Comportamiento reactivo-evolutivo en navegación de robots móviles y su síntesis en VHDL para FPGA

FERNÁNDEZ LEÓN, J. A.; TOSINI, M.; ACOSTA, GERARDO G.; ACOSTA, N.

San Carlos, Brasil

Simposio; I Simposio Latinoamericano en aplicaciones de lógica programable y procesadores digitales de señales en procesamiento de video, visión computacional y robótica, SLALP 04; 2004

Univ. Estatal de San Pablo en San Carlos - CNPQ

Mobile robot’s navigation and obstacle avoidance in an unknown environmentis analyzed in this paper. From the guidance of position sensors, artificial neuralnetwork (ANN) based controllers settle the desired trajectory between current and atarget point. Evolutionary algorithms were used to choose the best controller. Thisapproach, known as Evolutionary Robotics (ER), commonly resorts to very simple ANNarchitectures. Although they include temporal processing, most of them do not considerthe learned experience in the controller’s evolution. Thus, the ER research presented inthis article, focuses on the specification and testing of the ANN based controllersimplemented when genetic mutations are performed from one generation to another.Discrete-Time Recurrent Neural Networks based controllers were tested. Also the wayin which evolution was performed was analyzed. In addition, an automatic hardwarecontroller generation is described here. As a result, controlled mutation do not exhibitmajor advantages against the non controlled one, showing that diversity is morepowerful than controlled adaptation.