Construction method of oil chromatography multi-physical field coupling model for multi-target evaluation of transformer

Abstract

The invention discloses a construction method of an oil chromatography multi-physics field coupling model for transformer multi-target evaluation. The construction method comprises the following steps: acquiring oil chromatography analysis data and operation parameters of a transformer; constructing a comprehensive state index model of multi-physics field coupling; defining a multi-objective optimization function and constraint conditions; constructing a non-dominated sorting genetic algorithm model with an elitist strategy, obtaining an optimal solution set of key parameters of the multi-physics-field coupled comprehensive state index model, and solidifying the multi-physics-field coupled comprehensive state index model; the optimized model is applied to transformer state evaluation, and finally the evaluation accuracy reaches 97.4%. The model has higher sensitivity and robustness, reliable evaluation results can be provided under various complex conditions, the accuracy of transformer operation state evaluation is high, and health monitoring and preventive maintenance of the transformer can be achieved.

Description

Technical Field

[0001] This invention relates to the field of power equipment condition monitoring and fault diagnosis technology, and in particular to a method for constructing an oil chromatography multiphysics coupling model for multi-objective evaluation of transformers. Background Technology

[0002] Power transformers are core equipment in the power grid, and their operating status directly affects the safety and stability of the power system. Real-time and accurate condition assessment and fault early warning of transformers are of great significance for implementing predictive maintenance, avoiding major accidents, and extending equipment life.

[0003] Currently, dissolved gas analysis (DGA) in insulating oil is one of the most widely used methods for diagnosing internal faults in oil-immersed transformers. Traditional diagnostic methods mainly rely on the characteristic gas composition method and the characteristic gas content ratio method. The characteristic gas composition method is experience-based, distinguishing fault types by the gas composition in the oil, primarily providing a qualitative description and lacking quantitative analysis. The characteristic gas content ratio method establishes a mapping relationship based on the relationship between gas concentration, temperature, and fault type. Among them, the modified three-ratio method is a commonly used fault discrimination method, which compensates for the order-of-magnitude differences in gas data during analysis and can accurately identify faults where the gas ratio deviates from the boundary value. However, due to the influence of various factors such as temperature, electric field, and humidity, the dissolved gas concentration in insulating oil varies non-linearly, increasing the complexity of the mapping between fault type and gas concentration. This leads to ambiguity in the relationship between equipment faults and codes, and the critical region of gas ratio coding is uncertain, resulting in low accuracy of this method in diagnosing fault severity. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method for constructing an oil chromatography multiphysics coupling model for multi-objective evaluation of transformers. This method can improve the accuracy of fault condition assessment and reduce the maintenance cost of transformers, thereby improving the safety, reliability and economy of transformer operation.

[0005] The present invention adopts the following technical solution: A method for constructing an oil chromatography multiphysics coupling model for multi-objective evaluation of transformers includes: S1: Obtain oil chromatography analysis data and operating parameters of the transformer; S2: Construct a comprehensive state index model with multi-physics coupling; S3: Define the multi-objective optimization function and constraints; S4: Construct a non-dominated sorting genetic algorithm model with an elite strategy to obtain the optimal solution set of key parameters of the integrated state index model with multi-physics coupling, and solidify the integrated state index model with multi-physics coupling. S5: Apply the optimized model to transformer condition assessment.

[0006] Furthermore, in step S1, the oil chromatographic analysis data includes the concentrations of hydrogen, methane, ethylene, and acetylene.

[0007] Further, in step S2, the expression for the comprehensive state index S is:

[0008] Where D is the discharge risk intensity, γ is the overheat risk intensity, and A is the aging risk intensity; α, β, and τ are the corresponding optimizable weight coefficients.

[0009] Furthermore, the discharge risk intensity D is obtained by using the gas ratio of acetylene to ethylene, R1, and by performing a continuous mapping based on the coding rules of the IEC 60599 standard; wherein the formula for R1 is:

[0010] and These are the volume concentrations of acetylene and ethylene in transformer insulating oil (unit: μL / L). Furthermore, the overheating risk intensity γ is obtained by continuous mapping based on the relative percentages of methane, ethylene, and acetylene in the total hydrocarbons and after locating the fault area using the Duval triangulation method. Furthermore, the aging risk intensity A is obtained based on the proportions of methane and hydrogen in the total combustible gas through a weighted calculation model, specifically as follows:

[0011] In the formula, and These are the weighting coefficients. To assess the risk intensity of methane aging, The intensity of hydrogen aging risk.

