Power istribution network fault line selection method based on fractal dimension calculation and mallat decomposition

The invention discloses a power istribution network fault line selection method based on fractal dimension calculation and mallat decomposition. The method comprises obtaining zero-sequence voltages of a power distribution network bus and zero-sequence currents of each outlet wire under the condition that a fault occurs in a power distribution network; performing multilayer mallat decomposition on the obtained zero-sequence voltages of the power distribution network bus and the zero-sequence currents of each outlet wire, and according to a decreasing sequence of zero-sequence current transient modulus maximum values, successively screening a plurality of outlet wires of the power distribution network bus; extracting node coefficients in a plurality of transient frequency sets after the mallat decomposition from the zero-sequence voltages and zero-sequence currents of the screened outlet wires; according to a wavelet maximum polarity line selection theory, obtaining a primary line selection result; determining whether the primary line selection result is one outlet wire of the power distribution network bus, if so, determining that the outlet wire is a fault line, and ending line selection; and otherwise, calculating fractal dimensions corresponding to the zero-sequence currents of each outlet wire in the primary line selection result, and determining that the outlet wire, with a maximum fractal dimension, of the power distribution network bus, is a fault line. The method provided by the invention improves the accuracy and reliability of the line selection result.