现有的基于 CMAC的自学习控制器能够有效地减小跟踪误差 ,但是在跟踪连续变化信号如正弦波时 ,由于累积误差的影响会产生过学习现象 ,进而导致系统的不稳定 .为此 ,提出一种新的基于 CMAC的自学习控制器 ,它以系统的动态误差作为 CMAC的激励信号 ,从而避免了累积误差的影响 .仿真结果表明 ,该控制器不仅是有效的 ,而且具有很强的鲁棒性 .此外 ,它可以使用较高的学习速率 ,实时性强 . Existing CMAC-based self-learning controller can effectively reduce the tracking error, but when tracking continuously changing signals such as sine wave, over-learning will occur due to the influence of accumulated error, which leads to system instability.Therefore, A new self-learning controller based on CMAC is proposed, which uses the dynamic error of the system as the excitation signal of the CMAC to avoid the influence of the cumulative error. The simulation results show that the controller is not only effective but also has strong In addition, it can use a higher learning rate, real-time and strong.