[0012] Furthermore, in step S4, the key parameters to be optimized include: weighting coefficients in the aging risk intensity calculation model. , And the weighting coefficients α, β, τ in the comprehensive state index model.

[0013] The beneficial effects of this invention are: By constructing a comprehensive state index model, oil chromatographic characteristics reflecting discharge, overheating, and aging are deeply correlated with operating physical fields such as voltage, current, and temperature. This overcomes the limitations of traditional methods with their single-feature approach and can more comprehensively characterize the complex internal states of equipment. For the first time, the transformer condition assessment problem is modeled as a multi-objective optimization problem involving safety, reliability, and lifespan, and solved using the advanced NSGA-II algorithm. This achieves a scientific trade-off among multiple operational objectives, resulting in assessment results that better align with actual engineering needs. The optimal model parameters are automatically determined through a data-driven optimization process, reducing reliance on subjective experience. The final model achieved an accuracy rate of 97.4% in testing, significantly outperforming the traditional ratio method, and enabling precise, objective, and quantitative assessment of transformer conditions. The constructed model can be directly integrated into transformer online monitoring systems to achieve real-time intelligent condition assessment and early warning, providing strong decision support for predictive maintenance, optimized maintenance strategies, and ensuring the safe and economical operation of the power grid. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the overall process of the method of the present invention. Figure 2 This is a schematic diagram illustrating the gas ratio characteristics obtained by the method of the present invention through oil chromatography monitoring; Figure 3 A coupled model architecture diagram for the multiphysics integrated state index is provided for this invention; Figure 4 This is a schematic diagram of the Duval triangle of the present invention; Figure 5 This is a flowchart illustrating the solution process based on the NSGA-II algorithm of this invention. Figure 6 This is a schematic diagram of the multi-objective evaluation of transformers based on the NSGA-II algorithm of this invention; Figure 7 This is a schematic diagram of the comprehensive evaluation index for transformers according to the present invention. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0016] Example 1 like Figures 1-7 As shown in the figure, this embodiment proposes a method for constructing an oil chromatography multiphysics coupling model for multi-objective evaluation of transformers, including: like Figure 1As shown in the figure, this embodiment proposes a method for constructing an oil chromatography multiphysics coupling model for multi-objective evaluation of transformers, including: S1: Transformer oil chromatography and operational data acquisition and preprocessing Acquire historical and real-time oil chromatographic analysis data of the target transformer, with key monitoring characteristic gases including hydrogen, methane, ethylene, and acetylene; simultaneously, collect relevant operating parameters of the transformer, including but not limited to operating voltage U and rated voltage Uo. n Hot spot temperature T, load current I, rated current I n wait; S2: Constructing a comprehensive state index model with multiphysics coupling The comprehensive state index S is composed of three coupled components: discharge risk intensity D, overheat risk intensity γ, and aging risk intensity A. Its model is shown in the formula:

[0017] In the formula, D, γ, A ∈ [0, 1], which respectively quantify the severity of discharge faults, overheating faults, and insulation aging risks. The calculation methods for each component are detailed below.

[0018] S2.1: Calculate the discharge risk intensity D The discharge risk intensity D is used to quantify the severity of discharge faults occurring inside a transformer. This invention is based on the three-ratio method in International Electrotechnical Commission IEC 60599:2015 (which is also equivalent to Chinese standard DL / T 722-2014), with the core basis being the ratio of acetylene to ethylene. / ).

