Modeling, Design and Construction of Articulated Hand for Use in Prosthetics, with Adaptive Control in Neural Networks Based on Mathematical Model for Finger

An Articulated hand model was designed and built by CAD-CAM software, analyzed by finite element in the structure; the hand was made for a 3D printer in a first time, the construction of each finger suggested a mathematical model that describes the Inertia Matrix, Gravitational Matrix, and the Coreolis Matrix; the resulting mathematical model is three degrees of freedom, the finger control is based for Neural Networks that determine the behavior of the actuators independent of each phalanx. Such movements react to capture a myoelectric signal, from amputee arm trans-radius, processing the myoelectric signal determines the closure or opening of the hand with fingers. The electrode design was created with a structural materials in carbon fiber to delete the spurious signals, and obtain a clean command signal.