En este trabajo se desarrolla una variación del algoritmo de campos potenciales para la planificación de trayectorias que evita obstáculos aplicado a un manipulador antropomórfico de 6 grados de libertad. En primera instancia se generó el modelo de cinemática inversa basada en un proceso de control iterativo multivariable, luego se modificó el modelo agregando un vector de rotación obtenido por las fuerzas repulsivas entre el obstáculo y las seis articulaciones del robot de manera que el manipulador pueda encontrar una ruta que evite el obstáculo y alcance una posición objetivo. Este modelo final de cinemática inversa con campos potenciales genera trayectorias que dependen del parámetro de velocidad de optimización y de los coeficientes de campos vectoriales. Con el objetivo de optimizar las trayectorias se generó una base de datos con los puntos iniciales, finales y de obstáculos de diferentes trayectorias con sus respectivos parámetros optimizados para entrenar una red neuronal supervisada. Los resultados muestran que la red neuronal debe entrenarse con una mayor cantidad de datos debido a que se calculan parámetros erróneos para ciertas posiciones iniciales y finales. Finalmente, la simulación del manipulador de 6 grados de libertad que sigue la trayectoria generada por el modelo de cinemática inversa y campos potenciales con los parámetros optimizados calculados empíricamente resultan en un modelo con un comportamiento adecuado logrando evitar los obstáculos.
In this document, we propose a system architecture that aims to improve the performance and learning of basic sciences in STEM undergraduate students at Peruvian national universities. For this purpose, we first analyze the problems that arise in the most representative STEM university of Peru, the National University of Engineering (UNI by its initials in Spanish). Then, we review some papers with topics on STEM education, Human-Agent Interaction and Artificial Intelligence to finally propose the mentioned system and explain how it works. So, since the present document contains the first progress of a project that is currently under development, it will serve as a basis for future experiments, which will be defined and discussed at the end of this paper.
Abstract-In this document, we propose a system architecture that aims to improve the performance and learning of basic sciences in STEM undergraduate students at Peruvian national universities. For this purpose, we first analyze the problems that arise in the most representative STEM university of Peru, the National University of Engineering (UNI by its initials in Spanish). Then, we review some papers with topics on STEM education, Human-Agent Interaction and Artificial Intelligence to finally propose the mentioned system and explain how it works. So, since the present document contains the first progress of a project that is currently under development, it will serve as a basis for future experiments, which will be defined and discussed at the end of this paper.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.