CN-122000869-A - Actual measurement correction method, device, system and medium for large power grid magnetic bias model parameters

CN122000869ACN 122000869 ACN122000869 ACN 122000869ACN-122000869-A

Abstract

The invention discloses an actual measurement correction method, device, system and medium for large power grid magnetic bias model parameters, and provides an accurate correction scheme aiming at the problem of insufficient accuracy of a magnetic bias evaluation model caused by the complexity of the earth resistivity. The method comprises the steps of constructing a large alternating current power grid magnetic bias theoretical model, outputting a node voltage equation and a transformer neutral point loop current expression, introducing a transformer substation ground surface potential correction column vector to generate a corrected model and a corrected current column vector, constructing a function by taking the minimum sum of corrected absolute values and measured current deviation as a target, and solving by adopting a hybrid quasi-Newton method to obtain the corrected column vector. The method can effectively compensate model errors, effectively solve correction parameters, dynamically adapt to power distribution network changes, and remarkably improve bias magnetic evaluation reliability and efficiency.

Inventors

WU CHUANQI
REN JIESHUAI
YU YANG
ZHANG YINGYING
ZHANG YI
XIAO MIAN
PAN ZHUOHONG
GOU XIAOTONG
TANG YU
SHU LONG
LI KAIYU
LI JINBIN
MENG YI

Assignees

国网湖北省电力有限公司电力科学研究院
华北电力大学（保定）
国网经济技术研究院有限公司

Dates

Publication Date: 20260508
Application Date: 20260105

Claims (12)

