Parameter correction method for transformer thermal circuit model based on extended kalman filter

Abstract

The application discloses a transformer thermal circuit model parameter correction method based on an extended Kalman filter, belongs to the technical field of transformer temperature rise tests, and comprises the following steps: S1: defining a state equation for representing the change of dynamic parameters in a transformer thermal circuit model with time according to a state transition matrix; S2: defining a nonlinear equation for representing the nonlinear relationship between observation parameters and dynamic parameters based on the physical relationship between the observation parameters and the dynamic parameters in the transformer thermal circuit model; S3: linearizing the nonlinear equation to obtain a linear equation based on the extended Kalman filter; S4: obtaining a first predicted value of the dynamic parameters based on the state equation, obtaining a second predicted value of the observation parameters based on the linear equation, and correcting the first predicted value of the dynamic parameters based on the second predicted value and the actual value of the observation parameters to obtain a final value of the dynamic parameters. The technical problem that the temperature dependence of thermal resistance is not considered in the prior art, resulting in low transformer temperature prediction accuracy is overcome.

Description

Technical Field

[0001] This invention relates to the field of transformer temperature rise testing, specifically to a method for correcting transformer thermal circuit model parameters based on extended Kalman filtering. Background Technology

[0002] With the rapid development and increasing complexity of power systems, transformers, as key equipment in these systems, directly impact the stability and reliability of the entire power grid. Transformers generate heat during operation; excessively high temperatures or abnormal temperature fluctuations can lead to aging of insulation materials, deterioration of equipment performance, and even serious malfunctions. Therefore, accurate prediction of transformer temperature is not only crucial for ensuring the safe operation of transformers but also a key measure for improving power system stability, extending equipment lifespan, and reducing operation and maintenance costs. Existing technologies primarily predict transformer temperature using transformer thermal circuit models. For example, patent number CN106595884A describes a method for predicting transformer winding hot spot temperature under low-temperature conditions. This method first establishes an improved three-thermal-circuit model; then, it corrects the improved model based on different oil flow cooling methods; it compares the temperature rise of the transformer oil detected by a temperature sensor with the temperature value calculated using the corrected three-thermal-circuit model to obtain a more realistic transformer thermal circuit model; based on the ambient temperature, it calculates the average oil temperature of the transformer using the near-realistic thermal circuit model; then, it uses this average oil temperature as a reference temperature for the middle layer oil temperature in the improved three-thermal-circuit model to calculate the top layer oil temperature; finally, it calculates the hot spot temperature value of the transformer winding. This improves the accuracy of the transformer winding hot spot temperature value. However, the thermal resistance parameter in the above scheme's transformer thermal circuit model remains constant, neglecting the nonlinear characteristics of thermal resistance changing with temperature during the temperature rise process. This is especially problematic in oil-immersed transformers, where the transformer thermal resistance changes significantly with increasing temperature. If a constant thermal resistance parameter is still used for temperature prediction, the transformer thermal circuit model will not be able to accurately reflect the actual heat conduction process, thus introducing a large prediction error. Summary of the Invention

[0003] To address the problem of low accuracy in transformer temperature prediction caused by existing technologies failing to consider the temperature dependence of thermal resistance, this invention provides a parameter correction method for transformer thermal circuit models based on extended Kalman filtering. This method defines a nonlinear equation based on the physical relationship between observed and dynamic parameters in the transformer thermal circuit model, accurately reflecting the nonlinear characteristics of thermal resistance changing with temperature. Then, through the local linearization property of extended Kalman filtering, the nonlinear equation is linearized to obtain a linear equation. Furthermore, the predicted values ​​of the observed parameters are obtained based on the linear equation, and the predicted values ​​of the dynamic parameters are corrected by combining the actual values ​​of the observed parameters to obtain the final values ​​of the dynamic parameters. This method overcomes the technical problem of low accuracy in transformer temperature prediction caused by the failure to consider the temperature dependence of thermal resistance in existing technologies, significantly improving the accuracy of transformer temperature prediction.

