レオロジーに基づく転がり摩擦モデリングとニューラルネットワークによるパラメータ同定

概要

論文の詳細を見る
This paper presents novel methodologies for a rheology-based rolling friction modeling and its parameterization in ball screw-driven table systems. The rolling friction exists in the table drive mechanism as a nonlinear component, which deteriorates the control performance in the fine positioning. The rolling friction behaviors, therefore, should be clarified to provide a precise numerical simulator and to design effective compensators. In the rolling friction modeling, a rheology-based friction model is introduced, considering the characteristics of contact points at the friction surface. The proposed friction model, in addition, is formulated by a neural network (NN), where the network parameters can be identified to express the actual friction behaviors. The proposed rolling friction modeling has been evaluated by a series of numerical simulations and experiments using a prototype for industrial positioning devices.
社団法人精密工学会の論文