Dry-type transformer temperature prediction method based on accurate thermal network model

Abstract

A dry-type transformer temperature prediction method based on an accurate thermal network model is characterized in that the structure of a transformer is accurately divided according to the difference between the actual structure and the material attribute of the dry-type transformer, and convective thermal resistance and radiant thermal resistance are iteratively calculated in different time in the fine thermal network model, so that the heat transfer process of the transformer is more practical. The established precise thermal network model shortens the solving time and accelerates the temperature prediction process under the condition of ensuring the solving precision; the method can achieve the prediction of the overall temperature of the transformer under special working conditions such as overload and short circuit, and verifies whether the maximum steady-state temperature of an insulation system of the transformer is lower than the maximum allowable temperature of an insulation material or not and whether the transient temperature rise rate is lower than the strength attenuation rate of the insulation material or not under all expected severe working conditions. The grade design of the insulation system of the dry-type transformer is completed, and guidance can be provided for ventilation and heat dissipation design of the dry-type transformer.

Description

Technical Field

[0001] This invention belongs to the field of marine shipbuilding and design, and specifically relates to a method for predicting the temperature of dry-type transformers based on an accurate thermal network model. Background Technology

[0002] The harsh marine environment, cramped engine room, and insufficient ventilation limit heat dissipation. During navigation, dry-type transformers may be exposed to prolonged high temperatures. Furthermore, overloads and malfunctions in the ship's electrical system can cause rapid load fluctuations, potentially leading to emergency situations. In such circumstances, transformer temperatures can rise dramatically, damaging the insulation structure and increasing the risk of fire. Accurately predicting the temperature of the hottest spot is crucial to preventing insulation damage and mitigating fire risks.

[0003] The insulation class of dry-type transformers is highly dependent on the expected hot spot temperature, which is obtained by adding the ambient temperature, the average temperature rise of the winding, and the hot spot temperature margin. Traditional temperature prediction methods rely solely on standard formulas or empirical values, which are usually highly simplified empirical models. These models cannot accurately account for complex internal heat exchange, non-uniform heat dissipation conditions, transient processes of rapid load changes, or the impact of poor heat dissipation conditions. Therefore, traditional temperature prediction methods cannot accurately reflect the actual temperature changes of the transformer. This may lead to insufficient insulation class, which will accelerate transformer aging, reduce service life, and increase maintenance costs. Alternatively, blindly increasing the insulation class may lead to a surge in transformer costs, increased size, and maintenance difficulties.

[0004] Accurate thermal network models based on electrothermal analogy can accurately predict the hottest temperatures of the windings and the temperature distribution in various regions of the core and windings, with fast calculation speed. These models can predict the global temperature of the transformer under special operating conditions such as overload and short circuits based on load current, ambient temperature, and heat dissipation conditions. By verifying whether the transformer's hottest temperatures are below the maximum allowable temperature of the insulation material under all anticipated severe operating conditions, and with sufficient safety margins, the insulation design of dry-type transformers can be completed by comprehensively considering both economic and reliability aspects. Accurate thermal network models can also show the effects of natural air cooling, forced air cooling, and different heat dissipation configurations in harsh high-temperature environments, helping to select the most reliable cooling method. Understanding the temperature characteristics of transformers under different loads and environments helps ship operators develop better operating strategies, maximizing equipment utilization while ensuring safety. Summary of the Invention

[0005] To address the above problems, this invention provides a method for predicting the temperature of dry-type transformers based on an accurate thermal network model. The technical solution adopted is as follows: A method for predicting the temperature of a dry-type transformer based on an accurate thermal network model is proposed. The transformer is a three-phase dry-type transformer with three magnetic core columns fixed between the upper and lower jaws of the magnetic core. The magnetic core columns are wrapped with low-voltage windings and insulators in sequence. Multiple high-voltage windings are arranged from top to bottom inside the insulators, and there are gaps between the multiple high-voltage windings.

[0006] The specific steps are as follows: S1: The dry-type transformer structure is finely divided, with a high-voltage winding and its corresponding low-voltage winding as the unit, to obtain the smallest thermal network conductor unit.

[0007] S2: During transformer operation, the heat sources include core loss and winding loss. The input core loss Pv in the thermal network model is calculated as follows: .

[0008] In the formula K h This is the hysteresis loss coefficient. B is the hysteresis loss coefficient. m K represents the peak magnetic flux density. e K is the eddy current loss coefficient. a This is the abnormal loss coefficient.

