulating gas is greater than the saturated vapor pressure... Then, the pressure of the main insulating gas in the multi-dimensional state parameters is corrected to the saturated vapor pressure. Synchronously correct the gas composition of the main insulating gas ;in, The gas composition after correction for the main insulating gas. Pressure of insulating gas before correction.
[0030] It needs further explanation that, in the specific implementation process, the process includes: constructing digital twin models of high and low voltage electrical equipment and several single-physics models; binding the single-physics models to the core computational domains in the digital twin models; constructing correlation functions between multi-source heterogeneous data and material parameters for each core computational domain; dynamically assigning values to each single-physics model based on real-time multi-source heterogeneous data; determining the coupling coefficient solution logic between each single-physics model; and binding the coupling coefficient solution logic to the corresponding core computational domains to generate a digital twin. Extract 3D design drawings, comprehensive material parameters, equipment ratings and boundary parameters, and test databases (including equipment factory type test data, fluorocarbon mixed gas basic test data, and equipment field operation history data) of high and low voltage electrical equipment. Based on the 3D design drawings, construct 3D geometric models of the high and low voltage electrical equipment in the ANSYS simulation platform. All dimensional parameters are set as variables that can be called in real time, supporting subsequent dynamic modifications (such as real-time adjustment of gas chamber pressure and electrode spacing). Divide the 3D geometric model into multiple physical domains, splitting it into several core computational domains (including conductor domain / solid insulation domain / gas domain / metal shell domain, etc.). Set API interfaces for each core computational domain. Assign attribute values to each core computational domain based on the comprehensive material parameters and equipment ratings and boundary parameters, determining the material parameters and equipment ratings and boundary parameters of each core computational domain, laying the foundation for subsequent multiphysics field assignment. Match multi-source heterogeneous data with the core computational domains of the 3D geometric model to generate digital twin models. Construct basic single-physics field models of electrical equipment, including electric field, temperature field, and fluid flow field. Bind these basic single-physics field models to the core computational domains of the digital twin model via API interfaces. Taking the electric field model as an example: The computational domain is bound to the high-voltage conductor / contact / equalizing shield metal body domain, epoxy resin, rubber and other solid insulator domain, fluorocarbon mixed gas filled gas chamber fluid body domain, and grounding metal shell domain; Material parameters are associated with geometrical domains and specific properties of the directional electric field: dielectric constant, conductivity, volume resistivity, and surface resistivity. Strongly bound geometric boundaries: High voltage potential boundaries: bound to the outer surface of the high voltage conductor in the geometric model; Zero-potential grounding boundary: Binds to all outer surfaces of the equipment housing and the grounded metal structural surface; Floating potential boundary: internal shielding of the binding geometry, and floating electrode geometry; Insulator surface boundary: The specific curved surface used to bind pot-type insulators and support insulators.
[0031] It should be further explained that the modeling framework, governing equations, solution algorithms, and independent single-field solution logic of the single-physics field basic simulation model are all existing technologies in the fields of high-voltage electrical appliances and multi-physics field simulation. This invention only uses the existing mature single-physics field modeling architecture as the underlying simulation carrier, and addresses the inherent defects of the existing technology: the existing conventional single-physics field modeling schemes all use fixed static material parameter assignments. The gas dielectric constant, thermal conductivity, viscosity, specific heat capacity, solid insulation dielectric parameters, thermal conductivity, etc. are all pre-entered as fixed constants and do not change with the real-time operating conditions of the equipment, environmental conditions, or gas state. Abandoning the fixed physical property parameters of existing technologies, and while maintaining the existing single-physics-field basic model framework, the physical property parameters of all media are dynamically assigned through formulaization, correlation, and real-time linkage. A working condition linkage parameter mapping relationship is implanted into each single-physics-field model. Specifically, a correlation function is constructed between multi-source heterogeneous data and material parameters in each core computational domain. Based on the correlation function and multi-source heterogeneous data, the material parameters of each core computational domain are dynamically updated. Based on the dynamically updated material parameters of each core computational domain, the material parameters of the single-physics-field basic models attached to each core computational domain are dynamically assigned. Taking the dynamic assignment of material parameters in the electric field model as an example: For metallic conductors: Considering the effect of temperature on conductivity, a conductivity-temperature correlation function is constructed, and the conductivity is dynamically updated based on the conductivity-temperature correlation function and the measured temperature. For fluorocarbon mixtures: Construct a real-time linkage function between dielectric constant and gas composition, pressure, and temperature. This enables real-time dynamic updating of dielectric