[0004] To address the aforementioned technical problems, this invention provides a method for correcting transformer thermal circuit model parameters based on extended Kalman filtering, comprising the following steps:

[0005] S1: Define the state equations that characterize the changes of dynamic parameters in the transformer thermal circuit model over time based on the state transition matrix;

[0006] S2: Based on the physical relationship between observed parameters and dynamic parameters in the transformer thermal circuit model, define a nonlinear equation characterizing the nonlinear relationship between observed parameters and dynamic parameters;

[0007] S3: Linearize the nonlinear equations using extended Kalman filtering to obtain linear equations;

[0008] S4: Obtain the first predicted value of the dynamic parameter based on the state equation, obtain the second predicted value of the observed parameter based on the linear equation, and correct the first predicted value of the dynamic parameter based on the second predicted value and the actual value of the observed parameter to obtain the final value of the dynamic parameter.

[0009] By adopting the above technical solution, the present invention has the following advantages:

[0010] By defining a nonlinear equation based on the physical relationship between observed and dynamic parameters in a transformer thermal circuit model, the nonlinear characteristics of thermal resistance changing with temperature are accurately reflected. Then, by utilizing the local linearization property of the extended Kalman filter, the nonlinear equation is linearized to obtain a linear equation. This allows the predicted values ​​of observed parameters to be obtained from the predicted values ​​of dynamic parameters. The predicted and actual values ​​of observed parameters are then combined for analysis. Finally, the predicted values ​​of dynamic parameters are corrected using the linear equation and the combined analysis results to obtain the final values ​​of the dynamic parameters. The obtained linear equation bridges the gap between the predicted and actual values ​​of dynamic and observed parameters, enabling dynamic correction of the predicted values ​​of parameters such as thermal resistance by comparing the predicted and actual values, ultimately yielding accurate estimates of the dynamic parameters. This overcomes the technical problem of low transformer temperature prediction accuracy caused by the failure to consider the temperature dependence of thermal resistance in existing technologies, significantly improving the accuracy of transformer temperature prediction and the adaptability of the transformer thermal circuit model.

[0011] Furthermore, by linearizing the nonlinear equations—that is, transforming complex nonlinear problems into easily solvable linear problems—the efficiency of transformer temperature prediction is also improved.

[0012] Preferably, S1 includes:

[0013] S11: Based on the characteristics of the random distribution variables in the transformer thermal circuit model, obtain the cumulative distribution function of the random distribution variables, and then obtain the uniform distribution variables. Map the uniform distribution variables to Gaussian distribution variables through the inverse Gaussian distribution.

[0014] S12: Obtain the state equation based on the Gaussian distributed variables and the state transition matrix.

[0015] In this scheme, by mapping the random distribution variable to a Gaussian distribution variable suitable for extended Kalman filtering, the influence of external interference data on transformer temperature prediction is preserved, and it is also easy to transform it into a linear influence through extended Kalman filtering, thereby improving the efficiency and accuracy of transformer temperature prediction.

[0016] Preferably, in S12, the state equation is:

[0017] ;

[0018] In the formula, x k+1 Let x represent the value of the dynamic parameter at time k+1, A represent the state transition matrix, and x represent the value of the dynamic parameter at time k+1. k w represents the value of the dynamic parameter at time k. k Let k represent the Gaussian distribution variable at time k.

[0019] Preferably, in S2, the observed parameters include at least the winding temperature, top oil temperature, oil tank temperature, and ambient temperature;

[0020] The dynamic parameters include at least the thermal resistance between the winding and the oil, the thermal resistance between the oil and the oil tank, and the thermal resistance between the oil tank and the ring.

[0021] Preferably, in S2, the nonlinear relationship includes:

[0022] ;

[0023] ;

[0024] ;

[0025] In the formula, For winding temperature, For transformer load losses, For the winding heat capacity, The top oil temperature, For time step, For the winding to oil thermal resistance, For oil heat capacity, For no-load loss, For the fuel tank temperature, The thermal resistance between the oil and the oil tank. For the heat capacity of the fuel tank, For ambient temperature, This represents the thermal resistance between the oil tank and the ring.

[0026] Preferably, in S2, the nonlinear equation is:

[0027] ;

[0028] In the formula, z k Let v be the value of the observed parameter at time k, h represent the nonlinear relationship function between the dynamic parameter and the observed parameter, and v k This represents the observed disturbance variable at time k.

