Transformer thermal balance overload risk early warning system for extreme weather coupling scenarios

Abstract

The application relates to the technical field of transformer thermal balance overload risk early warning, and particularly discloses a transformer thermal balance overload risk early warning system for an extreme weather coupling scenario, which comprises a micro-weather dynamic acquisition module, which is used for acquiring the wind speed, the ambient temperature, the relative humidity and the solar radiation intensity of the micro-environment where the transformer is located in real time, identifying and removing abnormal outliers by adopting a median filtering method, uniformly quantifying the three factors of the wind speed, the humidity and the radiation by constructing an extreme weather coupling coefficient, and segmentally and dynamically correcting equivalent heat dissipation thermal resistance and thermal time constants according to the three factors, so that the heat path model has the following capacity: following the change of the weather, and the prediction accuracy of the top oil temperature under a complex weather scenario is improved, which avoids the prediction failure of a fixed thermal parameter model when the weather suddenly changes.

Description

Technical Field

[0001] This invention relates to the field of transformer thermal balance overload risk early warning technology, specifically a transformer thermal balance overload risk early warning system for extreme weather coupled scenarios. Background Technology

[0002] As a core hub device in the power system, the accurate monitoring of the thermal balance of transformers and the timely early warning of overload risks are crucial to ensuring the safe and stable operation of the power grid. Under extreme weather conditions, parameters such as wind speed, humidity, temperature and radiation change drastically, which significantly affects the heat dissipation conditions of transformers. Therefore, an early warning system is needed to provide timely warnings of transformer thermal balance overload risks.

[0003] In existing transformer thermal condition monitoring and early warning technologies, most thermal circuit models use fixed equivalent heat dissipation thermal resistance and thermal time constant, without taking into account the dynamic coupling effect of micro-meteorological factors. Early warning methods generally lack a rapid response mechanism for extreme gust events, and overload risk early warning indicators are usually based only on the comparison between the current instantaneous temperature value and a fixed threshold, failing to fully consider the temperature-time cumulative damage effect.

[0004] When using existing technologies, fixed thermal parameter models show a significant increase in prediction bias under meteorological change scenarios. When extreme gusts occur, the model prediction results show a serious inertial lag. The instantaneous temperature threshold method is difficult to fully characterize the cumulative thermal risk. Furthermore, fixed warning boundaries are either oversensitive or sluggish in all-weather scenarios and cannot adaptively adjust according to meteorological changes, ultimately reducing the accuracy, response speed, and environmental adaptability of the warning system. Summary of the Invention

[0005] The purpose of this invention is to provide a transformer thermal balance overload risk early warning system for extreme weather coupled scenarios, and to solve the following technical problems: How to improve the detection accuracy of transformer thermal balance overload risk early warning system.

[0006] The objective of this invention can be achieved through the following technical solutions: A transformer thermal balance overload risk early warning system for extreme weather coupled scenarios, the system comprising: The micro-meteorological dynamic acquisition module is used to acquire wind speed, ambient temperature, relative humidity and solar radiation illuminance of the micro-environment in which the transformer is located in real time, and to identify and remove outliers using median filtering. The load dynamic monitoring module is used to collect the current and voltage on the high-voltage side of the transformer in real time and calculate the real-time load factor. The fiber optic distributed temperature measurement module is used to acquire the measured values ​​of the top oil temperature and the hot spot temperature of the winding in real time. The dynamic thermal parameter correction module is used to dynamically correct the equivalent heat dissipation thermal resistance and thermal time constant of the transformer based on the coupled meteorological data obtained by the micro-meteorological dynamic acquisition module, and generate a dynamic thermal parameter set. The thermal equilibrium state analysis module is used to construct and solve the transient thermal equilibrium differential equation under extreme weather coupled scenario based on the dynamic thermal parameter set and the load coefficient, so as to obtain the predicted top oil temperature at the next moment. The overload risk step warning module is used to introduce the cumulative damage effect of temperature and time. Based on the predicted top oil temperature and the measured hot spot temperature, it calculates the real-time cumulative overload risk integral value and issues different levels of warning signals according to the step interval where the cumulative risk integral value is located. The dynamic threshold self-tuning module is used to connect the overload risk step early warning module and the dynamic thermal parameter correction module, and is configured to adjust the level judgment boundary of the overload risk cumulative integral value in real time according to the instantaneous value of the extreme weather coupling coefficient output by the dynamic thermal parameter correction module. The oil temperature residual feedforward compensation module is used to connect the fiber optic distributed temperature measurement module and the thermal balance state analysis module. It is configured to calculate the dynamic residual between the predicted top oil temperature and the measured top oil temperature in real time, and to start residual inertial coupling compensation when a sudden wind speed event is detected. The weighted differential form of the dynamic residual is superimposed on the heat source term on the right side of the transient thermal balance differential equation to form the compensated equivalent heat source.

[0007] Furthermore, the transient thermal equilibrium differential equation constructed by the thermal equilibrium state analysis module is as follows: ; in, The equivalent heat capacity of the transformer, Let t be the top oil temperature. The heat source is calculated based on the total load loss at time t. Let be the ambient temperature at time t. The equivalent thermal resistance at time t is the result of correction by the dynamic thermal parameter correction module. For time differentials, Differential element of top oil temperature Furthermore, the correction process of the equivalent heat dissipation thermal resistance by the dynamic thermal parameter correction module includes: Through formula Calculate the extreme weather coupling coefficient at time t; And it is corrected based on the following coupling correction formula: ; in, Let be the real-time wind speed at time t. The wind speed convection enhancement factor, Let be the relative humidity at time t. Humidity is a damaging factor. Let be the solar irradiance at time t. For reference irradiance, Add a temperature rise factor to radiation. The reference thermal resistance under standard reference conditions. The environmental temperature influence coefficient. For reference ambient temperature.