[0009] Input ambient temperature T, effective value of current per phase I rms Given the conductor cross-sectional area S and conductor perimeter Lc, calculate the winding loss Pc per phase, set in the thermal network model, as follows: .

[0010] .

[0011] In the formula, ρ(T) is the resistivity; ρT0 is the resistivity at the reference temperature T0; α is the temperature resistivity coefficient; the winding current changes under overload and short circuit conditions of the dry-type transformer, correcting the core loss and winding loss.

[0012] S3: Calculate the thermal resistances Rx and Ry and the heat capacity Cd, such as Figure 3 The minimum thermal network conductor unit obtained from S1 is shown below. Inputting the thermal network unit length L, and the inner and outer radii r1 and r0 of the conductor, the radial thermal resistance of the minimum thermal network conductor unit is calculated as follows: .

[0013] In the formula, λ is the thermal conductivity of the material.

[0014] The contact area Sy of the input unit is used to calculate the axial thermal resistance as follows: .

[0015] The heat capacity Cd can be used to quantitatively analyze the transient temperature changes of a transformer, such as... Figure 3 As shown in Figure 1.3, given the input conductor volume V, the electrical parameter heat capacity is calculated using the following formula: .

[0016] In the formula, Cth is the heat capacity of the conductor, and ρ is the density of the conductor.

[0017] S4: Construct an accurate thermal network model of the transformer using the transformer structure divided in step S1 with the smallest thermal network unit. Input winding loss Pc and core loss Pv are represented by current sources, and ambient temperature Ta is represented by voltage sources. Solve for the initial global node temperature T1(k,m) of the transformer, where k is the number of nodes in the thermal network model and m is the number of iterations.

[0018] S5: Convection thermal resistance Rconv represents the ability of heat exchange between the fluid and the solid surface, and radiation winding Rrad reflects the resistance to heat dissipation through radiation. Based on the solved node temperature T1(k,m), the convection thermal resistance Rconv and radiation winding Rrad of the fluid-solid interface are determined.

[0019] Under natural convection conditions (without a fan), the input surface dimension La along the airflow direction and the transformer surface temperature T are... s Transformer surface temperature T s The contact area between the transformer and the fluid is S. The convective thermal resistance Rconv1 is calculated using the following formula: .

[0020] .

[0021] In the formula h c1 is the convective heat transfer coefficient; C is an empirical coefficient, usually taken as 1.3~0.7; For the surface temperature difference, T a The ambient temperature.

[0022] Under forced convection conditions, given the input channel diameter D, air velocity v, and transformer-fluid contact area S, calculate the convection thermal resistance Rconv2 using the following formula: .

[0023] K a ρ is the thermal conductivity of air; a ρ is the air density; μ is the aerodynamic viscosity; Pr is the Planck coefficient, which is 0.7 for air.

[0024] Given the ambient temperature Ta, transformer surface temperature Ts, and transformer-fluid contact area S, calculate the radiation thermal resistance Rrad using the following formula: .

[0025] .

[0026] hr is the radiation coefficient; For surface emissivity, the value for epoxy resin is 0.9; The Stefan-Boltzmann constant has a value of 5.67 × 10⁻⁸ W / m²K⁴.

[0027] S6: Correct the winding loss Pc in step S2 using the initial global temperature node T1(k,m) from step S4.

[0028] S7: Using the convective thermal resistance Rconv, the radiative winding Rrad, and the corrected winding loss Pc obtained in step S5, resolve the thermal network model. The ambient temperature remains unchanged, and the global thermal network model T2(k,m) is obtained.

[0029] S8: Set the temperature accuracy ε1 for the solution, compare the global transformer temperatures T1(k,m) and T2(k,m), and if the conditions are met... If the condition is not met, proceed to the next step. If not, return T2(k,m) to step S4 as the initial temperature for iteration, and proceed to steps S5~S8 until the requirement is met.

[0030] S9: Using the node temperature T2(k,m) as the initial value, and the convective thermal resistance Rconv, radiation winding Rrad, and corrected winding loss Pc completed in S5~S8, solve for the global temperature T3(k,m) within the time interval Δt.

[0031] S10: Set the solution temperature accuracy ε2, compare T3(k,m) and T2(k,m), and if satisfied... If the condition is not met, proceed to the next step. If not, return T3(k,m) to step S4 as the initial temperature T1(k,m) for iteration, and execute steps S4 to S9 until the condition is met, and record the iteration number m.

