<p>With the increasing integration of distributed power sources, primarily clean energy, into distribution grids, their structures have become increasingly complex. During faults, traditional location methods struggle to achieve rapid and accurate fault location due to the influence of distributed power sources. This paper proposes a fault section location method based on negative-sequence components, utilizing negative-sequence voltage magnitude as the primary criterion and current direction as the auxiliary criterion. This approach effectively overcomes bidirectional power flow interference caused by photovoltaic power source integration. For symmetrical faults, a negative-sequence reconstruction method based on cell membrane genetic optimization algorithm is proposed. This algorithm finds the optimal solution for reconstruction weight coefficients, enabling fault location for all fault types. Concurrently, hierarchical dimensionality reduction is applied to the distribution network, partitioning the complex network into multiple fault zones and significantly enhancing the computational efficiency of the location model. Experimental results demonstrate that this method enables rapid and accurate fault section location in distribution networks incorporating DG, resolving the limitations of traditional methods in adapting to bidirectional power flows caused by DG integration.</p>