A precise fault location method incorporating Kernel Density Estimation is proposed to resolve the challenge of location failure induced by time asynchronization among measurement points in distributed traveling-wave ranging systems. Faulty segments are first identified by utilizing the polarity of current waves. Subsequently, wave arrival times from remote measurement points are projected onto both ends of the identified segment. An initial fault vector-matrix is constructed and then transformed into a fault location vector-matrix indicating precise fault position. This matrix, when mapped onto a coordinate system, forms a matrix scatter plot. Outliers within this scatter plot are eliminated via Kernel Density Estimation. The reliability of the remaining points is then comprehensively evaluated based on the propagation time between measurement points, line distance, and TW energy attenuation, leading to the optimal final location. The proposed method achieves high accuracy under severe time-synchronization deviations without requiring additional hardware, significantly enhancing the robustness of distributed TW systems. Its reliability and practical applicability are validated through PSCAD/EMTDC simulations.