[0008] Furthermore, the process by which the dynamic thermal parameter correction module corrects the thermal time constant includes: Through formula The thermal time constant at time t, corrected by the dynamic thermal parameter correction module, is calculated. ; in, The initial thermal time constant, Based on the reference heat capacity, The measured value of the top layer oil temperature obtained by the fiber optic distributed temperature measurement module. Inertial correction function based on oil temperature change rate Furthermore, the process of the oil temperature residual feedforward compensation module performing the residual inertial coupling compensation includes: Through the formula: The equivalent heat source after compensation at time t is calculated. ; in, The preset residual coupling gain coefficient, , where is the dynamic residual, and is the compensated equivalent heat source. During the wind speed abrupt change event, the original heat source term is replaced. Substituting into the right side of the transient thermal equilibrium differential equation, the thermal equilibrium state analysis module predicts the top oil temperature at the next moment based on the compensated heat source driven by the measured residual. For a symbolic function, it is defined as when The value is 1 at time. The value is -1. The value is 0 at this time. This is the preset saturation limiting coefficient.

[0009] Furthermore, the process by which the overload risk tiered early warning module calculates the real-time cumulative integral value of overload risk includes: Through formula Calculate the cumulative integral value of overload risk at time t. ; And establish rules for determining the level of early warning signals: when At that time, a warning at the attention level will be issued; when At that time, a severe warning was issued; when At that time, a crisis-level warning was issued; in, To provide a sliding evaluation time window, Let be the integral variable, representing the sliding time window. At any point in history, This is the reference value for the rated withstand temperature of the insulating material. for The winding hot spot temperature at any given moment. This is an aging rate index, with values ​​taking positive integers. For the accumulation of forgetting factors, and The level determination boundary is provided in real time by the dynamic threshold self-tuning module.

[0010] Furthermore, the process by which the dynamic threshold self-tuning module adjusts the level determination boundary of the overload risk cumulative integral value in real time includes: For the level determination boundary and The real-time adjustment satisfies the following inverse function cooperative equation: ; ; in, The preset collaborative sensitivity coefficient, and The baseline boundary value under standard meteorological conditions. This is the extreme weather coupling coefficient at time t, output by the dynamic thermal parameter correction module.

[0011] Furthermore, the micro-meteorological dynamic acquisition module also includes a wind speed change detection unit, the working process of which includes: Through formula Calculate the rate of change of wind speed at time t ; And when the absolute value of the rate of change of the wind speed Exceeding the preset mutation threshold When an extreme gust event is detected, the dynamic thermal parameter correction module is immediately triggered to perform abrupt enhancement correction on the equivalent heat dissipation thermal resistance, and the oil temperature residual feedforward compensation module is simultaneously triggered to start the residual inertial coupling compensation. The enhanced equivalent thermal resistance is smoothed for inertia using the following formula: ; in, This is the preset damping coefficient for sudden wind speed changes. This is the equivalent thermal resistance for heat dissipation after a sudden increase following an extreme gust event. This is the preset mutation threshold.

[0012] The beneficial effects of this invention are: (1) This invention quantifies the three factors of wind speed, humidity and radiation by constructing an extreme weather coupling coefficient, and then performs segmented dynamic correction on the equivalent heat dissipation thermal resistance and thermal time constant accordingly, so that the thermal circuit model has the ability to follow weather changes, thereby improving the prediction accuracy of top oil temperature in complex weather scenarios. This avoids the prediction failure of the fixed thermal parameter model when the weather changes suddenly.

[0013] (2) This invention triggers a dual-channel coordinated response of thermal resistance enhancement correction and residual feedforward compensation by real-time monitoring of wind speed change rate and determination of sudden events. It uses residual differential information to form predictive correction and constrains the negative feedback direction with a sign function, thereby improving the prediction lag problem of the existing system in extreme gust scenarios and thus improving the system detection response speed.

[0014] (3) By introducing linear rectification cutoff into the overload risk accumulation integral, the integral kernel function is always zero in the safe temperature range, avoiding the mathematical risk of negative power operation in the traditional integral formula, so that the risk accumulation is focused only on the over-temperature range that has a real damage effect. At the same time, the introduction of the forgetting factor makes the risk assessment take into account the dual weight of historical accumulation and current state.

[0015] (4) By linking the early warning judgment boundary with the extreme weather coupling coefficient, the invention enables the early warning sensitivity to be automatically adjusted according to the weather scenario, thereby realizing intelligent management that increases vigilance under severe weather conditions and reduces the false alarm rate under favorable weather conditions. Combined with the progressive three-level early warning system, it enables maintenance personnel to take differentiated countermeasures according to the risk level, thereby improving the all-weather adaptability of the early warning system. Attached Figure Description

[0016] The invention will now be further described with reference to the accompanying drawings.

