Modeling and Parameter Identification Algorithm for Tree-Contact Single-Phase-to-Ground Fault in Distribution Networks

The tree-contact single-phase-to-ground fault (TSF) in 10 kV distribution networks has high transition resistance, weak fault currents, and nonlinear steady-state waveforms. As existing high-impedance fault models cannot accurately describe its complete physical evolution, this paper proposes a novel modeling and parameter identification algorithm for TSF. First, based on recorded data from full-scale experiments, the initiation and development processes of TSF are studied, revealing the main factors affecting fault electrical characteristics—such as moisture evaporation, pyrolysis carbonization, air gap breakdown, and tree body current dissipation. Then, a dynamic resistance series model for TSF is constructed, with parameters identified and calibrated using experimental data, objective functions, and physical constraints. Finally, a 10 kV TSF simulation model is built and verified. Furthermore, a cross-condition predictive validation is performed using different voltage and geometric boundaries. Results demonstrate that the proposed physics-constrained model can effectively reproduce the RMS fault current envelope with asymmetric moisture evaporation characteristics. It also accurately predicts steady-state nonlinear waveform features without parameter re-tuning, providing more physically consistent data support for future TSF identification studies.