[0032] S11: During the iteration process, the global temperature T3(k,m) is obtained for each iteration. The temperature T3(k,m) of each node within the time T=m*Δt is plotted as a temperature rise curve to obtain the global temperature change curve of the transformer.

[0033] S12: Based on the global temperature change curve, obtain the maximum steady-state temperature and the maximum transient temperature change rate of the insulation material, and evaluate the insulation system.

[0034] If the maximum steady-state temperature is less than the temperature index of the insulating material and the maximum transient temperature change rate is less than the insulation strength attenuation rate, the insulation system is assessed as reliable and the design is terminated.

[0035] If either of these conditions is not met, the insulation class should be increased, and steps S1 to S13 should be repeated until the insulation system assessment is reliable.

[0036] Furthermore, the aforementioned method for predicting the temperature of a dry-type transformer based on an accurate thermal network model further includes the existence of gaps between the magnetic core column and the low-voltage winding, and between the low-voltage winding and the high-voltage winding, forming air channels.

[0037] The aforementioned method for predicting the temperature of a dry-type transformer based on an accurate thermal network model further specifies that ε1 = 0.01. ， ε2=0.01.

[0038] Furthermore, in steps S4 and S8, the ambient temperature is used as a boundary condition when solving the thermal network model, and the boundary condition is corrected when the ambient temperature changes.

[0039] The above-mentioned method for predicting the temperature of a dry-type transformer based on an accurate thermal network model further includes step S9, in which the ambient temperature is used as a boundary condition to calculate the transient temperature field and solve for the global temperature T3(k,m) within time Δt.

[0040] Furthermore, the aforementioned method for predicting the temperature of a dry-type transformer based on an accurate thermal network model further includes four high-voltage windings arranged from top to bottom within the insulation body.

[0041] The beneficial effects of this invention are: 1. Based on the differences in the actual structure and material properties of dry-type transformers, the transformer structure is accurately divided, and a refined thermal network model is established. The thermal network model improves the efficiency of design iteration and optimization, and accelerates the temperature prediction process.

[0042] 2. Convection thermal resistance and radiation thermal resistance have different heat dissipation capabilities at different temperatures. Iterative calculation of convection thermal resistance and radiation thermal resistance at different time points in the fine thermal network model makes the heat transfer process of the transformer more realistic.

[0043] 3. Predict the rate of temperature rise and peak value of the transformer under conditions such as startup, overload, short circuit, ambient temperature change, and heat dissipation failure of the dry-type transformer. Based on whether the maximum steady-state temperature of the transformer's insulation system is lower than the maximum allowable temperature of the insulation material and whether the transient temperature rise rate is less than the strength decay rate of the insulation material, complete the insulation system level design of the dry-type transformer, evaluate the transformer's overload capacity and overload time, assess the damage risk, and provide a theoretical basis for the design of the ventilation and heat dissipation system. Attached Figure Description

[0044] Figure 1This is a schematic diagram of a three-phase dry-type transformer. Figure 2 A schematic diagram of local heat transfer in a three-phase dry-type transformer; Figure 3 This is a schematic diagram of the smallest heat network unit; Figure 4 This is a schematic diagram of the process of this invention; Among them, 1-magnetic core upper jaw, 2-magnetic core column, 3-high voltage winding, 4-low voltage winding, 5-air passage, 6-insulation, 7-magnetic core lower jaw, 1.1-axial thermal resistance, 1.2-radial thermal resistance, 1.3-heat source, 1.4-heat capacity, 2.1-convective thermal resistance, 2.2-radiative thermal resistance, 2.3-ambient temperature. Detailed Implementation

[0045] The present invention will be described in detail with reference to the accompanying drawings.

[0046] like Figure 1 The method shown is a method for predicting the temperature of a dry-type transformer based on an accurate thermal network model. The transformer has three core columns, which are fixed between the upper and lower core jaws. The core columns are wrapped with low-voltage windings and insulators in sequence. Multiple high-voltage windings are arranged from top to bottom in the insulators, and there are gaps between the multiple high-voltage windings.

[0047] The specific steps are as follows: S1: Dry-type transformer structure division. Based on the different materials used in the radial and axial directions of the dry-type transformer, a fine-grained regional division is performed, resulting in the smallest thermal network conductor unit, such as... Figure 2 As shown.

