The present invention relates to a fuzzy clustering analysis-based power distribution network fault line selection method. According to the method, the ratio of the variation quantity of the active component of the zero-sequence current of each outgoing line of a power distribution network to the total variation quantity of the active component of the zero-sequence current of the outgoing lines of the power distribution network is calculated; the simulation model of the actual operation of the power distribution network is constructed; and a fault line is determined through using a clustering analysis method. When a single-phase grounding fault occurs on the arc suppression coil grounding system of a neutral point, fault characteristics are not obvious; the grounding mode of the neutral point in the power distribution network is changed, fault signals are amplified, and therefore, line selection accuracy can be improved; a bus-bar potential transformer PT and a current transformer CT are adopted to measure the zero-sequence voltage of a bus-bar and the zero-sequence current of the outgoing lines before and after a resistor is switched; the variation quantity of the active component of the zero-sequence current of each outgoing line of lines is monitored; errors brought by measurement devices such as the current transformer CT of the line can be offset; and high anti-interference performance can be realized.