[0017] Figure 1 This is a schematic diagram of the transformer thermal balance overload risk early warning system for extreme weather coupling scenarios in this invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Please see Figure 1 As shown, in one embodiment, this application provides a transformer thermal balance overload risk early warning system for extreme weather coupled scenarios, the system comprising: The micro-meteorological dynamic acquisition module is used to acquire wind speed, ambient temperature, relative humidity and solar radiation illuminance of the micro-environment in which the transformer is located in real time, and to identify and remove outliers using median filtering. The load dynamic monitoring module is used to collect the current and voltage on the high-voltage side of the transformer in real time and calculate the real-time load factor. The fiber optic distributed temperature measurement module is used to acquire the measured values ​​of the top oil temperature and the hot spot temperature of the winding in real time. The dynamic thermal parameter correction module is used to dynamically correct the equivalent heat dissipation thermal resistance and thermal time constant of the transformer based on the coupled meteorological data obtained by the micro-meteorological dynamic acquisition module, and generate a dynamic thermal parameter set. The thermal equilibrium state analysis module is used to construct and solve the transient thermal equilibrium differential equation under extreme weather coupled scenario based on the dynamic thermal parameter set and the load coefficient, so as to obtain the predicted top oil temperature at the next moment. The overload risk step warning module is used to introduce the cumulative damage effect of temperature and time. Based on the predicted top oil temperature and the measured hot spot temperature, it calculates the real-time cumulative overload risk integral value and issues different levels of warning signals according to the step interval where the cumulative risk integral value is located. The dynamic threshold self-tuning module is used to connect the overload risk step early warning module and the dynamic thermal parameter correction module, and is configured to adjust the level judgment boundary of the overload risk cumulative integral value in real time according to the instantaneous value of the extreme weather coupling coefficient output by the dynamic thermal parameter correction module. The oil temperature residual feedforward compensation module is used to connect the fiber optic distributed temperature measurement module and the thermal balance state analysis module. It is configured to calculate the dynamic residual between the predicted top oil temperature and the measured value of the top oil temperature in real time, and to start residual inertial coupling compensation when a sudden wind speed event is detected. The weighted differential form of the dynamic residual is superimposed on the heat source term on the right side of the transient thermal balance differential equation to form the compensated equivalent heat source. Through the above technical solution, this embodiment provides a transformer thermal balance overload risk early warning system for extreme weather coupled scenarios. When the system is working, it first consists of a data perception layer composed of a micro-meteorological dynamic acquisition module, a load dynamic monitoring module, and a fiber optic distributed temperature measurement module. These modules acquire meteorological parameters, electrical load parameters, and transformer body temperature parameters of the microenvironment in which the transformer is located. The micro-meteorological dynamic acquisition module uses median filtering to identify and eliminate abnormal outliers, ensuring that the input data used for subsequent calculations has good quality. On this basis, the dynamic thermal parameter correction module uses the coupled meteorological data output by the micro-meteorological dynamic acquisition module as input to perform real-time dynamic correction of the transformer's equivalent heat dissipation thermal resistance and thermal time constant, generating a dynamic thermal parameter set under the current meteorological conditions. This correction process breaks the limitation of fixed thermal parameters in traditional models, enabling the thermal circuit model to adaptively adjust to changes in meteorological conditions. The thermal equilibrium state analysis module then receives the dynamic thermal parameter set and load coefficient, constructs the transient thermal equilibrium differential equation under the extreme weather coupling scenario, and solves it to obtain the predicted top oil temperature at the next moment. The overload risk step warning module, based on the predicted top oil temperature and the measured hot spot temperature, introduces the temperature and time cumulative damage effect, calculates the real-time overload risk cumulative integral value, and issues warning signals of attention level, severe level, or crisis level according to the step interval in which the integral value is located. At the same time as the warning assessment, the dynamic threshold self-tuning module adjusts the level judgment boundary of the risk cumulative integral value in real time according to the instantaneous value of the extreme weather coupling coefficient, so that the warning sensitivity adapts to the weather scenario.

[0020] Finally, the oil temperature residual feedforward compensation module calculates the dynamic residual between the predicted top oil temperature and the measured top oil temperature in real time, and starts residual inertial coupling compensation when a sudden wind speed event is detected. The weighted differential form of the residual is superimposed on the heat source term on the right side of the transient thermal balance differential equation to form the compensated equivalent heat source, thereby correcting the model prediction bias under extreme gust scenarios. By constructing a five-in-one early warning closed-loop architecture integrating perception, correction, analysis, evaluation, and compensation, the linkage between dynamic micro-meteorological data acquisition and dynamic thermal parameter correction enables the thermal circuit model to follow meteorological changes, improving the accuracy of top-level oil temperature prediction. The wind speed change response mechanism of the oil temperature residual feedforward compensation module improves the prediction lag problem of the existing system in extreme gust scenarios. The synergy between overload risk cumulative integration and dynamic threshold self-tuning realizes the upgrade from single-point temperature comparison to cumulative damage assessment. Moreover, the early warning sensitivity is adaptively adjusted according to the meteorological scenario, avoiding the false alarm and missed alarm problems of fixed thresholds in all-weather scenarios, thereby improving the environmental adaptability and practicality of the early warning system.

[0021] The transient thermal equilibrium differential equation constructed by the thermal equilibrium state analysis module is as follows: ; in, The equivalent heat capacity of the transformer, Let t be the top oil temperature. The heat source is calculated based on the total load loss at time t. Let be the ambient temperature at time t. The equivalent thermal resistance at time t is the result of correction by the dynamic thermal parameter correction module. For time differentials, The differential element of the top oil temperature, The first derivative of the top oil temperature with respect to time represents the instantaneous rate of change of the top oil temperature with time, i.e., the amount of change in oil temperature per unit time. Through the above technical solution, this embodiment provides the transient thermal balance differential equation constructed by the thermal balance state analysis module. The basic physical principle of transformer thermal balance state analysis is the law of conservation of energy: during the operation of a transformer, part of the heat generated by load losses is used to increase the transformer's own temperature (i.e., the heat capacity storage term), and the other part is dissipated to the surrounding environment through the heat dissipation medium (i.e., the heat dissipation term). The transient thermal balance differential equation is a mathematical characterization of this energy conservation relationship. The transient thermal balance differential equation is as follows: ; The first term on the right side of the equation The heat source item represents the heat input converted from load loss per unit time; it is the second item on the right. This is the heat dissipation item, representing the heat transferred from inside the transformer to the surrounding environment per unit time through the heat dissipation path. The temperature difference between the oil and the ambient temperature is the driving force for heat dissipation. The equivalent thermal resistance of the heat dissipation path is shown on the left side. This is the thermal capacity energy storage item, representing the net increase in the thermal energy of the transformer body per unit time; Based on this, when the heat source item When the heat dissipation term is greater than the heat source term, the left-hand term is positive, and the top oil temperature rises; conversely, when the heat source term is negative... When the value is less than the heat dissipation term, the left-hand term is negative, and the top oil temperature decreases. When the two are equal, the system is in thermal equilibrium, and the oil temperature remains stable. It is time-varying; the equation is a first-order ordinary differential equation with variable coefficients, and its temperature response characteristics depend on changes in meteorological conditions. By introducing time-varying equivalent thermal resistance, the meteorological coupling effect is incorporated into the model through the time-varying characteristics of the time-varying equivalent thermal resistance. This allows the thermal balance analysis process to reflect the dynamic influence of meteorological parameters such as wind speed, humidity, and radiation. On this basis, the variable coefficient design endows the model with the ability to respond to extreme weather scenarios, providing a highly accurate oil temperature prediction basis for subsequent overload risk assessment. This improves the prediction lag problem of traditional models in meteorological change scenarios and enhances the accuracy of the early warning system. It should be noted that the transient thermal equilibrium differential equation is solved by discretization using the first-order explicit Euler method, with a discretization time step of [missing information]. The value ranges from 1 second to 10 seconds. The discretized difference equation is: ; in, This refers to the predicted top-layer oil temperature for the next moment, within each calculation cycle. The load factor is calculated by combining the latest load factor provided by the load dynamic monitoring module with the transformer's factory loss parameters. and The data is updated in real time by the micro-meteorological dynamic acquisition module and the dynamic thermal parameter correction module, respectively.

