Feedforward control

Due to their small sizes, micromechatronic systems are very sensitive to the environment (environmental vibration, thermal disturbance, ...). Furthermore, many of them are prone to nonlinearities: hysteresis and creep, quadratic nonlinearity.... These nonlinearities make them lose the accuracy although the very high resolution offered. Finally, most of the structures used for microsystems are based on cantilever (piezocantilever, …). This renders them typify badly damped vibrations which are unwanted for different reasons (stability compromised, settling time increased although the high bandwidth, …). To enhance the performances and to reach the sever specifications required in the micro-world and micro-scale in general, the control of these micromechatronic systems are essential. Closed-loop control techniques (robust, optimal,...) would be the first idea that comes in mind. Unfortunately, there is a lack of convenient sensors in general at this scale which therefore prevents from doing the feedbacks. For many micromechatronic systems, the available sensors that have the necessary performances in term of bandwidth, resolution and accuracy are bulky and very expensive (for instance: optical sensors based on triangulation, interferometer,...). On the other hand, embeddable sensors do not have these necessary performances. To sum up, there is a technological difficulties impeding the use of feedback or closed-loop control techniques.

An alternative and very interesting solution to feedback is feedforward (or open-loop) control configuration. It consists in putting in cascade with the process a controller that will compensate for the nonlinearities, the vibrations and the disturbance sensitivity typified (Fig.1). Additionally to the technological limitation avoided, feedforward control also permits low cost and high level of packageability. Feedforward techniques are classical in robotics, for instance the employment of the inverse kinematic control. In micromechatronic systems, the challenge of feedforward control techniques is very high because the process exhibits several behaviors to be accounted simultaneously (strong hysteresis and creep nonlinearities, vibrations, high sensitive to the environment).

Fig.1 Feedforward control of nonlinear, badly damped and highly sensitive (to environmental disturbance) micromechatronic systems.

 

Feedforward control strategies developed at the CODE-team also include the control of digital microrobotic systems without using sensors (Fig.1). The development of digital microbots at the CODE-team permits not only to have well discretized (and thus determinist) reachable positions, but also easier control strategies. In fact, since the positions versus the input are well determined, the uncertainties are more mastered and the use of open-loop control is more acceptable. The team develops therefore advanced feedforward control devoted to such digital microrobots.

 

Fig.2 Feedforward control of digital microrobotic systems.

Samples of publications

Yassine Haddab, Vincent Chalvet and Micky Rakotondrabe, Open-loop control approaches of compliant micromanipulators, in 'Flexible robotics: Applications to Multiscale Manipulations', Wiley - ISTE, ISBN 978-1-84821-520-7, July 2013.

Mohammad Al Janaideh, Micky Rakotondrabe and Omar Al Janaideh 'Further Results on Hysteresis Compensation of Smart Micro-Positioning Systems with the Inverse Prandtl-Ishlinskii Compensator, IEEE - Transactions on Control Systems Technology (T-CST), doi:10.1109/TCST.2015.2446959.

 

Didace Habineza, Micky Rakotondrabe and Yann Le Gorrec, 'Bouc-Wen Modeling and Feedforward Control of multivariable Hysteresis in Piezoelectric Systems: Application to a 3-DoF Piezotube scanner', IEEE - Transactions on Control Systems Technology (T-CST), Vol 23, Issue 5, Page 1797-1806, Sept 2015.

 

Micky Rakotondrabe, 'Classical Prandtl-Ishlinskii modeling and inverse multiplicative structure to compensate hysteresis in piezoactuators', ACC, (American Control Conference), pp.1646-1651, Montréal Canada, June 2012.

 

Micky Rakotondrabe, 'Modeling and Compensation of Multivariable Creep in multi-DOF Piezoelectric Actuators', IEEE - ICRA, (International Conference on Robotics and Automation), pp.4577-4581, St Paul Minnesota USA, May 2012.

 

Micky Rakotondrabe, 'Bouc-Wen modeling and inverse multiplicative structure to compensate hysteresis nonlinearity in piezoelectric actuators', IEEE - Transactions on Automation Science and Engineering (T-ASE), Vol.8, Issue.2, pp.428-431, April 2011.

 

Micky Rakotondrabe, Cédric Clévy and Philippe Lutz, 'Complete open loop control of hysteretic, creeped and oscillating piezoelectric cantilever', IEEE - Transactions on Automation Science and Engineering (T-ASE), Vol.7(3), pp:440-450, July 2010.

 

People involved

Yasser Al Hamidi, Nicolas Chaillet, Cédric Clévy, Yann Le Gorrec, Didace Habineza, Philippe Lutz, Micky Rakotondrabe

 

Contact

Micky Rakotondrabe, mrakoton 'at' femto-st 'dot' fr