Assessment of fuzzy logic approach for the prediction of dynamic stability on a mass-sliding belt experiment

Abstract

This study aims to investigate the predictability of a friction-induced nonlinear dynamic behavior on a simplified yet controlled laboratory experiment through the fuzzy logic approach. First, a mass-sliding belt experiment is built to observe the effects of several operating parameters on the occurrence of nonlinear dynamic behavior. Second, experiments are performed at various levels of these operating parameter, and the data are recorded. Third, fuzzy logic model architectures with different membership functions are built, where these operating parameters are assumed as the input parameters. The output of the fuzzy logic model architecture is defined as a new parameter called squeal index. Finally, a fuzzy logic model with a 96.97% prediction accuracy is obtained. Hence, it is shown that the proposed model can provide insight about the dynamic behavior of the system of interest without solving the nonlinear governing equations. Furthermore, the proposed model allows the prediction of the system state at operating conditions where experimentation is not possible, and it can be used for the determination of the critical operating parameters at which the system behavior switches from one state to another.