Prototype design and performance analysis of genetic algorithm-based SLAM for indoor navigation using TETRIX Prizm mobile robot

Abstract

For any robot to navigate on its own, it is important to create a map of its surroundings and localise itself to detect the obstacles. The present paper focuses on building a universal robotic description format model for the TETRIX robot followed by optimisation of 2D simultaneous localisation and mapping parameters using genetic algorithm in ROS-based environment to make robot navigation efficient. The GMapping and adaptive Monte Carlo localisation parameters were optimised using multiobjective GA. Furthermore, map quality is validated by using no reference blind/referenceless image spatial quality evaluator factor. The statistical analysis based on the mean and standard deviation of particle filters was analysed to measure the spread the particles filters during localisation. The test environment was created in ROS, Gazebo and MATLAB and with optimal parameters good quality high-resolution map with accurate and efficient localisation was achieved.