During off-peak periods, passengers’ choice of the subway is more based on its accessibility and convenience. If the energy-saving operation mode can be integrated into the train automatic operation system, it is possible to reduce traction energy consumption without affecting the service level. For this purpose, this paper proposes an energy-saving operation strategy based on the traction characteristics of subway trains. By accurately describing the inter-station running time constraints and speed limits, as well as considering the allowable range of operation time fluctuations, a fuzzy dynamic programming model is constructed with the goal of minimizing inter-station energy consumption. The model is solved using the Capacity-constrained Non-dominated Sorting (CNS) algorithm, which further saves program execution time while ensuring solution accuracy. Taking the Beijing Subway Yizhuang Line as an example, numerical simulation results show that the optimized train operation sequence scheme reduces inter-station traction energy consumption compared to the current fastest operation scheme, and the line operation time is also controlled within an acceptable range. The effectiveness and practicality of the algorithm are verified by comparing and analyzing simulation results under different parameter settings.