[0022] The process by which the dynamic thermal parameter correction module corrects the equivalent heat dissipation thermal resistance includes: Through formula Calculate the extreme weather coupling coefficient at time t; And it is corrected based on the following coupling correction formula: ; in, Let be the real-time wind speed at time t. The wind speed convection enhancement factor, Let be the relative humidity at time t. Humidity is a damaging factor. Let be the solar irradiance at time t. For reference irradiance, Add a temperature rise factor to radiation. The reference thermal resistance under standard reference conditions. The environmental temperature influence coefficient. For reference ambient temperature; Through the above technical solution, this embodiment provides a correction process for the equivalent heat dissipation thermal resistance. The core concept of this embodiment is to establish an extreme weather coupling coefficient. The combined effects of three major meteorological factors—wind speed, relative humidity, and solar radiation intensity—are condensed into a dimensionless scalar, which is then used to correct the baseline thermal resistance to obtain the dynamic thermal resistance value under the current meteorological conditions. The formula for calculating the extreme weather coupling coefficient is as follows: ; The formula is constructed as a product of three factors, which respectively characterize the direction and relative weight of the influence of the three meteorological factors on heat dissipation conditions: First item As wind speed increases, the convective heat transfer coefficient on the transformer surface increases, heat dissipation capacity is enhanced, and equivalent thermal resistance decreases. Therefore, λ decreases exponentially with increasing wind speed, following an exponential function. It conforms to the nonlinear growth relationship between convective heat transfer coefficient and wind speed, and its decay rate is determined by... control.

[0023] Second item Humidity is a significant factor; as relative humidity increases, the water vapor content in the air increases, slightly increasing the air's thermal conductivity. Simultaneously, high humidity can exacerbate moisture absorption on insulating surfaces, reducing heat dissipation efficiency. Therefore, thermal resistance increases with increasing humidity. The value increases accordingly, and the square term is used to reflect the physical characteristic that humidity increases non-linearly under high temperature conditions.

[0024] Third item As the radiation impact factor increases, when solar radiation intensity increases, the transformer casing absorbs more radiant heat energy, which is equivalent to increasing the external heat load and thus increasing the equivalent thermal resistance. As irradiance increases linearly, This refers to the irradiance under standard reference conditions (usually taken as 1000W per square meter, which is the standard solar constant). Adding a temperature rise factor to radiation Based on this, obtain Then, the corrected formula for the equivalent heat dissipation thermal resistance is: ; The corrected formula is based on the reference thermal resistance. Based on this, a correction factor is multiplied by the ambient temperature deviation and the meteorological coupling coefficient. When the ambient temperature is higher than the reference temperature, the temperature difference between the air and the transformer decreases, the heat dissipation driving force weakens, and the equivalent thermal resistance increases; therefore, the correction term is positive. Conversely, when the ambient temperature is lower than the reference temperature, the correction term is negative. As a weighting coefficient for the ambient temperature term, the amplification effect of ambient temperature on thermal resistance is more significant under severe weather conditions (such as high temperature and humidity, strong radiation). By configuring the parameters in this way, the combined effects of wind speed, relative humidity, and solar irradiance on transformer heat dissipation are condensed into a dimensionless scalar. The wind speed influence factor accurately reflects the nonlinear relationship between wind speed and the convective heat transfer coefficient, with a solid physical foundation. The humidity influence factor characterizes the nonlinear aggravation effect of insulation moisture absorption and heat dissipation deterioration under high humidity conditions, closely reflecting engineering realities. The radiation influence factor has a simple form, facilitating calibration and calculation. Based on this, the following parameters are used... By applying a weighted correction to the environmental temperature influence coefficient, the physical logic is realized that the more extreme the weather conditions, the more significant the thermal resistance correction becomes, thereby improving the model accuracy and computational efficiency. It should be noted that, , , The four parameters , a, and b can be obtained in one of the following ways: (1) According to the thermal resistance data under different ambient temperatures in the transformer factory temperature rise test report, the least squares fitting method is used to identify the historical operating data offline; (2) In the early stage of transformer commissioning, by applying a known load step and recording the oil temperature response curve, the approximate values ​​of each parameter are back-calculated, and then the online recursive least squares method is used for continuous correction. In the absence of test data, the factory recommended values ​​can be used: Take 0.05, Take 0.05, Take 0.05, a takes And it adaptively fine-tunes based on the prediction error during operation.

