To the problem that the traditional PID controllers has low effectiveness when there is strong noise in feedback signal Quadrotor UAVs attitude control process, a PD controller based on Tracking-Differentiator (TD) is proposed, in which Particle Swarm Optimization (PSO) algorithm is applied to tune parameters. First, according to the tracking and filtering performance of TD to signals, the index for evaluating the tracking performance and filtering performance is proposed respectively based on tracking error and high-frequency information energy. Then, the weighted sum of them is used as fitness function of PSO algorithm. The simulation results verify that the proposed PD controller has good effectiveness in attitude control under the condition of strong noise.