Abstract

AbstractThe present research project consisted in the design of a humanoid robot with two ways of operation for displacement, the first one automatically with machine learning using Artificial Neural Networks and the second one controlled by voice commands. The methodology used is the concurrent design that contains five stages, conceptual design, kinematic analysis, dynamic analysis, mechanical design and simulation. The essential requirements for the design of the humanoid robot were determined, such as sensors to detect the presence of people or obstacles, DC motors for displacement, voice recognition module and for the neural network system the Arduino NANO was used. The trajectories of displacement of the humanoid robot were programmed, the neurons were trained for learning in matlab and all the simulated tests were performed with the voice recognition patterns, resulting in the desired performance. Keywords: humanoid robot, machine learning, speech recognition, artificial neural networks.