1. The actual measurement correction method of the large power grid magnetic bias model parameters is characterized by comprising the following steps: Constructing a large alternating current power grid magnetic bias theoretical model, and outputting a node voltage equation of the theoretical model and a transformer neutral point loop current expression; Introducing a transformer substation ground potential correction column vector based on the node voltage equation, generating a corrected node voltage equation, and outputting a corrected transformer neutral point loop current column vector based on the corrected node voltage equation and the transformer neutral point loop current expression; Constructing an objective function of a transformer substation ground potential correction column vector by taking the minimum sum of deviation absolute values of the corrected transformer neutral point loop current column vector and the actual transformer neutral point loop current column vector as a target; and solving the objective function by adopting a hybrid quasi-Newton method to obtain a correction column vector of the ground surface potential of the transformer substation.
2. The method for actually measuring and correcting the parameters of the magnetic bias model of the large power grid according to claim 1, wherein the steps of constructing a magnetic bias theoretical model of the large power grid, outputting a node voltage equation of the theoretical model and a current expression of a neutral point loop of a transformer include: the method comprises the steps that a bus of a transformer substation and a neutral point of a transformer are equivalent to be nodes, a direct-current grounding resistor of the transformer substation, a direct-current resistor of a transformer winding and a direct-current resistor of a circuit are equivalent to be branches, and an equivalent network is formed; Deriving a node voltage equation of a theoretical model based on the equivalent network; Solving to obtain a node voltage column vector based on the node voltage equation; And deducing branch current column vectors based on the node voltage column vectors, and further obtaining a transformer neutral point loop current expression under a theoretical model.
3. The method for actually measuring and correcting the parameters of the magnetic bias model of the large power grid as set forth in claim 2, wherein the step of solving the objective function by adopting the hybrid quasi-Newton method to obtain a corrected column vector of the ground surface potential of the transformer substation comprises the steps of: setting a control error epsilon=1e- 6 , an initial point x 1 =zero vector, an initial matrix B 1 =unit matrix, and the iteration number k=1; For the current iteration point x k , calculating the gradient g k of the objective function; Performing Cholesky decomposition on B k , and selecting a BFGS or DFP form update matrix B k+1 according to a solution increment s k =x k+1 -x k and a gradient increment y k =gk +1 -g k ; repeating gradient calculation and matrix updating until l k p k || 2 is less than or equal to epsilon, wherein l k is the step length and the p k searching direction, and outputting a correction column vector xi of the ground surface potential of the correction vector transformer substation.
4. The method for actually measuring and correcting the parameters of the magnetic bias model of the large power grid according to claim 1, further comprising outputting corrected direct current distribution and magnetic bias risk assessment results based on the correction column vectors, and adapting to power grid parameters, operation mode changes or governance measures to adjust scenes.
5. The method for actually measuring and correcting the parameters of the magnetic bias model of the large power grid according to claim 1, wherein the method for actually measuring and correcting the parameters of the magnetic bias model of the large power grid based on the corrected column vector output corrected DC current distribution and magnetic bias risk assessment results and adapting to the parameters of the power grid, the change of the operation mode or the adjustment scene of the treatment measures comprises the following steps: Substituting the corrected column vector into the corrected node voltage equation to obtain an updated node voltage equation; Calculating the direct current distribution of the power grid based on the updated node voltage equation; Based on the distribution of net direct current, finishing bias magnetic risk assessment; and repeating the steps aiming at the power grid parameter adjustment, the operation mode change or the treatment measure implementation scene, and outputting the updated evaluation result without re-actual measurement.
6. The actually measured correction device for the large power grid magnetic bias model parameters is characterized by comprising the following components: The theoretical model construction module is used for constructing a magnetic bias theoretical model of the large alternating current power grid and outputting a node voltage equation of the theoretical model and a transformer neutral point loop current expression; The correction model generation module is used for introducing a transformer substation ground surface potential correction column vector based on the node voltage equation, generating a corrected node voltage equation, and outputting a corrected transformer neutral point loop current column vector based on the corrected node voltage equation and the transformer neutral point loop current expression; the objective function construction module is used for constructing an objective function of the transformer substation ground potential correction column vector by taking the minimum sum of deviation absolute values of the corrected transformer neutral point loop current column vector and the actual transformer neutral point loop current column vector as an objective; and the correction column vector solving module is used for solving the objective function by adopting a hybrid quasi-Newton method to obtain the correction column vector of the ground surface potential of the transformer substation.
7. The device for actually measuring and correcting the parameters of the magnetic bias model of the large power grid according to claim 6, wherein the theoretical model construction module is specifically configured to: the method comprises the steps that a bus of a transformer substation and a neutral point of a transformer are equivalent to be nodes, a direct-current grounding resistor of the transformer substation, a direct-current resistor of a transformer winding and a direct-current resistor of a circuit are equivalent to be branches, and an equivalent network is formed; Deriving a node voltage equation of a theoretical model based on the equivalent network; Solving to obtain a node voltage column vector based on the node voltage equation; And deducing branch current column vectors based on the node voltage column vectors, and further obtaining a transformer neutral point loop current expression under a theoretical model.
8. The device for actually measuring and correcting the parameters of the magnetic bias model of the large power grid according to claim 7, wherein the correction column vector solving module is specifically configured to: setting a control error epsilon=1e- 6 , an initial point x 1 =zero vector, an initial matrix B 1 =unit matrix, and the iteration number k=1; For the current iteration point x k , calculating the gradient g k of the objective function; Performing Cholesky decomposition on B k , and selecting a BFGS or DFP form update matrix B k+1 according to a solution increment s k =x k+1 -x k and a gradient increment y k =gk +1 -g k ; repeating gradient calculation and matrix updating until l k p k || 2 is less than or equal to epsilon, wherein l k is the step length and the p k searching direction, and outputting a correction column vector xi of the ground surface potential of the correction vector transformer substation.
9. The device for actually measuring and correcting the parameters of the magnetic bias model of the large power grid according to claim 7, further comprising: The dynamic evaluation module is used for outputting corrected direct current distribution and magnetic bias risk evaluation results based on the correction column vectors, and adapting to power distribution network parameters, operation mode changes or governance measures adjustment scenes.
10. The device for actually measuring and correcting the parameters of the magnetic bias model of the large power grid according to claim 9, wherein the dynamic evaluation module is specifically configured to: Substituting the corrected column vector into the corrected node voltage equation to obtain an updated node voltage equation; Calculating the direct current distribution of the power grid based on the updated node voltage equation; Based on the distribution of net direct current, finishing bias magnetic risk assessment; and repeating the steps aiming at the power grid parameter adjustment, the operation mode change or the treatment measure implementation scene, and outputting the updated evaluation result without re-actual measurement.
11. An actual measurement correction system for large power grid bias magnetic model parameters comprises a computer readable storage medium and a processor; The computer-readable storage medium is for storing executable instructions; The processor is configured to read executable instructions stored in the computer readable storage medium, and execute the actual measurement correction method of the large power grid bias model parameter according to any one of claims 1 to 5.
12. A non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method of actually modifying a large grid bias model parameter according to any one of claims 1 to 5.