[0029] Preferably, S3 includes:

[0030] S31: Differentiate the nonlinear equation to obtain the initial linear equation, and input the historical data corresponding to the observed parameters and dynamic parameters into the initial linear equation to obtain the degree of linearization of the initial linear equation;

[0031] S32: When the linearization degree meets the preset conditions, the initial linear equation is a linear equation; otherwise, the initial linear equation is regarded as a nonlinear equation, and S31 is executed.

[0032] In this scheme, the linearization degree of the initial linear equation is verified by using historical data. If the condition is not met, the linearization degree of the initial linear equation is re-acquired until the preset condition is met. This overcomes the prediction error caused by insufficient linearization of the nonlinear equation and further improves the accuracy of transformer temperature prediction.

[0033] Preferably, in S4, the step of correcting the first predicted value of the dynamic parameter based on the second predicted value and the actual value of the observed parameter to obtain the final value of the dynamic parameter includes:

[0034] The inverse matrix to be determined is obtained by using the observation matrix corresponding to the linear equation, and the Kalman gain is obtained based on the inverse matrix to be determined.

[0035] The state estimate is obtained based on the Kalman gain, the second predicted value, and the actual value, and the state covariance matrix is ​​obtained. When the state covariance matrix meets the preset requirements, the first predicted value is the final value; otherwise, the state estimate is corrected based on the state covariance matrix to obtain the final value.

[0036] Preferably, the process of obtaining the Kalman gain based on the inverse matrix to be determined further includes:

[0037] The desingularity parameters are obtained through cross-validation, and the matrix to be inverted is updated based on the desingularity parameters.

[0038] In this scheme, the problem of high error rate in Kalman gain calculation caused by the singularity of the matrix to be inverted is overcome by updating the singular parameters. This not only improves the accuracy of the final value, but also enhances the adaptability and accuracy of the transformer thermal circuit model.

[0039] The beneficial effects of this plan are:

[0040] By defining a nonlinear equation based on the physical relationship between observed and dynamic parameters in a transformer thermal circuit model, the nonlinear characteristics of thermal resistance changing with temperature are accurately reflected. Then, by utilizing the local linearization property of the extended Kalman filter, the nonlinear equation is linearized to obtain a linear equation. This allows the predicted values ​​of observed parameters to be obtained from the predicted values ​​of dynamic parameters. The predicted and actual values ​​of observed parameters are then combined for analysis. Finally, the predicted values ​​of dynamic parameters are corrected using the linear equation and the combined analysis results to obtain the final values ​​of the dynamic parameters. The obtained linear equation bridges the gap between the predicted and actual values ​​of dynamic and observed parameters, enabling dynamic correction of the predicted values ​​of parameters such as thermal resistance by comparing the predicted and actual values, ultimately yielding accurate estimates of the dynamic parameters. This overcomes the technical problem of low transformer temperature prediction accuracy caused by the failure to consider the temperature dependence of thermal resistance in existing technologies, significantly improving the accuracy of transformer temperature prediction and the adaptability of the transformer thermal circuit model.

[0041] By mapping random distributed variables to Gaussian distributed variables suitable for extended Kalman filtering, the influence of external interference data on transformer temperature prediction is preserved, and it is also easy to transform it into a linear influence through extended Kalman filtering. Furthermore, the linearization degree of the initial linear equation is verified by historical data. When the condition is not met, the linearization degree of the initial linear equation is re-acquired until the preset condition is met, thus overcoming the prediction error caused by insufficient linearization of nonlinear equations and further improving the accuracy of transformer temperature prediction.

[0042] By updating the inverted matrix by removing singular parameters, the problem of high error rate in Kalman gain calculation caused by the singularity of the inverted matrix is ​​overcome. This not only improves the accuracy of the final value, but also significantly improves the adaptability and accuracy of the transformer thermal circuit model.

[0043] The present invention also provides a storage medium storing computer-executable instructions, which, when loaded and executed by a processor, implement the steps of the method for correcting transformer thermal circuit model parameters based on extended Kalman filtering. Attached Figure Description

[0044] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings. The drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings.

[0045] Figure 1 This is a flowchart of the transformer thermal circuit model parameter correction method based on extended Kalman filtering according to the present invention. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only one preferred embodiment of this invention and are only used to explain this invention. They do not limit the scope of protection of this invention. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0047] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe the operations (or steps) as sequential processes, many of the operations (or steps) can be performed in parallel, concurrently, or simultaneously. Furthermore, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but it may also have additional steps not included in the figures; the process may correspond to a method, function, procedure, subroutine, subroutine, etc.