[0048] S2: During transformer operation, the heat sources include core loss and winding loss. The input core loss Pv in the thermal network model is calculated as follows: .

[0049] In the formula K h This is the hysteresis loss coefficient. B is the operating frequency of the transformer. m K represents the peak magnetic flux density. e K is the eddy current loss coefficient. a This is the abnormal loss coefficient.

[0050] Input ambient temperature T, effective value of current per phase I rms Given the conductor cross-sectional area S and conductor perimeter Lc, calculate the winding loss Pc per phase, set in the thermal network model, as follows: .

[0051] .

[0052] In the formula, ρ(T) is the resistivity; ρT0 is the resistivity at the reference temperature T0; α is the temperature resistivity coefficient; the winding current changes under overload and short circuit conditions of the dry-type transformer, correcting the core loss and winding loss.

[0053] S3: Calculate the thermal resistances Rx and Ry and the heat capacity Cd, such as Figure 3 The minimum thermal network conductor unit obtained from S1 is shown below. Inputting the thermal network unit length L, and the inner and outer radii r1 and r0 of the conductor, the radial thermal resistance of the minimum thermal network conductor unit is calculated as follows: .

[0054] In the formula, λ is the thermal conductivity of the material.

[0055] The contact area Sy of the input unit is used to calculate the axial thermal resistance as follows: .

[0056] The heat capacity Cd can be used to quantitatively analyze the transient temperature changes of a transformer, such as... Figure 3 As shown in Figure 1.3, given the input conductor volume V, the electrical parameter heat capacity is calculated using the following formula: .

[0057] In the formula, Cth is the heat capacity of the conductor, and ρ is the density of the conductor.

[0058] S4: Construct an accurate thermal network model of the transformer using the minimum thermal network unit based on the transformer structure divided in S1. In input S1, winding losses Pc and core losses Pv are represented by current sources, and ambient temperature Ta is represented by a voltage source. Figure 3 As shown in Figure 1.4, the global node temperature T1(k,m) of the initial transformer is solved, where k is the number of nodes in the thermal network model and m is the number of iterations.

[0059] S5: When solving the thermal network model, the ambient temperature is used as the boundary condition. When the ambient temperature changes, the boundary condition is modified.

[0060] S6: Convection thermal resistance Rconv represents the ability of heat exchange between the fluid and the solid surface, and radiation winding Rrad reflects the resistance to heat dissipation through radiation. Based on the solved node temperature T1(k,m), the convection thermal resistance Rconv and radiation winding Rrad of the fluid-solid interface are determined.

[0061] Under natural convection conditions (without a fan), input the surface dimension La along the airflow direction, the transformer surface temperature Ts, and the transformer-fluid contact area S, and calculate the convective thermal resistance Rconv1, as shown in the following formula: .

[0062] In the formula h c1 is the convective heat transfer coefficient; C is an empirical coefficient, usually taken as 1.3~0.7; Ta represents the surface temperature difference, and Ta represents the ambient temperature.

[0063] Under forced convection (with a fan), given the input channel diameter D, air velocity v, and transformer-fluid contact area S, calculate the convective thermal resistance Rconv2 using the following formula: .

[0064] In the formula, Ka is the thermal conductivity of air; ρ a ρ is the air density; μ is the aerodynamic viscosity; Pr is the Planck coefficient, which is 0.7 for air.

[0065] Given the ambient temperature Ta, transformer surface temperature Ts, and transformer-fluid contact area S, calculate the radiation thermal resistance Rrad using the following formula: .

[0066] .

[0067] In the formula, hr is the radiation coefficient; For surface emissivity, the value for epoxy resin is 0.9; The Stefan-Boltzmann constant has a value of 5.67 × 10⁻⁸ W / m²K⁴.

[0068] S7: Correct the winding loss Pc in S2 based on the initial global temperature node T1(k,m) obtained from S4.

[0069] S8: Using the convective thermal resistance Rconv, the radiative winding Rrad calculated in S6, and the corrected winding loss Pc in S7, the thermal network model is solved again. The ambient temperature remains unchanged, and S5 is still used as the boundary condition. The global thermal network model T2(k,m) is obtained.

[0070] S9: Set the solution temperature accuracy The smaller the value, the more accurate the global temperature calculation. Too small a value leads to too many iterations, increasing the solution time. Comparing the solution for the global transformer temperature T1(k,m) in S8 and the solution for T2(k,m) in S4, the formulas are as follows: .

