Switched sliding mode approach for motion control systems with friction

Abstract

The chattering and offset phenomena, which may lead to acceleration disturbances and deteriorate control performance, are the main deficiencies caused by switching signal and mismatched uncertainties when the sliding mode control method is used for motion control systems with friction. The need is highlighted for simultaneous chattering suppression and mismatched uncertainties attenuation of sliding mode control. In this article, a switched sliding mode control method is proposed which uses three sliding surfaces to divide the error state space into three subspaces. Using different sliding mode controllers when error system state is located in different subspaces, the asymptotically stable of motion control system is guaranteed under friction and uncertainties (including matched and mismatched). In addition, through designing proper switching gain for each sliding mode controller, the control law of switched sliding mode control can be continuous. This means that the chattering phenomenon caused by switching gain will be suppressed effectively. For improving the tracking performance, a switched sliding mode control–based composite controller incorporating a LuGre model–based adaptive approximator is further presented. The predicted friction of LuGre model is used for friction compensation based on switched sliding mode control. Experimental results based on three different controllers verify the effectiveness of the proposed method on a test-bed.

Nonsmooth control based on finite-time disturbance estimator for the translational oscillator with a rotational actuator system with nonvanishing disturbances

“…Besides, robust control approaches like sliding mode control approaches take uncertain disturbances into account and exhibit strong robustness. 28,29 However, to reduce the effects of internal and external uncertainties, the switching gain of the sliding mode control approach should be chosen to be large enough, which could lead to the severe chattering problem. In addition, most existing methods tackle the uncertainties in a robust way and the price to pay is that their nominal control performances are degraded to some extent.…”