parameters, among which... The relative permittivity of the gas mixture is... Let Avogadro's constant be 1. For gas density, The molar mass of the gas mixture For the first mole fraction of the components For the first Molecular polarizability of the components; Boundary conditions are set for the single-physics basic models attached to and bound to each core computing domain based on the device ratings and boundary parameters of each core computing domain. Based on the physical characteristics of high and low voltage electrical equipment filled with a novel fluorocarbon insulating gas mixture, the coupling types between the basic models of each single physics field are obtained. The specific coupling types are shown in Table 2 below: Table 2 Coupled pair Coupling type Coupling mechanism Core linkage parameters Temperature field - flow field (thermal-fluid coupling) Bidirectional strong coupling 1. The temperature distribution of the temperature field alters the density and viscosity of the gas, affecting the flow characteristics of the flow field; 2. Convective heat transfer in the flow field, in turn, alters the temperature distribution, directly determining the temperature uniformity within the gas chamber and the risk of liquefaction. Temperature, gas density, viscosity, convective heat transfer coefficient Temperature field - electric field Bidirectional strong coupling 1. Temperature changes alter the dielectric parameters and conductivity of solid insulation and gases, affecting the electric field distribution and insulation strength; 2. Joule heating and dielectric losses in the electric field become heat sources for the temperature field, directly determining the equipment temperature rise. Temperature, dielectric constant, dielectric loss, conductivity, heat source power Flow field - electric field One-way coupling The pressure distribution in the flow field alters the gas density, which in turn changes the gas's insulation strength and dielectric parameters, affecting its electric field breakdown characteristics. Pressure, gas density, dielectric constant, insulation strength Chemical field (gas decomposition) - electric field - temperature field Multidirectional strong coupling 1. Partial discharge in the electric field and overheating in the temperature field drive the decomposition of fluorocarbon gas, generating decomposition byproducts; 2. The decomposition products alter the gas's dielectric parameters and insulation strength, which in turn affect the electric field distribution and degradation rate; 3. The thermal effect of the decomposition reaction further influences the temperature field distribution. Concentration of decomposition products, insulation strength, dielectric constant, partial discharge quantity, and temperature Structural field-electric field One-way coupling Contact erosion and electrode deformation alter the insulation structure, affecting the electric field distribution and the electric field non-uniformity coefficient. Electrode size, insulation distance, electric field non-uniformity coefficient The logic for solving the coupling coefficients between the basic models of each single physics field is set according to different coupling types, including: For bidirectional and multidirectional strongly coupled pairs: a direct coupling solver is used to solve the control equations of temperature field, flow field and electric field simultaneously to ensure real-time bidirectional linkage between physical fields. The convergence residual is set to ≤1e-6 to ensure solution accuracy. For unidirectional coupling pairs: a sequential solution strategy is adopted, first solving the upstream physics field, and using the output results as the input parameters of the downstream physics field to reduce the amount of computation and improve the solution speed; The coupling coefficient solution logic between each individual physics field basic model is attached and bound to the core computation domain corresponding to each individual physics field basic model through the API interface to generate a digital twin. The digital twin is subjected to two-level calibration (static + dynamic calibration) and accuracy verification based on the experimental database, and the digital twin that has completed accuracy verification is obtained.
[0032] It should be further explained that, in the specific implementation process, the construction of a digital twin for the new fluorocarbon insulating gas high and low voltage equipment is based on the equipment's factory type test data, basic test data of the fluorocarbon mixed gas, and full-process monitoring data of the equipment's recent stable operation (fluorocarbon gas parameters of each sensor monitoring node, equipment electrical operation parameters, multi-physics field state parameters, and insulation performance test data). Static and dynamic dual calibrations are performed on the digital twin to ensure that the model's parameter response, multi-field coupling, and insulation characteristics deviate from the physical equipment by ≤±3%, meeting the accuracy requirements for dynamic optimization of insulation performance and process testing. Static calibration: Match the three-dimensional model parameters of each core insulation unit (such as gas chamber volume - gas pressure, electrode spacing - electric field distribution, insulator structure - dielectric properties, control unit parameters - gas distribution accuracy) to the physical equipment; Dynamic calibration: Simulate the normal operating state of physical equipment to verify that the entire process parameter linkage response and multi-physics coupling law in the model of "fluorocarbon gas charging → electric field establishment → temperature field evolution → flow field circulation → insulation performance feedback" are consistent with the physical equipment. For example, after adjusting the gas chamber charging pressure, the gas dielectric constant and electric field distortion coefficient in the model need to match the physical equipment; after adjusting the ambient temperature, the dynamic decay law of gas liquefaction state and insulation strength in the model need to match the physical equipment.
