One kind from studies the controller based on the fuzzy radial direction primary function nerve network. 一种基于模糊径向基函数神经网络的自学习控制器。
Compared with existing approaches, auto-companding can help in controlling modeling error and reducing the number of basis functions employed. 与已有方法相比，采用自动压扩技术可以有效的控制模型误差并减少采用的基函数的个数，从而可以达到更好的建模效果。
