The objective of this activity is the design of control strategies for the control of devices employing "smart" materials, i.e. innovative materials (piezoceramics, megnetostrictives, shape memory alloys, electro-rehologic and magneto-rehologic fluids, polymer gels, carbon nanotubes) to implement on embedded microcontrollers, so as to realize "smart devices".
In a first phase, piezoelectric materials have been considered, a "self-sensing" circuit has been designed and implemented, a vibration controller has been designed and implemented on a Motorola DSP board in the dSPACE rapid prototyping environment.
Next, magnetostrictive materials have been considered, and a controller has been designed in order to compensate for the hysteresis and saturation effects. Specifically, a Preisach model has been used to mathematically model the hystheresis, then the model parameters have been identified by using a fuzzy identifier, finally a pseudo-compensator has been defined, such that its output signal, when fed to the magnetostrictive actuator, makes the overall input-output behavior (almost) linear. By inserting the pseudo-compensator into a position feedback control loop, a fine positioning without overshoot in the position and peaks in the driving current has been obtained, differently from what happens without compensator. The controller has next been implemented on a 16-bit C167 Infineon controller, and different controllers have been connected in a CAN-based network, so as to obtain a distributed intelligent device network.
The pseudocompensator has also been used in double control loop solution, comprising an inner force loop and an outer position loop in order to have an accurate control of the strain and stress variable characterizing the magnetostrictive actuator.