Description

Actual measurement correction method, device, system and medium for large power grid magnetic bias model parameters Technical Field The invention relates to the field of DC magnetic bias risk management, in particular to a method, a device, a system and a medium for actually measuring and correcting parameters of a magnetic bias model of a large power grid. Background Because the wiring operation mode of the power grid is complex, and the earth resistivity has non-uniformity of plane distribution and section distribution, the simulation result of DC magnetic bias evaluation is different from the measured data, especially in a large power grid, DC grounding greatly returns current with complex distribution, and the difficulty of magnetic bias risk evaluation and management is further increased. Before the direct current engineering is put into operation, the bias magnetic risk management needs to rely on a simulation model to predict the risk, the requirement on model precision is high, after the direct current engineering is put into operation, the management measure configuration depends on measured data, but the new invasion current distribution formed by the power grid after the management cannot be actually measured, the evaluation result can only be obtained through simulation, and if the deviation between the simulation model and the measured data is overlarge in the debugging stage, the follow-up risk evaluation precision is insufficient, and hidden danger is brought to the power grid operation. Disclosure of Invention The invention provides an actual measurement correction method, device, system and medium for parameters of a large power grid direct current magnetic bias evaluation model based on the problem of calculation deviation of the large power grid direct current magnetic bias evaluation model caused by the structural complexity of the earth resistivity, wherein actual measurement correction model is built by introducing transformer neutral point direct current actual measurement data as a core basis, and the correction amount is solved by combining a mixed form quasi-Newton method, so that the accurate correction of the magnetic bias evaluation model is realized. An actual measurement correction method for large power grid bias magnetic model parameters comprises the following steps: Constructing a large alternating current power grid magnetic bias theoretical model, and outputting a node voltage equation of the theoretical model and a transformer neutral point loop current expression; Introducing a transformer substation ground potential correction column vector based on the node voltage equation, generating a corrected node voltage equation, and outputting a corrected transformer neutral point loop current column vector based on the corrected node voltage equation and the transformer neutral point loop current expression; Constructing an objective function of a transformer substation ground potential correction column vector by taking the minimum sum of deviation absolute values of the corrected transformer neutral point loop current column vector and the actual transformer neutral point loop current column vector as a target; and solving the objective function by adopting a hybrid quasi-Newton method to obtain a correction column vector of the ground surface potential of the transformer substation. Further, the construction of the large alternating current power grid magnetic bias theoretical model, the node voltage equation of the output theoretical model and the transformer neutral point loop current expression, includes: the method comprises the steps that a bus of a transformer substation and a neutral point of a transformer are equivalent to be nodes, a direct-current grounding resistor of the transformer substation, a direct-current resistor of a transformer winding and a direct-current resistor of a circuit are equivalent to be branches, and an equivalent network is formed; Deriving a node voltage equation of a theoretical model based on the equivalent network; Solving to obtain a node voltage column vector based on the node voltage equation; And deducing branch current column vectors based on the node voltage column vectors, and further obtaining a transformer neutral point loop current expression under a theoretical model. Further, the solving the objective function by adopting the hybrid quasi-newton method to obtain a correction column vector of the ground surface potential of the transformer substation comprises: setting a control error epsilon=1e- 6, an initial point x 1 =zero vector, an initial matrix B 1 =unit matrix, and the iteration number k=1; For the current iteration point x k, calculating the gradient g k of the objective function; Performing Cholesky decomposition on B k, and selecting a BFGS or DFP form update matrix B k+1 according to a solution increment s k=xk+1-xk and a gradient increment y k=gk+1-gk; repeating gradient calculation and matrix updating until l