Controlling and learning motor functions

Abstract

Effective and adaptive motor functions are important for living beings and developing computational and learning mechanisms for roving robots is a crucial aspect in biorobotics. In this chapter we report a new architecture for motor learning to be applied in insect-like walking robots. The proposed model is based on the MB structure previously introduced able to memorize time evolutions of key parameters of the neural motor controller to improve existing motor primitives. The adopted control scheme enables the structure to efficiently cope with goal-oriented behavioural motor tasks. The problem of body-size evaluation is also considered and a model for the parallax-based estimation is provided. Finally, a six-legged structure, showing a steady-state exponentially stable locomotion pattern, was employed to modulate its motor commands implementing an obstacle climbing procedure. Simulation results on a Drosophila-inspired hexapod robot are reported.