[0033] It should be further explained that, in the specific implementation process, the process of quantifying the insulation performance of multi-source heterogeneous data to obtain the intrinsic insulation strength baseline value under the current operating conditions includes: Based on gas molecular dynamics and discharge theory, a fundamental calculation model for the insulation strength of a novel fluorocarbon insulating gas is constructed. The input parameters of the fundamental calculation model are gas composition, pressure, and temperature parameters, and the output parameter is the intrinsic insulation strength of the novel fluorocarbon insulating gas under standard operating conditions. ; The gas composition, pressure, and temperature parameters, after completing the pre-verification of liquefaction risk and partial pressure correction, are input into the basic calculation model of insulation strength. Based on the basic calculation model of insulation strength, the intrinsic insulation strength of the new fluorocarbon insulating gas under standard working conditions is output. The calculation formula for the basic calculation model of insulation strength is as follows: ; in, : The critical breakdown field strength under standard operating conditions is a fixed value of 89 kV / cm (uniform electric field). : Coefficient of synergistic effect of mixed gases =1.20, calibrated through preliminary tests; : The i-th gas component; The i-th gas component is relatively The insulation strength (under standard operating conditions); : Real-time absolute total pressure of the air chamber, unit MPa; Standard atmospheric pressure, fixed at 0.1 MPa; Real-time thermodynamic temperature of the gas chamber, in K; Standard thermodynamic temperature, fixed at 298.15 K (25 °C).
[0034] Based on real-time collected equipment operating electrical parameters, the current operating condition type is identified. Specifically, using the equipment operating electrical parameters as input, a three-layer discrimination logic of preset parameter threshold judgment, waveform feature recognition, and fault feature quantity matching is used to distinguish between power frequency steady-state operating conditions, lightning impulse operating conditions, switching impulse operating conditions, DC operating conditions, and broadband transient operating conditions. An impulse coefficient lookup table is constructed, which includes the impulse coefficients corresponding to different operating condition types. The impulse coefficient is determined based on the current operating condition type and the impulse coefficient lookup table. The intrinsic insulation strength baseline value under standard operating conditions is corrected based on the impulse coefficient to obtain the intrinsic insulation strength baseline value under the current operating condition.
[0035] For example, based on real-time collected electrical parameters, the current operating condition type can be identified, and... During the basic adaptation process: For steady-state power frequency operation: intrinsic insulation strength basic value Remain unchanged; For lightning strike conditions: intrinsic insulation strength basic value Multiply by the impact factor of 1.30; For operational shock conditions: intrinsic insulation strength baseline value Multiply by the impact factor of 1.05; For DC operating conditions: intrinsic insulation strength basic value Multiply by the impact factor of 0.90; The process of constructing the impact coefficient comparison table includes (taking lightning impact coefficient calibration as an example): Step 1: Perform vacuum drying on the test chamber, with a vacuum degree ≤1Pa, and maintain for 24 hours to remove trace amounts of water and impurities from the surface of the chamber and electrodes; install the target electrode structure, prepare the target ratio of fluorocarbon mixed gas, let it stand for 24 hours to ensure uniform gas mixing, and control the chamber pressure and temperature to the target operating conditions.
[0036] Step 2: Power Frequency Breakdown Voltage Measurement: The 50% breakdown voltage under steady-state power frequency is measured using the step-up / step-down method. Each test group should have ≥30 effective breakdown cycles; after each breakdown, change the electrode position or polish the electrode surface, and let it stand for 10 minutes; calculate the 50% breakdown field strength under steady-state power frequency. ,in This represents the electrode gap distance.
[0037] Step 3: Impulse breakdown voltage determination: Re-ventilate to ensure the gas state in the chamber is completely consistent with the power frequency test; use the rise and fall method to determine the 50% breakdown voltage under the corresponding impulse waveform. Prioritize testing negative polarity impacts. Each test group should have ≥30 effective breakdown cycles, and if the dispersion is large, the number of cycles should be increased to ≥40. The interval between each impact should be ≥1 minute. Calculate the 50% breakdown field strength under the impact. .
[0038] Step 4: Impact Coefficient The core formula for calculating the impact factor is: .
[0039] It should be further explained that, in the specific implementation process, the process of correcting the intrinsic insulation strength baseline value based on the global coupled simulation results output by the digital twin to obtain the real-time insulation strength, real-time insulation margin, and insulation state of the novel fluorocarbon insulating gas under the current operating conditions includes: Based on the global coupling simulation results output by the digital twin (the simulated output field quantities are the global distribution values of the real device, replacing the approximate values of traditional single-point sensors, which is the key to improving the accuracy of this invention), the global coupling simulation results include instantaneous field strength E(r), temperature T(r), gas pressure P(r), gas-liquid phase transition state, wall / gas domain temperature and humidity boundary, flow field pressure distribution, etc., at any location in space. A dynamic correction coefficient matrix is then obtained, which is derived from the electric field distortion correction coefficients. (Electric field-flow field coupling), correction factor for trace water content (Chemical field-temperature field coupling), decomposition product degradation correction coefficient (Chemical field-electric field coupling), temperature-liquefaction safety correction factor (Temperature field-flow field coupling), multi-field coupling cross-correction coefficient Composed of, etc. The specific process of obtaining the dynamic correction coefficient matrix based on the global coupled simulation results is exemplified by the electric field distortion correction coefficient. For example, ;in, The real-time electric field non-uniformity coefficient is obtained from the electric field distribution extracted from the global coupled simulation results. This refers to the real-time partial discharge quantity in multi-source heterogeneous data. This refers to the limit value for partial discharge of the equipment. The discharge influence weighting coefficient is set to 0.2~0.5, with a larger value for more non-uniform electric fields.
