Analysis of Partitional Clustering Methods for Nonlinear Hybrid Dynamical Systems
Nonlinear Hybrid Dynamical Systems (NHDS) are characterized by interacting dynamics of continuous and discrete domain. Application such systems has been reported in chemical systems, manufacturing systems, mechanical systems, electrical systems, telecommunication systems, automobile control and computer disk drive control.The nonlinear continuous dynamic in NHDS will change due to occurrence of some unknown discrete events. So, for identification of NHDS, it is required to classify the open loop data according to discrete events. Clustering is the process of organizing objects into groups whose members are similar in some way. A variety of algorithms have recently emerged that meet these requirements and were successfully applied to real life data mining problem. Fuzzy cmeans (FCM) and k-means are commonly used partitional algorithm based on unsupervised learning methods. This paper focuses on the analysis of FCM and k-means partitional clustering methods for the single tank NHDS data classification.
Ankit K. Shah, Dipak M. Adhyaru