The identification and location of neutral line breakage faults are important for fault diagnosis in low voltage distribution networks, but traditional breakage detection methods based on current or voltage fault characteristics can lead to missed faults and misjudgements because of fluctuations in load and its imbalance. With the access of large-scale distributed new energy sources, fault characteristics change with the change of low voltage distribution network topology, and traditional fault identification criteria are no longer accurate nor effective. In this paper, a multi-criteria fusion method based on weight-correction D-S theory is proposed for neutral line break detection considering new energy access. Based on the analysis of the neutral break fault characteristics, the current fault characteristics based on the neutral and third harmonic current variations and the voltage fault characteristics based on the neutral offset voltage are improved. The impact of the new energy access on the multiple criteria is corrected. Finally, based on dynamic proximity history data, the D-S evidence theory is used to fuse the improved current and voltage multiple criteria information based on the results of the data. The D-S evidence theory is used to fuse the information of the improved current and voltage criteria to form a comprehensive neutral break fault criterion. The method improves the adaptability to different scenarios and solves the problem of misclassification caused by traditional detection. Simulation results verify the effectiveness of the solution.