[0040] It should be further clarified that the calculation formulas for the electric field distortion correction coefficient, micro-water correction coefficient, and temperature-liquefaction correction coefficient are all existing well-known technologies, and those skilled in the art can conventionally calculate them based on the electric field, temperature, and pressure distribution parameters output by the coupled simulation. This invention does not focus on protecting the specific calculation formulas for the correction coefficients, but rather on using the real-time dynamic updating of each correction coefficient driven by the full-domain simulation field output by a digital twin, replacing the traditional static correction mode of single-point monitoring data, and achieving a high-precision dynamic quantitative assessment of the insulation strength of fluorocarbon insulating gas electrical equipment.
[0041] The intrinsic insulation strength baseline value under the current operating condition is corrected based on the dynamic correction coefficient matrix to obtain the real-time insulation strength under the current operating condition. , Simultaneously obtain the minimum allowable insulation strength under the current operating conditions. If the real-time insulation strength is less than the minimum allowable insulation strength, an insulation anomaly alarm will be triggered immediately, and the emergency control procedure will be initiated. If the real-time insulation strength is greater than or equal to the minimum allowable insulation strength, the real-time insulation margin will be obtained based on the real-time insulation strength and the minimum allowable insulation strength. , The insulation status is obtained based on the insulation margin, specifically: when K≥1.2, the insulation status is excellent; when 1.0≤K<1.2, the insulation status is qualified; when K<1.0, the insulation status is warning.
[0042] It should be further explained that, in the specific implementation process, the process of constructing the insulation degradation mechanism model includes: The insulation degradation mechanism model includes a bottom-level mechanism sub-model, a data-driven self-correcting model, and a total rate solution model. The input of the bottom-level mechanism sub-model (including the electrical degradation rate sub-model, thermal degradation rate sub-model, chemical degradation rate sub-model, and surface insulation degradation rate sub-model) is the global coupled simulation result output by the digital twin, and the output is the insulation performance degradation rate. All underlying mechanism sub-models output the insulation performance degradation rate per unit time, quantifying the contribution of different degradation paths to the attenuation of insulation strength, and comparing it with the previous correction coefficients. One-to-one correspondence, with the electrical degradation rate sub-model (corresponding to the electric field distortion correction coefficient). For example: Input source: Global instantaneous field strength output by the digital twin Real-time electric field non-uniformity coefficient Partial discharge quantity Related to discharge repetition rate and the proportion of maximum field strength in critical insulation regions; Quantification formula (mechanism-driven): ;in: The electrical degradation rate at time t is expressed in % / h (relative attenuation rate of insulation strength). The reference rate constant for electrical degradation is calibrated by an accelerated decomposition test of fluorocarbon gas discharge (existing technology). The maximum electric field strength in the critical insulation region obtained from digital twin simulation; This represents the gas reference breakdown field strength under standard operating conditions. The field strength nonlinearity index is 8~12 for fluorocarbon gas (according to existing experimental calibration). This refers to the partial discharge degradation weight and partial discharge limit (based on existing experimental calibration).
[0043] The quantification formula (mechanism-driven) of the underlying mechanism sub-model is a known existing technology; the core innovation of this invention lies in deeply binding the full-domain multi-physics field simulation results of the digital twin with the specific degradation mechanism of fluorocarbon gas, constructing a degradation rate quantification system, and realizing advanced and accurate prediction and graded proactive early warning of insulation performance. The data-driven self-correcting model takes as input the insulation performance degradation rate sequence of different degradation paths within a historical period (72 hours), the global coupled simulation result sequence, and the measured insulation margin sequence, and outputs the correction coefficients of the insulation performance degradation rate of different degradation paths.
[0044] The data-driven self-correcting model is based on the Attention-LSTM temporal deep learning framework. It corrects the calculation results of the mechanistic model in real time, and solves the bias caused by complex working conditions, material aging and individual differences of equipment that cannot be covered by the pure mechanistic model. The pre-training dataset of the data-driven self-correcting model includes fluorocarbon gas accelerated aging test data, equipment life cycle operation data and extreme working condition simulation data.
[0045] The overall rate solution model is used to correct the insulation performance degradation rate output by the underlying mechanism sub-model for different degradation paths based on the correction coefficients for the insulation performance degradation rate of different degradation paths (in terms of electrical degradation rate). For example, ,in, This is the corrected electrical degradation rate. (This is a correction factor for the electrical degradation rate). The maximum insulation degradation rate is selected from the corrected insulation degradation rates as the total insulation degradation rate.
