摘要
研究以模糊聚类和径向基函数网络结合的模糊径向基函数网络ＦＲＢＦＮ，并用主元分析对高维输入变量进 行预处理，降低了模型的输入变量维数，进而构造基于ＰＣＡ－ＦＲＢＦＮ的估计模型。这一方法通过对加氢裂化装置分 馏塔航空煤油干点估计得到验证。
In this paper, a fuzzy radial basis function neural network (FRBFN) is presented, which combining bob fuzzy clustering technique and radial basis function neural network. The principle component analysis (PCA) technique is applied to preprocessing high dimensional input variable so that the dimension of input variable is decreased, and an estimator model based on PCAFRBFN is constructed. This approach has been shown by the example of jet fuel endpoint estimation for a fractionator of the hydrocracking unit in the oil refining industry.
出处
《机电工程》
CAS
2000年第3期73-76,共4页
Journal of Mechanical & Electrical Engineering
关键词
航空煤油
干点估计
径向基函数
神经网络
RBF neural network
fuzzy clustering
principle component analysis
estilnator model
