For large wind farms, the unit combinatorial optimization research can improve the level of wind farm operation and increase the economic efficiency of wind farms. The blades are one of the key components of wind turbines, which not only account for 20% of total wind turbine cost, but also have much influence on the wind turbine life. This paper makes a stress analysis of blades root at four different load conditions, quantifies damage of the blades, determines its relationship with blade life, and establishes a mathematic model of total blade damage value of each wind turbine unit. Moreover, based on the theory of improved binary particle swarm optimization (BPSO) algorithm, a unit commitment optimization model is proposed. For this model, the minimum of blade damage is taken as optimization objective, and wind power prediction results and load dispatch demand of electrical system are fully considered. A 49.5 MW wind farm is taken as an example and combinatorial optimization of 33 units is made. Case study shows that while meeting the load requirements, it can reduce the number of start-stop times and extend the unit life, verifying the feasibility and reliability of the proposed optimal dispatching model.