[0046] It should be further explained that, in the specific implementation process, the process of obtaining the insulation margin time-series decay curve within a future set time window includes: The global coupled simulation results of the digital twin at the current moment are input into the insulation degradation mechanism model. The total insulation degradation rate is obtained based on the insulation degradation mechanism model. The global coupled simulation results of the digital twin at the current moment and the real-time insulation strength are then input into the model. Insulation margin The total insulation degradation rate is used as the initial state for prediction, and the prediction time window is set according to the current operating conditions. For example: Transient impact condition: time window 1h, step size 1s; Normal operating conditions: Time window 24h, step size 1min; Long-term maintenance prediction: time window 72h, step size 10min.
[0047] Based on historical operating data of high and low voltage electrical equipment, power grid dispatch plans, and weather forecast data, a time-series prediction model is used to obtain the operating condition boundary condition sequence of high and low voltage electrical equipment within future time windows (the predicted operating condition boundary sequence typically has ≤10 variables; low-dimensional operating condition boundaries are easier to predict and have higher accuracy, avoiding the dimensionality curse and poor generalization of directly predicting multi-source heterogeneous data, perfectly meeting the high reliability requirements of high-voltage equipment insulation prediction). The operating condition boundary condition sequence within the future time window is input into a digital twin, and the digital twin outputs the global coupled simulation result sequence for each time step within the future time window: , The total number of time steps contained in the time window; The sequence of global coupled simulation results at each time step is input into the insulation degradation mechanism model, and the total insulation degradation rate at each time step is output based on the insulation degradation mechanism model. The degradation path corresponding to the total insulation degradation rate is used to obtain the cumulative insulation strength attenuation based on the total insulation degradation rate at each time step and the real-time insulation strength in the initial state. ,in Using the time step as the basis, the real-time insulation strength for each future time step is obtained based on the cumulative insulation strength attenuation and the real-time insulation strength in the initial state. The insulation margin for each future time step is obtained based on the real-time insulation strength at each future time step. ,in Minimum allowable insulation strength; Based on the insulation margin at each future time step and the insulation margin in the initial state, construct the insulation margin time-series decay curve within the future time window.
[0048] It should be further explained that, in the specific implementation process, the process of constructing the control queue based on the insulation margin time-series decay curve includes: Within the insulation margin range, threshold points are selected to divide the area into sub-intervals with different warning levels. Corresponding to blue alert (normal). Corresponding to the yellow alert (Caution). Corresponding to orange alert (abnormal). For red alerts (emergency), the insulation margin time-series decay curve within the future time window is compared one by one with the sub-intervals of different alert levels. The alert level at different times within the future time window is obtained, as well as the degradation path corresponding to the total insulation degradation rate at different times within the future time window. A preset control priority lookup table is established, which includes the control priorities corresponding to different alert levels and degradation paths. For example: Level 1 Priority (Emergency Safety): Liquefaction risk, insulation margin about to fall below the minimum threshold (predicted within 1 hour). In cases of severe partial discharge, immediate intervention is necessary to prioritize equipment safety. Secondary priority (deterioration inhibition): Excessive levels of trace water / decomposition products, continuous loss of main components, and continuous aggravation of electric field distortion require regulation within 1 hour to inhibit the chain of deterioration. Level 3 Priority (Performance Optimization): Insulation margin is qualified but there is a trend of deterioration; there is room for optimization in terms of environmental protection / economic efficiency; control can be performed smoothly within 24 hours, balancing performance and cost. Level 4 Priority (Life Optimization): Equipment has been aging due to long-term operation and has sufficient insulation margin. Preventive adjustments can be implemented in conjunction with the maintenance plan to extend the equipment's life.
[0049] Based on the control priority comparison table and the warning level and degradation path at different times within the future time window, the control priorities at different times within the future time window are generated. The control priorities at different times are sorted to generate a control queue (first priority > second priority > third priority > fourth priority). Control operations are performed sequentially starting from the first position of the control queue to avoid system oscillations and misoperations caused by simultaneous adjustment of multiple parameters.
[0050] It should be further explained that, in the specific implementation process, the process of performing multi-objective coordinated dynamic control of insulation performance starting from the first position in the control queue, and obtaining the optimal control parameter scheme corresponding to each control priority, includes: Construct several objective optimization functions, including: Objective 1: Maximize insulation performance, quantified as maximizing the redundancy of the insulation margin after adjustment, while minimizing the total rate of insulation degradation. ,in The target insulation margin after regulation (the minimum insulation margin threshold that must be reached after the regulation action is executed under the current regulation priority). This is the minimum insulation margin. The total insulation degradation rate was calculated for the degradation mechanism model.
[0051] Objective 2: Minimize the total lifecycle cost of equipment. ,in For the cost of purchasing and replenishing fluorocarbon gases, In order to control the energy consumption cost of unit operation, For equipment maintenance costs. Costs incurred due to power outages caused by insulation failures.
[0052] Objective 3: Optimize the remaining lifespan of the equipment, quantified as minimizing the aging rate of the solid insulation and gaseous medium after regulation: ,in The rates are thermal aging, electrical aging, and chemical aging, respectively, obtained in conjunction with an insulation degradation mechanism model.
