Parameter Identification Of Equivalent Circuit And Fault Diagnosis For Distribution Transformer

Transformer is an important equipment in the energy transportation system,and off-line detection is the main diagnostic method.The realization of on-line status monitoring has great significance for the stable operation of the transformer.However,the high cost and the poor stability of the monitoring system obstacles its development;Using the current intelligent meter data to realize the transformer status monitoring without additional measuring device can solve the problem very well,which is the main content of this paper.Winding deformation,inter-turn short circuit and core failure are the most common transformer failures.Through the analysis of the relationship between the short circuit impedance,the excitation impedance in the equivalent circuit of the transformer and the above fault.Established the theoretical basis of discovering the occurrence of the transformer failure and the severity by the offset of the equivalent parameter.Firstly,the advantages and disadvantages of the ? and T-shaped equivalent circuits of the transformer are analyzed.Although the ?-shaped equivalent circuit is linear and the least squares method is easy to be used.The simulation results show that the parameters vary with the fluctuation of the load and cannot be used as the characteristic quantity of fault diagnosis.The T-shaped equivalent circuit can establish the parameter identification equation through the relationship between the low-side load impedance and the equivalent impedance of the highvoltage side,and solve the problem that the data is not synchronized on both sides.However,according to the sensitivity analysis of measurement error,it is found that the algorithm has high requirement for measurement accuracy,and it is proved that the equivalent parameters of high and low voltage side windings can not be identified separately by derivation.Then,based on the principle of power balance,a new T-shaped equivalent circuit linear identification equation is established.The relationship between the equation parameters and the equivalent circuit parameters can be proved by derivation,the excitation impedance and the sum of the high and low side winding impedances can be identified by the least squares method.According to the simulation analysis of load and parameter asymmetry,it is proposed to ensure the correctness of parameter identification by identifying the positive sequence component method.According to the sensitivity analysis of measurement error,the algorithm can tolerate the normal measurement error,and through the combination of positive sequence and negative sequence identification,the identification accuracy of short circuit impedance can be improved when the parameters are symmetrical,the occurrence of asymmetric failure can be indicated as soon as possible when the parameters are asymmetrical.According to the simulation results of the step mutation and continuous change of the transformer fault parameters,the equivalent parameters identification results and the least squares fitting error of the proposed algorithm have reference significance for transformer fault diagnosis.Finally,the linearized identification algorithm is preliminary validated by the actual monitoring data from the electric information collection terminal of a water supply center transformer.It is proving that the algorithm can obtain the transformer parameters accurately in different load intervals and have practical application value.