[0048] Example 1:

[0049] like Figure 1 As shown, the method for correcting the parameters of a transformer thermal circuit model based on extended Kalman filtering includes the following steps:

[0050] S1: Define the state equations that characterize the changes of dynamic parameters in the transformer thermal circuit model over time based on the state transition matrix.

[0051] S1 includes:

[0052] S11: Based on the characteristics of the random distribution variables in the transformer thermal circuit model, obtain the cumulative distribution function of the random distribution variables, and then obtain the uniform distribution variables. Map the uniform distribution variables to Gaussian distribution variables through the inverse Gaussian distribution.

[0053] S12: Obtain the state equation based on the Gaussian distributed variables and the state transition matrix.

[0054] In S12, the state equation is:

[0055] ;

[0056] In the formula, x k+1 Let x represent the value of the dynamic parameter at time k+1, A represent the state transition matrix, and x represent the value of the dynamic parameter at time k+1. k w represents the value of the dynamic parameter at time k. k Let k represent the Gaussian distribution variable at time k.

[0057] In this embodiment, the random distribution variable is the process noise, which represents the uncertainty of system state changes. Obtaining the cumulative distribution function of the random distribution variable based on its characteristics in the transformer thermal circuit model specifically involves obtaining the cumulative distribution function of the random distribution variables that do not conform to a Gaussian distribution. By using an inverse Gaussian distribution, the random distribution variables that do not conform to a Gaussian distribution are mapped to Gaussian distribution variables, thus ensuring that all random distribution variables conform to a Gaussian distribution. While preserving the influence of external interference data on transformer temperature prediction, this also facilitates the transformation of this influence into a linear effect through extended Kalman filtering, thereby improving the efficiency and accuracy of transformer temperature prediction.

[0058] In this embodiment, taking a constant ambient temperature as an example, then , This represents the value of the winding-to-oil thermal resistance at time k. This represents the thermal resistance between the oil and the tank at time k. The state transition matrix describes the change of the system state between different time steps. The value of the thermal resistance at the next time step can be obtained from the state transition matrix. Since the temperature rise process generally progresses slowly and the oil temperature time constant is large, the change in the thermal resistance value within a small time interval is small. Therefore, the state transition matrix can be set as an identity matrix. It is understandable that in practical applications, the change in the thermal resistance value may be affected by factors such as time lag and oil temperature changes. Therefore, the state transition matrix and process noise in the state equation will differ in different systems.

[0059] S2: Based on the physical relationship between observed parameters and dynamic parameters in the transformer thermal circuit model, define a nonlinear equation characterizing the nonlinear relationship between observed parameters and dynamic parameters.

[0060] In S2, the observed parameters include at least the winding temperature, top oil temperature, oil tank temperature, and ambient temperature;

[0061] The dynamic parameters include at least the thermal resistance between the winding and the oil, the thermal resistance between the oil and the oil tank, and the thermal resistance between the oil tank and the ring.

[0062] In S2, the nonlinear relationship includes:

[0063] ;

[0064] ;

[0065] ;

[0066] In the formula, For winding temperature, For transformer load losses, For the winding heat capacity, The top oil temperature, For time step, For the winding to oil thermal resistance, For oil heat capacity, For no-load loss, For the fuel tank temperature, The thermal resistance between the oil and the oil tank. For the heat capacity of the fuel tank, For ambient temperature, This represents the thermal resistance between the oil tank and the ring.

[0067] In S2, the nonlinear equation is:

[0068] ;

[0069] In the formula, z k Let v be the value of the observed parameter at time k, h represent the nonlinear relationship function between the dynamic parameter and the observed parameter, and v k This represents the observed disturbance variable at time k.

[0070] In this embodiment, the physical relationship specifically includes:

[0071] ;

[0072] ;

[0073] ;

[0074] To correlate thermal resistance with actual temperature, a nonlinear equation is introduced. In the transformer thermal circuit model, the winding temperature, top oil temperature, and tank temperature, which are closely related to thermal resistance, are all affected by thermal resistance. Therefore, the nonlinear equation needs to be established based on the physical relationship between thermal resistance and temperature. In this embodiment, , This represents the winding temperature at time k. This represents the top oil temperature at time k. This represents the value of the fuel tank temperature at time k. Approaching 0, the observation interference variable is the observation noise.