[0025] The process by which the dynamic thermal parameter correction module corrects the thermal time constant includes: Through formula The thermal time constant at time t, corrected by the dynamic thermal parameter correction module, is calculated. ; in, The initial thermal time constant, Based on the reference heat capacity, The measured value of the top layer oil temperature obtained by the fiber optic distributed temperature measurement module. This is an inertial correction function based on the rate of change of oil temperature. Specifically, ,in, As a heating acceleration factor, it controls the sensitivity of the oil temperature change rate during the heating phase to inertial correction. The larger the value, the more significant the acceleration of the heating response; the typical value range is [0.01, 0.2]. As a cooling hysteresis factor, it controls the sensitivity of the oil temperature change rate during the cooling phase to inertial correction. The larger the value, the more significant the acceleration of the cooling response; the typical value range is [0.01, 0.1]. Through the above technical solution, this embodiment provides the thermal time constant at time t after correction by the dynamic thermal parameter correction module. The principle is: the thermal time constant of the transformer Thermal time constant is a key parameter characterizing thermal response speed. It is defined as the product of equivalent heat capacity and equivalent heat dissipation thermal resistance. The larger the thermal time constant, the slower the transformer responds to external thermal disturbances (such as sudden load increases or sudden weather changes), and the smoother the temperature change. The smaller the thermal time constant, the faster the response.

[0026] This claim, based on the traditional product relationship between equivalent heat capacity and equivalent thermal resistance, further introduces an inertial correction function based on the measured rate of change of oil temperature. The core physical basis is that the equivalent heat capacity of a transformer is not a constant value, but varies with the direction of temperature change (heating or cooling). During the heating stage, when the oil temperature rises rapidly, the effective value of thermal inertia decreases (response speeds up); during the cooling stage, when the oil temperature drops, thermal inertia increases (response lags). Therefore, it is necessary to dynamically adjust the value of the thermal time constant according to the current temperature change trend. Based on this, the piecewise formula The physical logic is as follows: When the measured oil temperature is in the heating stage ( At that time, the faster the oil temperature rises, the more molecules... The larger, The smaller the value, the shorter the thermal time constant. This corresponds to the physical phenomenon that the oil temperature accelerates in response to changes in the heat source during the heating process. For example, when the load suddenly increases, the oil layer near the heat source heats up faster, and the overall equivalent thermal time constant is slightly shortened. When the measured oil temperature is in the cooling stage ( When the oil temperature drops at a faster rate, it indicates that the current external heat dissipation conditions are significantly better than the heat dissipation state reflected by the model, meaning that the cooling rate predicted by the model lags behind the actual cooling rate. In this case, a cooling hysteresis factor can be introduced. make As the value decreases, the thermal time constant shortens accordingly, and the model responds faster to the cooling process, thus allowing the predicted oil temperature to converge to the measured value more quickly. The physical essence of this correction is that during the rapid cooling process, the effective heat capacity of the transformer exhibits an apparent decrease, that is, the oil layer near the heat dissipation surface cools down faster than the overall average temperature, which shortens the overall equivalent thermal time constant in the early stage of cooling. This physical phenomenon has been verified in the temperature rise test of oil-immersed transformers. By introducing a piecewise inertial correction function based on the measured rate of change of oil temperature on the basis of the traditional constant thermal time constant model, the adaptive adjustment of the thermal time constant with the dynamic trend of temperature is realized. Through the piecewise function design, the model is matched with different thermal inertial characteristics in the heating and cooling stages, and the predicted response is more consistent with the actual thermal dynamics. Furthermore, the thermal time constant correction is superimposed on the thermal resistance correction, forming a more complete two-parameter dynamic thermal parameter set, which provides more reliable coefficient support for the accurate solution of the transient thermal balance equation, thereby improving the transient response accuracy of oil temperature prediction under extreme weather coupling scenarios.

[0027] The process of residual inertial coupling compensation performed by the oil temperature residual feedforward compensation module includes: Through the formula: The equivalent heat source after compensation at time t is calculated. ; in, The preset residual coupling gain coefficient is used to globally scale the magnitude of the compensation term, making the dimensions of the compensation term consistent with... To maintain consistency with the power dimension (W), the specific value was obtained by offline identification of the transformer's historical operating data. , where is the dynamic residual, and is the compensated equivalent heat source. During the wind speed abrupt change event, the original heat source term is replaced. Substituting into the right side of the transient thermal equilibrium differential equation, the thermal equilibrium state analysis module predicts the top oil temperature at the next moment based on the compensated heat source driven by the measured residual. For a symbolic function, it is defined as when The value is 1 at time. The value is -1. The value is 0 at this time. The preset saturation limiting coefficient is used to nonlinearly attenuate extreme residuals and prevent the compensation term from increasing indefinitely and causing the system to diverge when the sensor is abnormal or the model is severely mismatched. Through the above technical solution, this embodiment provides the process of residual inertial coupling compensation performed by the oil temperature residual feedforward compensation module. In actual operation, although the transient thermal balance differential equation has been solved... and The dynamic correction enhances the response capability to meteorological coupling. However, due to the complexity of the transformer's internal temperature field, sensor measurement noise, model simplification errors, and drastic changes in heat dissipation conditions during extreme gusts, a dynamic residual inevitably exists between the predicted and measured top-layer oil temperature. If this residual is not utilized, the model prediction error will accumulate over time, reducing the reliability of the early warning system. The core principle of the residual inertial coupling compensation mechanism proposed in this embodiment is: to calculate the residual between the predicted value and the measured value. Differential with respect to time (Reflecting the trend of residual change) As a feedforward correction signal, after gain amplification and time constant normalization, it is superimposed on the right-hand heat source term of the transient thermal equilibrium differential equation in a sign-directed manner. In the middle, a compensated equivalent heat source is formed. ; The physical logic of this compensation mechanism lies in: This reflects whether the model's prediction bias is widening or converging. When the measured temperature is higher than the predicted temperature and the residuals are widening, it indicates that the current model underestimates the actual heat generation or overestimates the heat dissipation, requiring positive compensation of the heat source. When the measured temperature is lower than the predicted temperature and the residuals are widening negatively, negative compensation of the heat source is required. (Sign function) Ensure that the compensation direction always points towards the zeroing direction of the residual, thus forming a negative feedback mechanism.

