Underactuated fingers controlled by robust and adaptive trajectory following methods

Abstract

Choosing an appropriate control scheme to alleviate nonlinearities and uncertainties is not a trivial task, especially when models are not easily available and practical evaluation provides the only means for actual performance assessment. Various factors can contribute to these nonlinearities and uncertainties, such as friction and stiction. Thus, this article investigates four different control schemes, namely PID, adaptive, conventional sliding mode control (SMC) and integral sliding mode control (ISMC) which are implemented in the Bristol Elumotion Robot Hand (BERUL) to analyse and overcome the aforementioned problems. The hand has five fingers with 16 joints and all fingers are underactuated. The implementation of the proposed control schemes are challenging since the BERUL fingers have significant friction, stiction and unknown parameters. The fingers are light in weight and fragile. Comparative performance characteristics have shown that the ISMC is the most suitable candidate to provide good experimental trajectory following and positioning control for underactuated BERUL fingers.

Keywords:

• trajectory following
• integral sliding mode control
• sliding mode control
• PID control
• adaptive control
• underactuated robotic fingers

Acknowledgements

The CHRIS (Cooperative Human Robot Interaction Systems) project is funded by the European Commission's Seventh Framework Programme (FP7) and will run from 2008–2012. This research is also funded by the Malaysian Government.

Notes

1. This example model can be easily generalised for fingers with more links.

2. Note that, using r from (9), the control in (19) is indeed a PID controller.

3. These assessment criteria are practically motivated where model-based design approaches fail. This is needed here in this case since nonlinearities due to friction and stiction are significant and changing, i.e. practical designs for uncertain systems are required.