[0075] S3: Linearize the nonlinear equations using extended Kalman filtering to obtain linear equations.

[0076] S3 includes:

[0077] S31: Differentiate the nonlinear equation to obtain the initial linear equation, and input the historical data corresponding to the observed parameters and dynamic parameters into the initial linear equation to obtain the degree of linearization of the initial linear equation;

[0078] S32: When the linearization degree meets the preset conditions, the initial linear equation is a linear equation; otherwise, the initial linear equation is regarded as a nonlinear equation, and S31 is executed.

[0079] In this embodiment, obtaining the initial linear equation by differentiating the nonlinear equation specifically involves: using the formula... The Jacobian matrix of the nonlinear equation is obtained by differentiating the nonlinear relationship function between the dynamic parameters and the observed parameters. The initial linear equation is then obtained through the Jacobian matrix. The Jacobian matrix represents the nonlinear equation. This represents the dependent variable in a nonlinear relationship function, specifically the observed parameters, i.e., various temperatures. The independent variable in the nonlinear relationship function is specifically a dynamic parameter, i.e., each thermal resistance. For example, taking winding temperature and winding-to-oil thermal resistance, the specific calculation method for the elements in the Jacobian matrix is ​​as follows: Similarly, the partial derivatives of other temperatures with respect to thermal resistance can also be calculated, and these partial derivatives form the Jacobian matrix. The essence of the nonlinear equation is to calculate the predicted temperature value at each time step based on the physical relationship between thermal resistance and the transformer thermal circuit model. To correct the predicted thermal resistance using actual temperatures, an extended Kalman filter is used to linearize the nonlinear equation. Furthermore, the linearization degree of the initial linear equation is verified using historical data. If the condition is not met, the linearization degree of the initial linear equation is re-acquired until the preset condition is met, thus overcoming the prediction error caused by insufficient linearization of the nonlinear equation and further improving the accuracy of transformer temperature prediction. In this embodiment, when the output data after inputting the historical data corresponding to the dynamic parameters into the initial linear equation matches the historical data corresponding to the observed parameters with a degree greater than a preset matching degree, it indicates that the preset condition is met. This preset matching degree can be flexibly set according to user needs or combined with experimental results of historical data, thereby indirectly improving the flexibility of the transformer thermal circuit model.

[0080] S4: Obtain the first predicted value of the dynamic parameter based on the state equation, obtain the second predicted value of the observed parameter based on the linear equation, and correct the first predicted value of the dynamic parameter based on the second predicted value and the actual value of the observed parameter to obtain the final value of the dynamic parameter.

[0081] In S4, the step of correcting the first predicted value of the dynamic parameter based on the second predicted value and the actual value of the observed parameter to obtain the final value of the dynamic parameter includes:

[0082] The inverse matrix to be determined is obtained by using the observation matrix corresponding to the linear equation, and the Kalman gain is obtained based on the inverse matrix to be determined.

[0083] The state estimate is obtained based on the Kalman gain, the second predicted value, and the actual value, and the state covariance matrix is ​​obtained. When the state covariance matrix meets the preset requirements, the first predicted value is the final value; otherwise, the state estimate is corrected based on the state covariance matrix to obtain the final value.

[0084] Before obtaining the Kalman gain based on the inverse matrix, the following steps are also included:

[0085] The desingularity parameters are obtained through cross-validation, and the matrix to be inverted is updated based on the desingularity parameters.