[0028] By using the differential information of the residuals instead of the residuals themselves as the basis for compensation, the compensation effect becomes predictive. That is, it can not only respond to the deviations that have already occurred, but also correct the tendency of the deviations to expand in advance. The introduction of the sign function ensures that the compensation direction is always negative feedback, which guarantees the stability of the system in principle. The saturation limiting term makes the compensator robust against abnormal interference, avoiding false alarms caused by occasional sensor failures. Moreover, the compensation is only activated when a sudden wind speed event is triggered, and does not interfere with the main model under normal conditions, which ensures the determinism of the system's predictive behavior under stable weather conditions. At the same time, the computational resources are concentrated on the extreme scenarios that require the most correction, and the design trade-offs are reasonable.

[0029] The process by which the overload risk tiered early warning module calculates the real-time cumulative integral value of overload risk includes: Through formula Calculate the cumulative integral value of overload risk at time t. ; And establish rules for determining the level of early warning signals: when At that time, a warning at the attention level will be issued; when At that time, a severe warning was issued; when At that time, a crisis-level warning was issued; in, To provide a sliding evaluation time window, Let be the integral variable, representing the sliding time window. At any point in history, This is the reference value for the rated withstand temperature of the insulating material. for The winding hot spot temperature at any given moment. This is an aging rate index, with values ​​taking positive integers. For the accumulation of forgetting factors, and The boundary for determining the level is provided in real time by the dynamic threshold self-tuning module; Through the above technical solution, this embodiment provides a process for calculating the real-time cumulative integral value of overload risk, which can be achieved through the formula. The calculation yields this formula, obtained through integration over a sliding time window. The system conducts a cumulative assessment of historical overheating events, including: Operator For linear rectification cutoff, only when the winding hot spot temperature exceeds the rated withstand temperature. Only when the temperature exceeds the limit is the excess amount included in the risk accumulation; when When less than When the integrand is zero, it does not contribute a risk value at that moment. This ensures that normal temperature fluctuations within the safe operating range will not generate false alarms, avoiding the defect of undefined negative exponentiation in the traditional integral formula. Then divide the over-temperature difference by To achieve normalization, It has dimensionless physical meaning, which facilitates comparison between transformers with different rated temperature levels; And the exponent term Forgetting factor, where As the cumulative forgetting rate, the weight of historical overheating events that are further away from the current time decays exponentially, reflecting the engineering intuition that "recent overheating has a greater impact on current risk than long-term overheating". Based on this, the calculation is obtained Then, warning signals can be output according to preset rules. Through this setting, an overload risk accumulation assessment method that combines physical rationality and engineering practicality is constructed. Linear rectification cutoff ensures that the integral kernel function is always zero within the safe temperature range, completely avoiding the mathematical risk of negative power operation, so that risk accumulation is focused only on the over-temperature range that has a real damage effect. The introduction of the forgetting factor makes the risk assessment take into account both the historical accumulation effect and the current state weight, which is consistent with the physical reality that insulation aging has both accumulation and recoverability. The tiered warning level combined with dynamic threshold constitutes a progressive alarm system, enabling maintenance personnel to take differentiated response measures according to the warning level. It should be noted that the sliding time window With forgetting factor Together, they determine the effective memory duration for risk accumulation. They must satisfy a synergistic relationship to eliminate parameter redundancy. It is recommended... and The product satisfies Within this range, the window cutoff effect and the exponential forgetting effect are balanced, ensuring sufficient sensitivity to recent overheating events while retaining appropriate historical information. The parameter tuning order is as follows: first, determine the parameters based on the transformer thermal time constant. (It is recommended to use a heat time constant of 2 to 5 times), and then according to Constraint determination .

[0030] The process by which the dynamic threshold self-tuning module adjusts the level determination boundary of the accumulated overload risk value in real time includes: For the level determination boundary and The real-time adjustment satisfies the following inverse function cooperative equation: ; ; in, The preset collaborative sensitivity coefficient, and The baseline boundary value under standard meteorological conditions. The extreme weather coupling coefficient at time t is output by the dynamic thermal parameter correction module. Through the above technical solution, this embodiment provides a process for adjusting the level determination boundary of the overload risk cumulative integral value. Under different meteorological conditions, the actual degree of danger corresponding to the same overload risk cumulative integral value is different. Under severe meteorological conditions (such as high temperature, high humidity, strong radiation, or no wind), the heat dissipation conditions of the transformer deteriorate, and the smaller A high value may indicate a higher actual risk; however, under favorable weather conditions (such as strong winds, low temperatures, and low humidity), the transformer's heat dissipation conditions improve, resulting in a greater risk. The value may still be within a safe range. Therefore, a fixed warning boundary is either too conservative (resulting in frequent false alarms under severe weather conditions) or too aggressive (resulting in missed alarms under favorable weather conditions) in all-weather scenarios. Therefore, by establishing the coupling coefficient between the early warning boundary and extreme weather... To solve the above problem, a quantitative relationship between them is needed, because This comprehensively reflects the coupled effects of wind speed, humidity, and radiation on heat dissipation conditions, namely... A higher value indicates more severe weather conditions. A smaller value indicates more favorable weather conditions; therefore, the warning boundary should be adjusted accordingly. The increase in the magnitude of ... By setting it up in this way, a dynamic self-tuning mechanism for the early warning boundary based on the extreme weather coupling coefficient is established, breaking the limitation of the fixed judgment boundary of the traditional early warning system. This enables the early warning sensitivity to be automatically adjusted according to changes in the weather scene, and realizes intelligent management that increases vigilance under severe weather conditions and reduces the false alarm rate under favorable weather conditions.

