Parameter Identification Method for Distribution Systems with a High Proportion of Distributed Renewable Energy

[Objective] The integration of a high proportion of distributed generation into the distribution system leads to frequent and significant voltage fluctuations， which negatively affect the accuracy of voltage measurement data and increase the difficulty of parameter identification based on full measurement data of the distribution system. [Methods] This paper proposes a solution to the parameter identification challenge based on full measurement data in distribution systems by establishing an adaptive extended Kalman filtering （AEKF） model based on current measurements. Specifically， the state equations for the line parameters are derived by incorporating historical temperature data， whereas those for the transformer parameters are established using a first-order exponential smoothing method. By applying Kirchhoff's law， the current amplitude measurement equations for the lines and transformers are constructed. A voltage pseudomeasurement strategy is introduced to construct pseudomeasurement equations based on prior parameter estimates. An adaptive noise mechanism is designed in which the noise covariance matrix is dynamically estimated based on the differences in adjacent measurements， thereby enhancing the robustness of the algorithm against time-varying noise. [Results] The results showed that the relative mean identification errors for line resistance and reactance were 1.45% and 1.61%， respectively， a reduction of 61.2% and 61.7% compared with those of the traditional EKF method with errors of 3.7% and 4.2%. The maximum relative error decreased from 12.6% to 2.4%. The peak identification errors for transformer resistance and reactance were 5.2% and 5.53%， and the single iteration time was optimized from 2.2 to 1.04 s. [Conclusion] This method effectively addresses the impact of deficiencies in voltage measurement data quality on parameter identification by integrating current amplitude data with historical operational information. Pseudomeasurement modeling and adaptive noise estimation techniques work synergistically to enhance the stability and computational efficiency in complex power grid environments， helping address the challenges posed by the randomness and volatility of distributed generation and improving the fine-tuned control capabilities of the distribution system.