[0053] Obtain the preset hard constraints corresponding to the current control priority (e.g., the minimum insulation margin threshold that must be reached after the control action is executed under the current control priority; the control process must not affect the normal operation of the equipment, and must not trigger the power grid protection action). Set adaptive weights for different objective optimization functions according to the control priority (the adaptive weights of different objective optimization functions are determined based on expert experience; for example, for level 1 / 2 priority conditions: the insulation performance objective weight is set to 0.6, and the economic / lifetime objective weight is compressed to 0.2, prioritizing safety and degradation suppression; for level 3 / 4 priority conditions: the economic objective weight is set to 0.4, and the insulation / lifetime objective weight is set to 0.30, balancing performance and low-carbon economy). Construct a fitness function F based on several objective optimization functions and adaptive weights. The fitness function is as follows: ,in, , , Indicates adaptive weights; Several preset control parameter schemes are provided. These schemes consist of five adjustable parameters across five dimensions, each with clearly defined upper and lower limits to match the equipment's hardware capabilities, as shown in Table 3 below. Table 3 Regulation Dimensions Optimize variables Value range constraints Corresponding degradation path to be addressed Gas component ratio control Volume ratio of fluorocarbon gas in main insulation 2%~30%, while meeting liquefaction safety constraints. Component loss, liquefaction risk, insufficient insulation margin Pressure-temperature control Absolute total pressure of the air chamber, target temperature of the air chamber Pressure: 0.8~1.1 times the rated pressure; Temperature: -30℃~70℃ Liquefaction risk, temperature-induced insulation strength fluctuations Electric field distortion control Gradient electrode driving voltage, particle capture unit trigger parameters 0~Rated drive voltage, matching equipment hardware capabilities Insulation degradation caused by electric field concentration and partial discharge Purification, Regeneration, and Regulation Purification unit circulation flow rate, catalytic regeneration module operating power, adsorbent activation cycle 0~Rated maximum flow / power, matching equipment hardware capabilities Media deterioration caused by excessive moisture and accumulation of decomposition products. Coordinated Control of Operating Conditions Equipment load limits, heat dissipation unit operating power, and switch action interlocking logic 0~rated load, matching power grid dispatch requirements Thermal degradation caused by overheating, electrical degradation caused by transient shocks Several regulatory parameter schemes are encoded by chromosomes, hard constraint conditions are verified, and populations are initialized to generate an initial population that meets the hard constraints. Based on the initial population and fitness function, the optimal regulatory parameter scheme corresponding to the current regulatory priority is obtained through a multi-objective genetic algorithm.
[0054] The specific process of obtaining the optimal regulatory parameter scheme corresponding to the current regulatory priority using a multi-objective genetic algorithm includes: The regulation parameter scheme is encoded as a chromosome. For example, a string of numbers represents information such as the volume ratio of fluorocarbon gas in the main insulation, the target temperature of the gas chamber, and the circulation flow rate of the purification unit. A certain number of chromosomes are randomly generated to form an initial population. Each chromosome represents a possible regulation parameter scheme. The fitness function is obtained. Using a tournament selection method, chromosomes with higher fitness are selected from the current population as parents. The parent chromosomes are crossovered, exchanging some genes to generate new offspring chromosomes. This simulates the exchange of genes in biological genetics to generate new regulation parameter schemes. The offspring chromosomes are mutated, randomly changing some genes to increase population diversity and avoid getting trapped in local optima. The above steps are repeated iteratively until the termination condition is met, such as reaching the maximum number of iterations or the fitness value no longer significantly improving. The optimal regulation parameter scheme is then output.
[0055] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.
Claims
1. A method for dynamic optimization of the insulation performance of a new fluorocarbon insulating gas in high and low voltage equipment, characterized in that, Includes the following steps: Step s1: For high and low voltage electrical equipment filled with new fluorocarbon insulating mixed gas, collect multi-source heterogeneous data in high and low voltage electrical equipment in real time, and perform liquefaction risk pre-verification and voltage division correction on the multi-source heterogeneous data; Step s2: Construct digital twin models of high and low voltage electrical equipment and several single-physics field models, attach and bind the several single-physics field models to the core computing domain in the digital twin model, dynamically assign material parameters of each single-physics field model based on real-time multi-source heterogeneous data, determine the coupling coefficient solution logic between each single-physics field model, attach and bind the coupling coefficient solution logic to the corresponding core computing domain to generate a digital twin; Step s3: Quantify the insulation performance of the multi-source heterogeneous data, obtain the intrinsic insulation strength base value under the current working condition, and correct the intrinsic insulation strength base value according to the global coupling simulation results output by the digital twin to obtain the real-time insulation strength, real-time insulation margin and insulation state of the new fluorocarbon insulating gas under the current working condition. Step s4: When the insulation state is in the warning state, based on the global coupling simulation results of the digital twin, the real-time insulation strength and the real-time insulation margin, and combined with the pre-trained insulation degradation mechanism model, obtain the insulation margin time-series decay curve within the future set time window, and construct the control queue based on the insulation margin time-series decay curve. Step s5: Starting from the first position in the control queue, perform multi-target coordinated dynamic control of insulation performance in sequence, obtain the optimal control parameter scheme to be executed in sequence, and execute the corresponding control action according to the optimal control parameter scheme to be executed in sequence.
