Desarrollo de un algoritmo de trayectoria para un robot seguidor de línea destreza de competencia mediante visión e inteligencia artificial

Abstract

Throughout time, artificial vision technology has developed rapidly, promoting various applications. However, in our country, the development of dexterity line follower robot prototypes based on artificial vision is scarce during national competitions. This has created the opportunity to employ, test, and implement intelligent algorithms that allow the control in the trajectory of a prototype robot follower of the dexterity line using cutting-edge technology. This work of research and experimentation is directed towards the development of a dexterity line follower mobile robot with differential architecture manufactured with SLA technology in conjunction with a neural network with artificial vision, based on different studies of free software for this application, considering the different characteristics and regulations of the prototype in a determined category. It was determined that the OpenMV CamH7 card with an ARM Cortex A8 Sitara AM335X core was suitable to allocate the intelligent algorithms such as neural networks to control the trajectory of the prototype robot for the development of the case study. Through the creation of the prototype in conjunction with neural networks to control the trajectory through artificial vision, as a result, 100% effectiveness was obtained in the application of a neural network with 12 neurons in the hidden layer in a controlled scenario with a high degree of difficulty.