To solve the problems that it is difficult to extract fault features in strong noise background and fault conditions randomly occur when signal phase ground faults occur in non-solid-earthed network

we proposed a fault line selection method based on optimized bistable denoising for non-solid-earthed network. Firstly

based on the analysis of the influence of noise on transient zero-sequence current（TZSC）

the genetic algorithm and bistable system were applied in the denoising of TZSC. Secondly

according to the peak value of TZSC

the fault situation was divided into three mode: strong fault mode

medium fault mode

and weak fault mode

and the three optimized bistable systems were obtained. Thirdly

the TZSC for each line was denoised by these three bistable systems

and the transient characteristic zero-sequence current（TCZC） was acquired by the principle of peak current. Finally

the polarity parameters for each lines were determined by the TCZC

and a fault line selection criterion based on the polarity parameter and polarity standard deviation was proposed. The simulation results show that

the proposed method can select fault line exactly in different fault environments including different fault resistances

fault angle

ault distance

and so on

even in the strong noise background of which the signal-to-noise ratio is-1 d B

the proposed method can filter the noise

when the fault is not busbar fault

it can make the polarity standard deviation bigger than 0.01 and the polarity parameter of faulty line is smaller than healthy line; when the fault is busbar fault

it can make the polarity standard deviation smaller than 0.01. The research results verified the effectiveness of the proposed fault line selection method.