[0019] a) Calculate the gas ratio R1 of acetylene to ethylene:

[0020] in and These are the volume concentrations of acetylene and ethylene in transformer insulating oil (unit: μL / L). b) According to the IEC 60599 standard, R1 is mapped to a specific fault type code, and the coding rules are shown in Table 1: Table 1. IEC 60599 Three-Ratio Method Gas Ratio Coding Rules

[0021] c) Convert discrete codes into continuous discharge risk intensity values To facilitate mathematical operations and optimization within the integrated model, the discrete IEC codes are converted into a continuous discharge risk intensity value, ranging from [0, 1]. The conversion function is defined as follows:

[0022] When D=0.1, it indicates that there is no significant risk of discharge and the equipment is in normal condition; When D=0.5, it indicates a moderate level of discharge risk, corresponding to low-energy discharge, requiring enhanced monitoring. When D=1.0, it indicates a serious risk of discharge, corresponding to high-energy discharge, and maintenance measures should be taken immediately. S2.2: Calculate the overheating risk intensity γ The overheat risk intensity γ is used to quantify the severity and temperature range of overheating faults occurring inside the transformer. This invention is based on the Duval Triangle Method for quantification, which locates the fault by analyzing the relative percentages of three gases: methane, ethylene, and acetylene.

[0023] a) Calculate the relative percentages of methane, ethylene, and acetylene in the total hydrocarbons (CH4+C2H4+C2H4):

[0024] The three percentages calculated satisfy: b) Using coordinates Locate the fault area in the Duval triangle diagram; like Figure 4 As shown, the type and severity of the overheating fault are determined based on the area where the coordinate point falls in the Duval triangle diagram. The key regions related to overheating defined by the Duval triangle method include: T1 region: thermal failure, temperature t < 300℃ (low-temperature overheating) T2 Zone: Thermal fault, temperature 300℃≤t≤700℃ (medium-temperature overheating) T3 region: Thermal failure, temperature t > 700℃ (high temperature overheating) c) Based on the landing point region, convert it into a continuous overheating risk intensity value γ∈ [0, 1], and the conversion rules are as follows:

[0025] In this quantification scheme, When γ=0.1, it indicates that the current fault mode is non-overheating type (such as partial discharge or arc discharge), and the risk of overheating can be ignored; When γ=0.4, it indicates a risk of low-temperature overheating, usually caused by overheating of the magnetic circuit or connector, which requires attention; When γ=0.7, it indicates a risk of overheating at medium temperature, which involves solid insulation and should be scheduled for maintenance. When γ=1.0, it indicates a serious risk of overheating, which may endanger equipment safety and requires immediate action. S2.3: Calculate the aging risk intensity A The aging risk intensity A is used to quantify the overall aging status of the transformer oil-paper insulation system. This invention quantifies the aging status of the insulation system based on the proportion of methane and hydrogen in the total combustible gas (TCG). a) Calculate the total concentration of combustible gases:

[0026] b) Calculate the proportion of key gases:

[0027] In the formula, The percentage of methane, The percentage of hydrogen; c) Constructing an aging risk intensity calculation model: The proportion of key gases is converted into a continuous aging risk intensity A, with a value range of [0, 1]. The conversion function is defined as follows: Methane is the main product of insulating oil and insulating paper during the thermal aging process, and its relative content directly reflects the degree of degradation of the insulating material.

[0028] An abnormally high hydrogen content may be related to the dehydration or cracking of cellulose materials and is an auxiliary indicator for judging insulation aging.

[0029]

[0030]

[0031] In the formula, and The A value is a weighting coefficient that reflects the difference in the contribution of methane and hydrogen to aging. The higher the A value, the deeper the thermal aging or cracking of the insulating paper system. S3: Define the multi-objective optimization function and constraints To achieve a comprehensive and optimal evaluation of the transformer's operating status, the following three maximization objective functions and corresponding physical constraints are established; The objective function for maximizing the operational safety of the transformer, as described in Objective Function 1, is: (1) In the formula, As a risk driver, Operating voltage Rated voltage, For hot spot temperature, It is the influence coefficient of temperature; Objective function 2: The objective function for maximizing the reliability performance of the transformer is: (2) In the formula, It is a scaling factor that adjusts the weighting of the overheating risk. For load current, This is the rated current.

[0032] Objective function 3: The objective function for maximizing the transformer's lifespan is: (3) In the formula, It is a coefficient representing economic loss, which can transform the risk of aging into economic costs. It is the lifespan depreciation rate caused by temperature.

