Abstract: To address the difficulty of quantifying the complex interactions and optimizing the starting strategy of multiple induction motors during short-term consecutive starting in mine power grids, a multi-induction motor starting strategy optimization method based on an improved Decision-Making Trial and Evaluation Laboratory (DEMATEL) approach is proposed. First, the interaction-related indicator data among motors are obtained through scanning simulations, and an improved DEMATEL direct influence matrix is constructed via discrete data aggregation. Then, the centrality and causality of each motor are calculated based on the comprehensive influence matrix, and the motor starting strategy is optimized accordingly. Finally, a case study of a mine power grid is conducted, where a multi-motor starting model is established on the MATLAB/Simulink platform, and different starting strategies are evaluated through simulation using selected performance indices. The results show that, compared with the empirical starting strategy, the optimized starting strategy based on the proposed method reduces the voltage sag energy by approximately 23.6% and shortens the total starting time by 8.5%. The findings demonstrate the feasibility and superiority of the proposed method in optimizing multi-induction motor starting in mine power grids, and provide a novel theoretical approach and technical pathway for multi-motor starting strategy optimization.