滑动窗与修剪技术的动态模糊神经网络方法研究

 提出了一种新型的动态模糊神经网络算法，该动态模糊神经网络的结构基于扩展的径向基网络。其算法的最主要特点是：使用滑动窗技术保持固定长度的数据来调整参数，避免了数据饱和；使用了修剪技术，使得网络结构不会持续增长，避免了过拟合及过训练现象，因而确保了系统的泛化能力。最后通过对血压的控制来验证动态模糊神经网络的逼近能力及所提控制方案的有效性。由仿真结果表明使用滑动窗与修剪技术的D-FNN控制器具有良好的性能。

Dynamic Fuzzy Neural Network Method Research of the Glide Window and Pruning Technology

Based on expansion radial basis function network,one kind of new dynamic fuzzy neural network algorithm was proposed.The algorithm's main feature is to maintain the fixed length of data by using glide window technology and avoid the data being saturated.The Pruning technology was used to make the network architecture unsustained growth,and avoid the fitting and the training phenomenon,thus,it has guaranteed system's pan-ability.Finally,the approach ability of the dynamic fuzzy neural network was confirmed t...