[0033] The constraints are as follows: Voltage constraint (4) Load current constraint (5) Temperature constraint (6) Electrothermal Co-constraint (7) S4: Multi-objective optimization solution based on NSGA-II algorithm The key adjustable parameter set X in the comprehensive state index model and objective function is used as the decision variable, specifically including: The weighting coefficients ω1 and ω2 in the aging risk intensity calculation, and the adjustment coefficients in the three objective functions—discharge risk intensity coefficient α, overheating risk intensity coefficient β, and aging risk intensity coefficient τ—are used to solve the above multi-objective optimization problem. A fast non-dominated sorting genetic algorithm with an elitist strategy (NSGA-II) is employed to select the final solution set X that best matches the decision preference from the Pareto optimal solution set. optimal = [ω1, ω2,α, β, τ]. will optimal The determined parameter values ​​are incorporated into the multiphysics coupling integrated state index model formula and the calculation model formula for each component of this invention.

[0034] A transformer intelligent assessment model, optimized through multiple objectives and capable of comprehensively balancing safety, reliability, and lifespan, has been constructed and can be used for online monitoring and diagnostic analysis.

[0035] S5: Model Evaluation and Application. The optimized model is applied to an independent test set. The comprehensive state index S is calculated for each sample. By comparing the model output with the actual state label, the numerical range of S is obtained to classify the transformer state according to the following criteria: normal state, warning state, abnormal state, and severe fault state.

[0036] Experimental results show that the optimized comprehensive evaluation index can accurately reflect the operating status of the transformer, with an evaluation accuracy of up to 97.4%. Compared with traditional methods, the proposed model has higher sensitivity and robustness, and can provide reliable evaluation results under various complex conditions.

Claims

1. A method for constructing an oil chromatography multiphysics coupling model for multi-objective evaluation of transformers, characterized in that: include: S1: Obtain oil chromatography analysis data and operating parameters of the transformer; S2: Construct a comprehensive state index model with multi-physics coupling; S3: Define the multi-objective optimization function and constraints; S4: Construct a non-dominated sorting genetic algorithm model with an elite strategy to obtain the optimal solution set of key parameters of the integrated state index model with multi-physics coupling, and solidify the integrated state index model with multi-physics coupling. S5: Apply the optimized model to transformer condition assessment.

2. The method for constructing an oil chromatography multiphysics coupling model for multi-objective evaluation of transformers according to claim 1, characterized in that: In step S1, the oil chromatographic analysis data includes the concentrations of hydrogen, methane, ethylene, and acetylene.

3. The method for constructing an oil chromatography multiphysics coupling model for multi-objective evaluation of transformers according to claim 1, characterized in that: In step S2, the expression for the comprehensive state index S is: ； Where D is the discharge risk intensity, γ is the overheat risk intensity, and A is the aging risk intensity; α, β, and τ are the corresponding optimizable weight coefficients.

4. The method for constructing an oil chromatography multiphysics coupling model for multi-objective evaluation of transformers according to claim 3, characterized in that: The discharge risk intensity D is obtained by using the gas ratio of acetylene to ethylene, R1, and through continuous mapping based on the coding rules of the IEC 60599 standard; where the formula for R1 is: ； and These represent the volume concentrations of acetylene and ethylene in transformer insulating oil, respectively, in μL / L.

5. The method for constructing an oil chromatography multiphysics coupling model for multi-objective evaluation of transformers according to claim 3, characterized in that: The overheating risk intensity γ is obtained by continuous mapping based on the relative percentages of methane, ethylene, and acetylene in the total hydrocarbons, after locating the fault area using the Duval triangulation method.

6. The method for constructing an oil chromatography multiphysics coupling model for multi-objective evaluation of transformers according to claim 3, characterized in that: The aging risk intensity A is obtained based on the proportions of methane and hydrogen in the total combustible gas, using a weighted calculation model, specifically: ； In the formula, and These are the weighting coefficients. To assess the risk intensity of methane aging, The intensity of hydrogen aging risk.

7. The method for constructing an oil chromatography multiphysics coupling model for multi-objective evaluation of transformers according to claim 1, characterized in that: In step S4, the key parameters to be optimized include: weighting coefficients in the aging risk intensity calculation model. , And the weighting coefficients α, β, τ in the comprehensive state index model.

Images