The Use of a Statistical Filter and Metaheuristics to Model and Control The DC Motor of the Mobile Robot Used on NXP Cup

Abstract

Over the past decades, Robotics is one of the research fields with more advances. From methodologies of low-level control, used on actuators, to high-level control, used with artificial intelligence approaches. One of the most interesting problem that a mobile robot faces is autonomous navigation. For the robot be able to navigate on the environment autonomously, it have to make use of sensory input from one or more sensors that are used to perceive the robot surroundings as well as sensors that measure internal states of the robot. One of the most used control theory methodology is the proportional-integrative-derivative (PID) controller, where its parameters are estimated through a variety of ways, from raw mathematical modeling to the application of hybrid approaches that uses both a mathematical model and metaheuristics such as genetic algorithms. This paper aims to estimate the parameters of the direct current motor through a Kalman filter and use those to estimate the PID parameters to control the DC motor by the usage of a genetic algorithm. Results shows that the derived PID controller is quite efficient on the control of the DC motor used, thus validating the methodology.