摘要
针对复杂的化工过程,在原有的自适应神经模糊推理系统(ANF IS)的基础上,结合主元分析和神经网络,提出了一种改进的自适应神经模糊推理故障诊断系统,并且分别将ANF IS和改进的ANF IS方法应用于TE(T ennessee E astm an)模型的故障诊断。两种方法均具有较高的精度,但改进的ANF IS具有运算速度快、结果清晰的优点,所以更适用于实际工业中。
This paper mainly focuses on the complicated chemical process and proposes an improved method, which integrates artificial neural networks and principal component analysis into the adaptive neural fuzzy inference system （ANFIS）. Then, ANFIS and improved ANFIS are applied in fault diagnosis of the TE model. Finally we conclude that both methods are very accurate, but the improved ANFIS has advantages with shorter time and clearer results. So it is more suitable for the industry.
出处
《华东理工大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2006年第8期985-988,共4页
Journal of East China University of Science and Technology
关键词
故障诊断
自适应模糊神经系统(ANFIS)
主元分析
神经网络
TE
fault diagnosis
adaptive neural fuzzy inference system （ANFIS）
principal component analysis
artificial neural networks
tennessee eastman （TE）
