A Measurement based Predictive Maintenance Algorithm for Rigid-Body Dynamical Systems using Radial Basis Function Approximation and Information Theory

This paper presents new viewpoints to solve predicative maintenance problems for arbitrarily rigid-body mechanical systems. Reliable predictions for changes in physical parameters are highly dependent on system model accuracy. Hence, costs for physical system modeling are expensive and time-consuming. To overcome this problem, we replaced the modeling process with a full measurement based procedure in conjunction with radial basis functions (RBF) approximation to create the system model. Subsequently, information theory is employed to generate a meaningful metric quantity that reports about the changes in physical parameters. A rigid-body mechanical system is used as an example to verify the presented theory.