[0086] In this embodiment, the first predicted value is input into the linear equation to obtain the second predicted value. The first predicted value of the dynamic parameter, i.e., thermal resistance, can be obtained not only from the state equation but also calculated from the state estimate at the previous time step. Another step in the prediction stage is the prediction of the state covariance matrix to measure the uncertainty of the thermal resistance estimate at the current time step. The prediction of the state covariance matrix is ​​based on the state covariance matrix at the previous time step. Covariance of process noise The calculation yields the following formula: Finally, the state estimate at the current time step is adjusted based on the error between the actual value and the second predicted value. This includes calculating the Kalman gain, updating the state estimate using the actual value, and updating the state covariance matrix. The Kalman gain is used to balance the weights between prediction and observation errors, determining the degree of influence of the second predicted value on the state update. The formula for calculating the Kalman gain is: ,in: It is the Kalman gain; Let H be the state covariance matrix; H is the observation matrix, representing the relationship between the state parameters and the second predicted value. It is the observation noise covariance matrix. Let be the matrix to be inverted. After the Kalman gain is calculated, the state estimate of the thermal resistance is updated based on the difference between the actual value and the second predicted value. The new state estimate is given by the following formula: ,in: It is the updated state estimate (i.e., the thermal resistance value at the current time k); It is the state estimate in the prediction step; It is the actual value at the current moment; It is based on the predicted state The second predicted value is calculated, i.e., the value obtained through the observation equation. Then, the state covariance matrix is ​​updated. The calculation of the Kalman gain affects not only the update of the state estimate but also the update of the state estimate uncertainty. The updated state covariance matrix... Calculated using the following formula: ,in: It is the identity matrix; H is the Kalman gain; H is the observation matrix. It is the predicted state covariance matrix. For example, the predicted thermal resistance at time k is... Predicting the covariance matrix The observation matrix is ​​the identity matrix. ; Observation noise covariance matrix The actual observed value at the current moment The Kalman gain is then: Update state estimate: Update the covariance matrix: The process of correcting the state estimate based on the state covariance matrix to obtain the final value includes: correcting the state estimate using the state covariance matrix to obtain a corrected estimate, then obtaining the corrected covariance matrix, and determining whether the corrected covariance matrix meets preset requirements. If it does, the corrected estimate is the final value; otherwise, the process continues until the corrected covariance matrix meets the preset requirements, at which point the iteration stops. Through these steps, the change in thermal resistance in the transformer thermal circuit model can be estimated in real time, and a more accurate prediction of future temperature distribution can be made. The updated covariance matrix reflects the accuracy of the state estimate after correction using the actual value and the second predicted value. If the Kalman gain is large, it means a strong dependence on the second predicted value, and the updated state covariance matrix will be smaller, indicating reduced uncertainty; conversely, if the Kalman gain is small, it means the contribution of the second predicted value is small, and the state covariance matrix is ​​large. The preset requirements are set according to the accuracy of transformer temperature prediction. If the accuracy of transformer temperature prediction needs to be maximized, the preset requirements are set relatively low.

[0087] In this embodiment, the singularity removal parameters are obtained through cross-validation, and the matrix to be inverted is updated based on the singularity removal parameters as follows:

[0088] Cross-validation was used to divide the historical dataset into training and validation sets to evaluate the performance of the transformer thermal circuit model under different singularity removal parameters, and then the optimal singularity removal parameters were selected using the formula. The optimal desingularity parameters are incorporated into the matrix to be inverted to update the matrix. Let represent the updated inverse matrix to be found. This represents the matrix to be inverted before the update. This represents the optimal desingularity parameters. This represents the identity matrix. By updating the matrix to be inverted through desingonic parameters, the problem of high error rate in Kalman gain calculation caused by the singularity of the matrix to be inverted is overcome. This not only improves the accuracy of the final value but also enhances the adaptability and accuracy of the transformer thermal circuit model.

[0089] This invention dynamically adjusts the thermal resistance estimate during transformer operation to adapt to changes in factors such as temperature, oil flow, and environmental conditions. Furthermore, by introducing an extended Kalman filter, it can acquire measurement data in real time and adjust the estimate accordingly, thereby achieving precise tracking and adaptation to changes in thermal resistance. Moreover, with real-time adjustment of thermal resistance, it can more accurately predict temperature changes in the windings, oil, and tank, thus achieving precise temperature management. Effective temperature prediction helps to detect transformer overheating and other abnormalities early, providing an early warning mechanism, avoiding equipment failure, and extending the transformer's service life.

[0090] Example 2:

[0091] This embodiment also provides a storage medium storing computer-executable instructions. When the computer-executable instructions are loaded and executed by a processor, they implement the steps of the transformer thermal circuit model parameter correction method based on extended Kalman filtering.

[0092] The specific embodiments described above are preferred embodiments of the transformer thermal circuit model parameter correction method based on extended Kalman filtering of the present invention, and are not intended to limit the specific scope of the present invention. The scope of the present invention includes but is not limited to the specific embodiments described above. All equivalent changes made in accordance with the shape and structure of the present invention are within the protection scope of the present invention.

