用模糊神经网络构造煤粉流量模型

USING FRACTIONAL BROWNIAN MODEL (FBM) TO DESCRIBE THE CONSTRUCTION ORGANISM OF THE BIOLOGICAL MACROMOLECULE

提出了用模糊神经网络的方法来建立煤粉流量模型．针对煤粉喷吹系统，模糊神经网络可以把它的工作空间分成若干个模糊区间，在每个区间中，用一局部模型来代表此系统；此方法克服了神经网络无法解释其联接权物理意义的不足．利用现场采集的数据进行实验，发现此方法不仅提高了准确度，而且可充分运用工作经验、专家知识。

The paper uses the FBM model to study the construction organism, cataract as the sample. First of all , hold that the best description of biological macromolecule -protein is FBM model. Secondly, apply the image texture analysis to the cataract macro properties; extend the two dimensional cataract data into one dimensional data according to the different window size and spreading style;use the maximum likelihood estimation (MLE) method to estimate the H value;then accomplish the segmentation and feature selection. Finally , point out that the best window size and spreading style is 6×6 and circle respectively, and that the Hurst parameter is the quantitative parameter of cataract diagnosis.