[0071] If the above equation is satisfied, proceed to the next step; otherwise, return T2(k,m) to S4 as the initial temperature T1(k,m) and iterate through S6 to S9 until the above equation is satisfied.

[0072] S10: Using the node temperature T2(k,m) as the initial value, the convective thermal resistance Rconv, radiation winding Rrad, and corrected winding loss Pc completed by the iterative meter in S6~S9, and the ambient temperature in S5 as the boundary condition, the transient temperature field is calculated to solve the global temperature T3(k,m) within the time interval Δt.

[0073] S11: Set the temperature calculation precision ε2 = 0.01. The smaller the value, the more accurate the global temperature calculation. However, too small a value for ε2 leads to too many iterations, increasing the solution time. Compare the calculation of the global transformer temperature T3(k,m) in S10 with the calculation of T2(k,m) in S8. The formulas are as follows: .

[0074] If the above equation is satisfied, proceed to the next step. If not, return T3(k,m) to S4 as the initial temperature T1(k,m) and iterate through S4 to S10. Continue this iterative calculation until the difference between two temperature values ​​is less than the given precision value ε2.

[0075] S12: If the conditions in S11 are not met, return T3(k,m) to S4 as the initial temperature T1(k,m) and iterate through S4~S10, recording the iteration number m.

[0076] S13: Iterate through S4 to S10 until the temperature judgment formula in S11 is satisfied, and obtain the global temperature T3(k,m) for each time. Plot the temperature T3(k,m) of each node within time T=m*Δt as a temperature rise curve, and you can obtain the global temperature change curve of the transformer.

[0077] S14: Based on the global temperature change curve obtained in S13, obtain the maximum steady-state temperature and the maximum transient temperature change rate of the insulating material, and evaluate the insulation system: the maximum steady-state temperature is less than the temperature index of the insulating material, and the maximum transient temperature change rate is less than the insulation strength attenuation rate. If both are satisfied, the insulation system design is considered reliable. If either of them is not satisfied, the insulation level is increased, and S1~S14 are repeated until the insulation system evaluation is reliable.