Claims

1. A method for correcting parameters of a transformer thermal circuit model based on extended Kalman filtering, characterized in that, Includes the following steps: S1: Define the state equations that characterize the changes of dynamic parameters in the transformer thermal circuit model over time based on the state transition matrix; S2: Based on the physical relationship between observed parameters and dynamic parameters in the transformer thermal circuit model, define a nonlinear equation characterizing the nonlinear relationship between observed parameters and dynamic parameters; S3: Linearize the nonlinear equations using extended Kalman filtering to obtain linear equations; S4: Obtain the first predicted value of the dynamic parameter based on the state equation, obtain the second predicted value of the observed parameter based on the linear equation, and correct the first predicted value of the dynamic parameter based on the second predicted value and the actual value of the observed parameter to obtain the final value of the dynamic parameter. S1 includes: S11: Based on the characteristics of the random distribution variables in the transformer thermal circuit model, obtain the cumulative distribution function of the random distribution variables, and then obtain the uniform distribution variables. Map the uniform distribution variables to Gaussian distribution variables through the inverse Gaussian distribution. S12: Obtain the state equation based on the Gaussian distributed variables and the state transition matrix; In S11, obtaining the cumulative distribution function of the random distribution variables based on the characteristics of the random distribution variables in the transformer thermal circuit model specifically means: obtaining the cumulative distribution function of the random distribution variables that do not conform to the Gaussian distribution. In S12, the state equation is: ； In the formula, This represents the value of the dynamic parameter at time k+1. Represents the state transition matrix. This represents the value of the dynamic parameter at time k. Represents the Gaussian distribution variable at time k; S3 includes: S31: Differentiate the nonlinear equation to obtain the initial linear equation, and input the historical data corresponding to the observed parameters and dynamic parameters into the initial linear equation to obtain the degree of linearization of the initial linear equation; S32: When the linearization degree meets the preset condition, the initial linear equation is a linear equation; otherwise, the initial linear equation is regarded as a nonlinear equation and S31 is executed. In S2, the observed parameters include at least the winding temperature, top oil temperature, oil tank temperature, and ambient temperature; The dynamic parameters include at least the thermal resistance between the winding and the oil, the thermal resistance between the oil and the oil tank, and the thermal resistance between the oil tank and the ring. In S2, the nonlinear relationship includes: ； ； ； In the formula, For winding temperature, For transformer load losses, For the winding heat capacity, The top oil temperature, For time step, For the winding to oil thermal resistance, For oil heat capacity, For no-load loss, For the fuel tank temperature, The thermal resistance between the oil and the oil tank. For the heat capacity of the fuel tank, For ambient temperature, This represents the thermal resistance between the oil tank and the ring.

2. The method for correcting transformer thermal circuit model parameters based on extended Kalman filtering according to claim 1, characterized in that, In S2, the nonlinear equation is: ； In the formula, Let k be the value of the observed parameter at time k. This represents a nonlinear relationship function between dynamic parameters and observed parameters. This represents the observed disturbance variable at time k.

3. The method for correcting transformer thermal circuit model parameters based on extended Kalman filtering according to claim 1, characterized in that, In S4, the step of correcting the first predicted value of the dynamic parameter based on the second predicted value and the actual value of the observed parameter to obtain the final value of the dynamic parameter includes: The inverse matrix to be determined is obtained by using the observation matrix corresponding to the linear equation, and the Kalman gain is obtained based on the inverse matrix to be determined. The state estimate is obtained based on the Kalman gain, the second predicted value, and the actual value, and the state covariance matrix is ​​obtained. When the state covariance matrix meets the preset requirements, the first predicted value is the final value; otherwise, the state estimate is corrected based on the state covariance matrix to obtain the final value.

4. The method for correcting transformer thermal circuit model parameters based on extended Kalman filtering according to claim 3, characterized in that, Before obtaining the Kalman gain based on the inverse matrix, the following steps are also included: The desingularity parameters are obtained through cross-validation, and the matrix to be inverted is updated based on the desingularity parameters.

5. A storage medium, characterized in that, The storage medium stores computer-executable instructions, which, when loaded and executed by a processor, implement the steps of the transformer thermal circuit model parameter correction method based on extended Kalman filtering as described in any one of claims 1 to 4.

Images