2. The method for dynamic optimization of the insulation performance of the novel fluorocarbon insulating gas in high and low voltage equipment according to claim 1, characterized in that, Multi-source heterogeneous data includes multi-dimensional state parameters of novel fluorocarbon insulating gases, equipment operating electrical parameters, and environmental parameters.
3. The method for dynamic optimization of the insulation performance of the novel fluorocarbon insulating gas in high and low voltage equipment according to claim 2, characterized in that, The process of performing liquefaction risk pre-verification and partial pressure correction on multi-source heterogeneous data includes: The main insulating gas, gas composition, pressure, and temperature parameters in the environmental parameters are obtained from the new fluorocarbon insulating gas. The saturated vapor pressure of the main insulating gas at the current temperature parameter is obtained. The pressure corresponding to the main insulating gas in the multi-dimensional state parameters is compared with the saturated vapor pressure. If the pressure of the main insulating gas is greater than the saturated vapor pressure, the pressure of the main insulating gas in the multi-dimensional state parameters is corrected to the saturated vapor pressure, and the gas composition of the main insulating gas is corrected simultaneously.
4. The method for dynamic optimization of the insulation performance of the novel fluorocarbon insulating gas in high and low voltage equipment according to claim 3, characterized in that, The process of building a digital twin includes: Extract 3D design drawings, system material parameters, equipment rating and boundary parameters, and test database of high and low voltage electrical equipment. Construct 3D geometric models of high and low voltage electrical equipment based on the 3D design drawings. Divide the 3D geometric model into multiple physical domains and split it into several core computational domains. Set API interfaces for each core computational domain. Assign attribute values to each core computational domain based on system material parameters and equipment rating and boundary parameters. Determine the material parameters and equipment rating and boundary parameters of each core computational domain. Match multi-source heterogeneous data with the core computational domains of the 3D geometric model to generate a digital twin model. A single-physics-field basic model is constructed, and the core computing domains of the single-physics-field basic model and the digital twin model are bound together through an API interface. A correlation function between multi-source heterogeneous data and material parameters of each core computing domain is constructed. The material parameters of each core computing domain are dynamically updated based on the correlation function and multi-source heterogeneous data. The material parameters of the single-physics-field basic model bound to each core computing domain are dynamically assigned based on the dynamically updated material parameters of each core computing domain. Boundary conditions are set for the single-physics-field basic model bound to each core computing domain based on the device rating and boundary parameters of each core computing domain. Based on the physical characteristics of high and low voltage electrical equipment filled with a new type of fluorocarbon insulating mixed gas, the coupling type between each single physical field basic model is obtained. According to different coupling types, the coupling coefficient solution logic between each single physical field basic model is set. The coupling coefficient solution logic between each single physical field basic model is mounted and bound to the core computing domain corresponding to each single physical field basic model through the API interface to generate a digital twin. The digital twin is subjected to secondary calibration and accuracy verification based on the experimental database to obtain a digital twin that has completed accuracy verification.
5. The method for dynamic optimization of the insulation performance of the novel fluorocarbon insulating gas in high and low voltage equipment according to claim 4, characterized in that, The process of quantifying the insulation performance of multi-source heterogeneous data to obtain the intrinsic insulation strength baseline value under the current operating conditions includes: A basic calculation model for the insulation strength of a novel fluorocarbon insulating gas is constructed. The input parameters of the basic calculation model are gas composition, pressure and temperature parameters, and the output parameter is the intrinsic insulation strength of the novel fluorocarbon insulating gas under standard operating conditions. The gas composition, pressure, and temperature parameters, after completing the pre-verification of liquefaction risk and partial pressure correction, are input into the basic calculation model of insulation strength. Based on the basic calculation model of insulation strength, the intrinsic insulation strength of the new fluorocarbon insulating gas under standard working conditions is output. Based on the real-time collected electrical parameters of the equipment operation, the current operating condition type is identified, and an impact coefficient reference table is constructed. The impact coefficient reference table includes the impact coefficients corresponding to different operating condition types. The impact coefficient is determined based on the current operating condition type and the impact coefficient reference table. The intrinsic insulation strength base value under the standard operating condition is corrected based on the impact coefficient to obtain the intrinsic insulation strength base value under the current operating condition.