[0031] The micro-meteorological dynamic acquisition module also includes a wind speed change detection unit, the working process of which includes: Through formula Calculate the rate of change of wind speed at time t ; And when the absolute value of the rate of change of the wind speed Exceeding the preset mutation threshold When an extreme gust event is detected, the dynamic thermal parameter correction module is immediately triggered to perform abrupt enhancement correction on the equivalent heat dissipation thermal resistance, and the oil temperature residual feedforward compensation module is simultaneously triggered to start the residual inertial coupling compensation. The enhanced equivalent thermal resistance is smoothed for inertia using the following formula: ; in, The preset damping coefficient for sudden wind speed changes (dimension: Controlling the rate of change of wind speed enhances the correction strength of the modified thermal resistance. This is the equivalent thermal resistance for heat dissipation after a sudden increase following an extreme gust event. The preset mutation threshold is used to determine whether an extreme gust event has occurred; the typical value range is [value range missing]. The specific value is determined based on the historical wind speed data of the transformer installation site and the range setting of the wind speed sensor. Through the above technical solution, this embodiment provides the working process of the wind speed change detection unit. Unlike slowly changing meteorological conditions, wind speed changes caused by extreme gust events (such as thunderstorms, the outer circulation of typhoons, and localized severe convection) usually occur within a few seconds to tens of seconds, and the wind speed change rate... This is far greater than normal weather fluctuations. Such a drastic weather disturbance will cause a rapid and dramatic change in the convective heat transfer coefficient between the transformer surface and the air, resulting in a significant increase in the equivalent thermal resistance. It should respond quickly physically; however, due to Wind speed item Based on instantaneous wind speed calculation, when the wind speed changes abruptly, the thermal resistance correction can follow instantly, which is the inherent advantage of this formula. However, at the same time, due to the existence of thermal inertia, the oil temperature model based on this thermal resistance prediction has a first-order lag in its prediction output, which cannot immediately reflect the full impact of thermal resistance change on oil temperature. Based on this, this embodiment monitors the rate of change of wind speed in real time. ,when Exceeding the preset mutation threshold The event was determined to be an extreme gust of wind. A trigger signal was simultaneously sent to the dynamic thermal parameter correction module and the oil temperature residual feedforward compensation module, and based on the formula... Immediately after a gust event is triggered, the equivalent thermal resistance is abruptly enhanced and corrected. This setup fills the technical gap in the response of conventional meteorological coupled models to instantaneous wind speed changes. By monitoring the wind speed change rate in real time, the system can identify extreme gust events with a response speed of seconds, which has a fundamental time advantage compared to relying on weather forecasts. The re-initialized thermal resistance introduces damping smoothing of the wind speed change rate on the basis of conventional correction, which makes the thermal resistance model further adjust towards heat dissipation enhancement under the impact of gusts. Furthermore, the dual-channel triggering of thermal resistance re-initialization and residual compensation achieves a coordinated response to wind speed change scenarios, thereby enhancing the robustness of the early warning system under the transient impact of extreme weather. It should be noted that during the wind speed abrupt change event, the thermal resistance enhancement correction ( ) and residual feedforward compensation ( Simultaneously acting on different physical components of the transient thermal equilibrium equation, when their directions of action are opposite, the system follows the direction of residual compensation. This is because residual compensation directly reflects the deviation direction between the measured and predicted temperatures, has a measured basis, and is always negative feedback due to the constraint of the sign function, thus belonging to the closed-loop correction category. Thermal resistance enhancement correction, on the other hand, is an open-loop feedforward correction based on the rate of change of wind speed, belonging to the auxiliary correction category. Therefore, when both act simultaneously, the overall effect is equivalent to a synergistic relationship where closed-loop compensation is dominant and feedforward thermal resistance correction is auxiliary, preventing system divergence. To enhance synergistic stability, further calculations can be performed... At that time, with Alternative Substituting into the compensation formula The calculations ensure that both corrections share the same thermal resistance basis, and this scheme is mainly designed to address scenarios with sudden increases in wind speed, while residual compensation dominates the correction when wind speed drops sharply.

[0032] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A transformer thermal balance overload risk early warning system for extreme weather coupled scenarios, characterized in that, The system includes: The micro-meteorological dynamic acquisition module is used to acquire wind speed, ambient temperature, relative humidity and solar radiation illuminance of the micro-environment in which the transformer is located in real time, and to identify and remove outliers using median filtering. The load dynamic monitoring module is used to collect the current and voltage on the high-voltage side of the transformer in real time and calculate the real-time load factor. The fiber optic distributed temperature measurement module is used to acquire the measured values ​​of the top oil temperature and the hot spot temperature of the winding in real time. The dynamic thermal parameter correction module is used to dynamically correct the equivalent heat dissipation thermal resistance and thermal time constant of the transformer based on the coupled meteorological data obtained by the micro-meteorological dynamic acquisition module, and generate a dynamic thermal parameter set. The thermal equilibrium state analysis module is used to construct and solve the transient thermal equilibrium differential equation under extreme weather coupled scenario based on the dynamic thermal parameter set and the load coefficient, so as to obtain the predicted top oil temperature at the next moment. The overload risk step warning module is used to introduce the cumulative damage effect of temperature and time. Based on the predicted top oil temperature and the measured hot spot temperature, it calculates the real-time cumulative overload risk integral value and issues different levels of warning signals according to the step interval where the cumulative risk integral value is located. The dynamic threshold self-tuning module is used to connect the overload risk step early warning module and the dynamic thermal parameter correction module, and is configured to adjust the level judgment boundary of the overload risk cumulative integral value in real time according to the instantaneous value of the extreme weather coupling coefficient output by the dynamic thermal parameter correction module. The oil temperature residual feedforward compensation module is used to connect the fiber optic distributed temperature measurement module and the thermal balance state analysis module. It is configured to calculate the dynamic residual between the predicted top oil temperature and the measured top oil temperature in real time, and to start residual inertial coupling compensation when a sudden wind speed event is detected. The weighted differential form of the dynamic residual is superimposed on the heat source term on the right side of the transient thermal balance differential equation to form the compensated equivalent heat source.

