针对可重复轨迹跟踪问题，提出了一种ＰＤ型模糊学习算法．该算法集成两种控制：作为基础的ＰＤ型模糊逻辑算法和改善系统性能的学习算法．模糊学习控制在模糊控制基础上引入迭代学习算法，使得模糊ＰＤ控制器可以精确地跟踪可重复轨迹以及消除周期性扰动．本文在能量函数和泛函分析的基础上，通过严格的推导表明ＰＤ型模糊学习算法可达到：１）系统跟踪误差一致收敛到零；２）学习控制序列几乎处处收敛到理想的控制信号． Aiming at the problem of repeatable trajectory tracking, a PD-based fuzzy learning algorithm is proposed. The algorithm integrates two types of control: the underlying PD-based fuzzy logic algorithm and the learning algorithm to improve system performance. Fuzzy learning control introduces iterative learning algorithm based on fuzzy control, which makes the fuzzy PD controller can accurately track repeatable trajectories and eliminate periodic perturbations. On the basis of energy function and functional analysis, this paper shows that the PD-type fuzzy learning algorithm can be achieved through strict derivation: 1) the tracking error of the system converges to zero uniformly; 2) the learning control sequence converges to the ideal control signal almost everywhere.