Nonlinear system analysis with Karhunen-Loève transform (Q2499472)

The authors apply some methods developed for random systems to the analysis of nonlinear dynamical systems. The Karhunen-Loéve transform (KLT) is a well-known linear transform which approximates a stochastic process in least square sense. The authors use the discrete-time formulation of KLT to derive an individual orthonormal basis for each signal, and thus to decompose the deterministic signals. The KLT is used to study the transient dynamics of railway wheelset. The accelerations of the wheelset suspension are measured and analyzed. It is proved that the eigenvalues change in the wheelset rolling state, and the characteristic functions only can describe the principal directions of wheelset vibrations.