Fault Detection Using Radial Basis Function Network and Polygonal Line

A Novel Modular Nonlinear Network for Fault Diagnosis and Supervised Pattern Classification

“…A non-parametric approach based on kernel density estimation (KDE) is used to determine the confidence limit. The detail of this methodology can be found in Bhushan et. al.…”

“…The number of hidden layer nodes was decided using cross validation scheme. A methodology based on "well -defined" architecture of radial basis function (RBF) network and polygonal line (PL) has been suggested by Bhushan and Romagnoli (2005) for dimensionality reduction and fault detection. The data of the normal operating region is used to fit the polygonal lines and the output generated has been used for determining the architecture of the network and to train the model.…”