6. The method for dynamic optimization of the insulation performance of the novel fluorocarbon insulating gas in high and low voltage equipment according to claim 5, characterized in that, Based on the global coupled simulation results output by the digital twin, the intrinsic insulation strength baseline value is corrected, and the process of obtaining the real-time insulation strength, real-time insulation margin, and insulation state of the novel fluorocarbon insulating gas under the current operating conditions includes: Based on the global coupled simulation results output by the digital twin, the dynamic correction coefficient matrix is obtained; The intrinsic insulation strength base value under the current operating condition is corrected according to the dynamic correction coefficient matrix to obtain the real-time insulation strength under the current operating condition. At the same time, the minimum allowable insulation strength under the current operating condition is obtained. If the real-time insulation strength is less than the minimum allowable insulation strength, an insulation abnormality alarm is immediately triggered. If the real-time insulation strength is greater than or equal to the minimum allowable insulation strength, the real-time insulation margin is obtained according to the real-time insulation strength and the minimum allowable insulation strength, and the insulation status is obtained according to the insulation margin.
7. The method for dynamic optimization of the insulation performance of the novel fluorocarbon insulating gas in high and low voltage equipment according to claim 6, characterized in that, The insulation degradation mechanism model includes a bottom-level mechanism sub-model, a data-driven self-correcting model, and a total rate solution model. The input of the bottom-level mechanism sub-model is the global coupled simulation result output by the digital twin, and the output is the insulation performance degradation rate. The data-driven self-correcting model takes as input the insulation performance degradation rate sequence of different degradation paths within the historical period, the global coupled simulation result sequence, and the measured insulation margin sequence, and outputs the correction coefficients for the insulation performance degradation rate of different degradation paths. The overall rate solution model is used to correct the insulation performance degradation rate output by the underlying mechanism sub-model corresponding to different degradation paths based on the correction coefficient of the insulation performance degradation rate of different degradation paths. The maximum insulation performance degradation rate is selected as the overall insulation degradation rate from the corrected insulation performance degradation rates.
8. The method for dynamic optimization of the insulation performance of the novel fluorocarbon insulating gas in high and low voltage equipment according to claim 7, characterized in that, The process of obtaining the insulation margin time-series decay curve within a future set time window includes: The global coupled simulation results of the digital twin at the current moment are input into the insulation degradation mechanism model. The total insulation degradation rate is obtained according to the insulation degradation mechanism model. The global coupled simulation results of the digital twin at the current moment, the real-time insulation strength, the insulation margin, and the total insulation degradation rate are used as the initial state for prediction. The prediction time window is set according to the current working conditions. Obtain the operating condition boundary condition sequence of high and low voltage electrical equipment within a future time window, input the operating condition boundary condition sequence within the future time window into a digital twin, and output the global coupled simulation result sequence for each time step within the future time window based on the digital twin; The sequence of global coupled simulation results for each time step is input into the insulation degradation mechanism model. Based on the insulation degradation mechanism model, the total insulation degradation rate and the degradation path corresponding to the total insulation degradation rate for each time step are output. Based on the total insulation degradation rate for each time step and the real-time insulation strength in the initial state, the cumulative insulation strength attenuation is obtained. Based on the cumulative insulation strength attenuation and the real-time insulation strength in the initial state, the real-time insulation strength for each future time step is obtained. Based on the real-time insulation strength for each future time step, the insulation margin for each future time step is obtained. Based on the insulation margin at each future time step and the insulation margin in the initial state, construct the insulation margin time-series decay curve within the future time window.
9. The method for dynamic optimization of the insulation performance of the novel fluorocarbon insulating gas in high and low voltage equipment according to claim 8, characterized in that, The process of constructing a control queue based on the insulation margin time-series decay curve includes: Within the insulation margin range, threshold points are selected to divide sub-intervals into different warning levels. The insulation margin time-series decay curve within the future time window is obtained and compared with the sub-intervals of different warning levels one by one. The warning level at different times within the future time window is obtained, and the degradation path corresponding to the total insulation degradation rate at different times within the future time window is obtained. A control priority comparison table is preset, which includes the control priority corresponding to different warning levels and degradation paths. Based on the control priority comparison table and the warning level and degradation path at different times within the future time window, the control priorities at different times within the future time window are generated, and the control priorities at different times are sorted to generate a control queue.
10. The method for dynamic optimization of the insulation performance of novel fluorocarbon insulating gases in high and low voltage equipment according to claim 9, characterized in that, Starting from the first position in the control queue, the process of performing multi-objective coordinated dynamic control of insulation performance to obtain the optimal control parameter scheme corresponding to each control priority includes: Several objective optimization functions are constructed, and the preset hard constraints corresponding to the current regulatory priority are obtained. Adaptive weights are set for different objective optimization functions according to the regulatory priority. A fitness function is constructed based on several objective optimization functions and adaptive weights. Several regulatory parameter schemes are preset. Chromosome encoding, hard constraint verification, and population initialization are performed on several regulatory parameter schemes to generate an initial population that meets the hard constraints. Based on the initial population and fitness function, the optimal regulatory parameter scheme corresponding to the current regulatory priority is obtained through a multi-objective genetic algorithm.
