Regensburg 2013 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

DS: Fachverband Dünne Schichten

DS 12: Poster Session II: Functionalized semiconductor nanowires (jointly with HL); Resistive switching (jointly with DF, KR, HL); Thermoelectric materials

DS 12.9: Poster

Montag, 11. März 2013, 17:00–20:00, Poster B1

Adaptive Robot learning via (simulated) memresistive Elements — •Marius Schirmer and Andy Thomas — Bielefeld University, Bielefeld, Germany

Memristors are new circuit elements that fill the gap in relations between the circuit variables flux φ and charge q. Like their name consisting of Memory and Resistor presumes, they are resistors with memory. The resistance depends on the current i(t) that was applied to an element earlier. This behavior is similar to the synapse of a living nerve cell and therefore networks of memristors can be used as neuronal networks in hardware circuits.

To visualize the learning process of such a memresistive neuronal network we used a Lego Mindstorms NXT robot and designed an interface for using memristors that can be connected to the NXT brick. We also designed a circuit board that is simulating memresistive behavior. The Software written in Java allows several different learning tasks.