With the enlargement of the wind farm scale, the ability of accommodating wind powers of network becomes a main factor of wind farm planning. This paper makes a scheduling model comprehensively in consideration of operation cost of conventional units and wind curtailment with a wind power acceptance level parameter to determine the optimal level of wind power acceptance. And a wind power acceptance level optimization based particle swarm algorithm is proposed in the angle of the economical operation of conventional units and wind powers. This paper calculates in a system with one wind farm and 10 conventional units and works out the optimal level of wind power acceptance respectively of the 12 groups of typical output data. And then a comprehensive optimal acceptance level is determined using approaching method and Newton interpolation. The system is in an optimal economic operation at this level and proportion coefficient can be a reference of wind power planning.