Abstract: A dynamic maintenance decision-making method, integrating Weibull proportional hazards model (WPHM) and multi-objective optimization algorithm, was proposed to address the reliability-economy trade-off in gas turbine blade maintenance. Health indicators were constructed from operational state parameters to establish a failure probability assessment model for the blades. The cumulative failure probability and total maintenance cost were selected as dual optimization objectives, and Pareto-optimal maintenance schemes were derived using non-dominated sorting genetic algorithm (NSGA)-Ⅱ, NSGA-Ⅲ, and multi-objective particle swarm optimization (MOPSO) algorithm. Case study results demonstrate that the proposed method is capable of dynamically optimizing maintenance timing, maintenance type, and maintenance duration, thereby satisfying reliability constraints while minimizing operational expenditures. Compared with conventional periodic maintenance strategy, the three algorithms achieve maintenance cost reductions of 29.6%, 36.2%, and 42.7%, respectively. Among them, NSGA-Ⅲ exhibits superior performance under high-reliability operational scenarios, offering an effective decision-support framework for turbine blade maintenance planning.