Claims

1. A method for predicting the temperature of a dry-type transformer based on an accurate thermal network model, characterized in that, There is a three-phase dry-type transformer. Three magnetic core columns are fixed between the upper jaw and the lower jaw of the magnetic core. Low-voltage windings and insulators are wrapped around the magnetic core columns in sequence. Multiple high-voltage windings are arranged from top to bottom inside the insulators, and there are gaps between the multiple high-voltage windings. The specific steps are as follows: S1: The dry-type transformer structure is finely divided, with a high-voltage winding and its corresponding low-voltage winding as the unit, to obtain the smallest thermal network conductor unit; S2: During transformer operation, the heat sources include core loss and winding loss. Input the core loss Pv into the thermal network model. ； In the formula K h This is the hysteresis loss coefficient. B is the operating frequency of the transformer. m K represents the peak magnetic flux density. e K is the eddy current loss coefficient. a This is the abnormal loss coefficient; By inputting the ambient temperature T, the effective value of the current per phase Irms, the conductor cross-sectional area S, and the conductor perimeter Lc into the thermal network model, the winding loss Pc of each phase can be obtained. ； ； In the formula, ρ(T) is the resistivity; ρT0 is the resistivity at the reference temperature T0; α is the temperature resistivity coefficient; the winding current changes under overload and short circuit conditions of the dry-type transformer, correcting for core loss and winding loss; S3: Based on the minimum thermal network conductor unit obtained in step S1, input the thermal network unit length L, the inner and outer radii r1 and r0 of the conductor, the contact area Sy of the unit, and the volume V of the unit conductor into the thermal network model, and calculate the radial conduction thermal resistance Rx, the axial heat transfer thermal resistance Ry, and the fusion temperature Cd of the minimum thermal network conductor unit. ； In the formula, λ is the thermal conductivity of the material; ； ； In the formula, Cth is the heat capacity of the conductor, and ρ is the density of the conductor; S4: Construct an accurate thermal network model of the transformer using the transformer structure divided in step S1 with the smallest thermal network unit. Input winding loss Pc and core loss Pv are represented by current sources, and ambient temperature Ta is represented by voltage sources. Solve for the initial global node temperature T1(k,m) of the transformer, where k is the number of nodes in the thermal network model and m is the number of iterations. S5: Convection thermal resistance Rconv represents the ability of heat exchange between the fluid and the solid surface, and radiation winding Rrad reflects the resistance to heat dissipation through radiation. Based on the solved node temperature T1(k,m), determine the convection thermal resistance Rconv and radiation winding Rrad of the fluid-solid interface. Under natural convection conditions, given the surface dimension La along the airflow direction, the transformer surface temperature Ts, and the transformer-fluid contact area S, calculate the convective thermal resistance Rconv1 using the following formula: ； ； In the formula h c1 is the convective heat transfer coefficient; C is an empirical coefficient, usually taken as 1.3~0.7; For the surface temperature difference, T a Ambient temperature; Under forced convection conditions, given the input channel diameter D, air velocity v, and transformer-fluid contact area S, calculate the convection thermal resistance Rconv2 using the following formula: ； ； In the formula K a ρ is the thermal conductivity of air; a air density; μ is aerodynamic viscosity; Pr is Planck's coefficient, which is 0.7 for air. Given the ambient temperature Ta, transformer surface temperature Ts, and transformer-fluid contact area S, calculate the radiation thermal resistance Rrad using the following formula: ； ； In the formula, hr is the radiation coefficient; For surface emissivity, the value for epoxy resin is 0.9; The Stefan-Boltzmann constant has a value of 5.67 × 10⁻⁸ W / m²K⁴. S6: Correct the winding loss Pc in step S2 using the initial global temperature node T1(k,m) in step S4; S7: Using the convective thermal resistance Rconv, the radiative winding Rrad, and the corrected winding loss Pc obtained in step S5, the thermal network model is solved again. The ambient temperature remains unchanged, and the global thermal network model T2(k,m) is obtained. S8: Set the temperature accuracy ε1 for the solution, compare the global transformer temperatures T1(k,m) and T2(k,m), and if the conditions are met... If the condition is not met, proceed to the next step. If not, return T2(k,m) to step S4 as the initial temperature for iteration, and proceed to steps S5~S8 until the requirement is met. S9: Using the node temperature T2(k,m) as the initial value, and the convective thermal resistance Rconv, radiative winding Rrad, and corrected winding loss Pc completed in S5~S8, solve for the global temperature T3(k,m) within the time interval Δt. S10: Set the solution temperature accuracy ε2, compare T3(k,m) and T2(k,m), and if satisfied... If the condition is not met, proceed to the next step. If not, return T3(k,m) to step S4 as the initial temperature T1(k,m) for iteration, and execute steps S4 to S9 until the condition is met, and record the iteration number m. S11: During the iteration process, the global temperature T3(k,m) is obtained for each iteration. The temperature T3(k,m) of each node within the time T=m*Δt is plotted as a temperature rise curve to obtain the global temperature change curve of the transformer. S12: Based on the global temperature change curve, obtain the maximum steady-state temperature and maximum transient temperature change rate of the insulation material, and evaluate the insulation system; If the maximum steady-state temperature is less than the temperature index of the insulating material and the maximum transient temperature change rate is less than the insulation strength attenuation rate, the insulation system is assessed as reliable and the design is terminated. If either of these conditions is not met, the insulation class should be increased, and steps S1 to S13 should be repeated until the insulation system assessment is reliable.

2. The method for predicting the temperature of a dry-type transformer based on an accurate thermal network model according to claim 1, characterized in that, There are gaps between the magnetic core column and the low-voltage winding, and between the low-voltage winding and the high-voltage winding, forming air passages.

3. The method for predicting the temperature of a dry-type transformer based on an accurate thermal network model according to claim 1, characterized in that, ε1=0.01 ， ε2=0.

01.

4. The method for predicting the temperature of a dry-type transformer based on an accurate thermal network model according to claim 1, characterized in that, In steps S4 and S8, the ambient temperature is used as a boundary condition when solving the thermal network model. When the ambient temperature changes, the boundary condition is corrected.

5. The method for predicting the temperature of a dry-type transformer based on an accurate thermal network model according to claim 1, characterized in that, In step S9, the ambient temperature is used as the boundary condition to calculate the transient temperature field and solve for the global temperature T3(k,m) within the time interval Δt.

6. The method for predicting the temperature of a dry-type transformer based on an accurate thermal network model according to claim 1, characterized in that, There are four high-voltage windings arranged from top to bottom inside the insulator.

Images