Due to the high fault probability of overhead lines in multi-terminal DC grids and limited overcurrent tolerance of equipment, non-selective reclosing can severely compromise equipment safety and grid stability. To mitigate these risks, a reclosing strategy utilizing dynamic time warping (DTW)-based sequence reconstruction is developed to enable accurate fault nature identification before reclosing. First, the time-domain equivalent modeling of DC lines using the Bergeron method is introduced. On this basis, equivalent Bergeron models of key primary equipment are established, and differences in equivalent time-domain waveform processes under transient and permanent fault conditions are analyzed. Taking transient faults as the benchmark, the theoretical calculations of the transient electrical components induced by the injection signals of the DC circuit breakers are conducted. The DTW reconstruction sequence is then used to quantify the setting thresholds and eliminate dispersion errors. This forms the basis for constructing a criterion to identify the nature of the fault. Simulation case studies based on a four-terminal DC grid using PSCAD/EMTDC demonstrate that the proposed adaptive reclosing method requires only single-ended measurements, does not rely on communication, eliminates dead zones, and possesses a sensitivity of 600 .