2. The transformer thermal balance overload risk early warning system for extreme weather coupled scenarios according to claim 1, characterized in that, The transient thermal equilibrium differential equation constructed by the thermal equilibrium state analysis module is as follows: ； in, The equivalent heat capacity of the transformer, Let t be the top oil temperature. The heat source is calculated based on the total load loss at time t. Let be the ambient temperature at time t. The equivalent thermal resistance at time t is the result of correction by the dynamic thermal parameter correction module. For time differentials, This is the differential element of the top oil temperature.

3. The transformer thermal balance overload risk early warning system for extreme weather coupled scenarios according to claim 1, characterized in that, The process by which the dynamic thermal parameter correction module corrects the equivalent heat dissipation thermal resistance includes: Through formula Calculate the extreme weather coupling coefficient at time t; And it is corrected based on the following coupling correction formula: ； in, Let be the real-time wind speed at time t. The wind speed convection enhancement factor, Let be the relative humidity at time t. Humidity is a damaging factor. Let be the solar irradiance at time t. For reference irradiance, Add a temperature rise factor to radiation. The reference thermal resistance under standard reference conditions. The environmental temperature influence coefficient. For reference to ambient temperature, The equivalent thermal resistance at time t is the result of correction by the dynamic thermal parameter correction module. Let t be the ambient temperature at time t.

4. The transformer thermal balance overload risk early warning system for extreme weather coupled scenarios according to claim 1, characterized in that, The process by which the dynamic thermal parameter correction module corrects the thermal time constant includes: Through formula The thermal time constant at time t, corrected by the dynamic thermal parameter correction module, is calculated. ; in, The initial thermal time constant, Based on the reference heat capacity, The measured value of the top layer oil temperature obtained by the fiber optic distributed temperature measurement module. This is an inertial correction function based on the rate of change of oil temperature. , As a heating acceleration factor, As a cooling hysteresis factor, The equivalent heat capacity of the transformer, The equivalent thermal resistance at time t is the result of correction by the dynamic thermal parameter correction module. This is the reference thermal resistance under standard reference conditions.

5. The transformer thermal balance overload risk early warning system for extreme weather coupled scenarios according to claim 1, characterized in that, The process of residual inertial coupling compensation performed by the oil temperature residual feedforward compensation module includes: Through the formula: The equivalent heat source after compensation at time t is calculated. ; in, The preset residual coupling gain coefficient has the following dimensions: , , where is the dynamic residual, and is the compensated equivalent heat source. During the wind speed abrupt change event, the total load loss at time t, which replaces the original heat source term, is converted into heat. Substituting into the right side of the transient thermal equilibrium differential equation, the thermal equilibrium state analysis module predicts the top oil temperature at the next moment based on the compensated heat source driven by the measured residual. For a symbolic function, it is defined as when The value is 1 at time. The value is -1. The value is 0 at this time. The preset saturation limiting coefficient, The thermal time constant at time t is the result of correction by the dynamic thermal parameter correction module. Initial thermal time constant.

6. The transformer thermal balance overload risk early warning system for extreme weather coupled scenarios according to claim 1, characterized in that, The process by which the overload risk tiered early warning module calculates the real-time cumulative integral value of overload risk includes: Through formula Calculate the cumulative integral value of overload risk at time t. ; And establish rules for determining the level of early warning signals: when At that time, a warning at the attention level will be issued; when At that time, a severe warning was issued; when At that time, a crisis-level warning was issued; in, To provide a sliding evaluation time window, Let be the integral variable, representing the sliding time window. At any point in history, This is the reference value for the rated withstand temperature of the insulating material. for The winding hot spot temperature at any given moment. This is an aging rate index, with values ​​taking positive integers. For the accumulation of forgetting factors, and The level determination boundary is provided in real time by the dynamic threshold self-tuning module.

7. The transformer thermal balance overload risk early warning system for extreme weather coupled scenarios according to claim 1, characterized in that, The process by which the dynamic threshold self-tuning module adjusts the level determination boundary of the accumulated overload risk value in real time includes: For the level determination boundary and The real-time adjustment satisfies the following inverse function cooperative equation: ； ； in, The preset collaborative sensitivity coefficient, and The baseline boundary value under standard meteorological conditions. This is the extreme weather coupling coefficient at time t, output by the dynamic thermal parameter correction module.

8. The transformer thermal balance overload risk early warning system for extreme weather coupled scenarios according to claim 1, characterized in that, The micro-meteorological dynamic acquisition module also includes a wind speed change detection unit, the working process of which includes: Through formula Calculate the rate of change of wind speed at time t ; And when the absolute value of the rate of change of the wind speed Exceeding the preset mutation threshold When an extreme gust event is detected, the dynamic thermal parameter correction module is immediately triggered to perform abrupt enhancement correction on the equivalent heat dissipation thermal resistance, and the oil temperature residual feedforward compensation module is simultaneously triggered to start the residual inertial coupling compensation. The enhanced equivalent thermal resistance is smoothed for inertia using the following formula: ； in, This is the preset damping coefficient for sudden wind speed changes. This is the equivalent thermal resistance for heat dissipation after a sudden increase following an extreme gust event. The preset mutation threshold, The equivalent thermal resistance at time t is the thermal resistance after correction by the dynamic thermal parameter correction module.

Images