Navigation and Control of Humanoid Robots in Unknown Environment Using Smart Methodologies

Abstract

The current research work focuses on design, development and implementation of effective motion planning strategies in humanoid robots for collision free and obstacle avoidance path traversal between two locations with high accuracy of task completion. Here, kinematic analysis of NAO H25 with trajectory generation is considered to perform analysis on motion control parameters of humanoid. Various navigational approaches like Bacterial Foraging Optimization Algorithm, Grey Wolf Optimization Algorithm, novel DAY-DEMOCRATIC technique have been designed with mathematical formulation for navigational analysis of single and multiple humanoid robots. For effective motion planning of multiple humanoids in complex environments, a Petri-Net architecture is designed and combined with the developed methods to avoid inter-collision among the humanoids. Using fuzzy logic system with Rule Based Technique, Prim’s Algorithm, and DAY-DEMOCRATIC technique, hybridized controllers are developed to improve the motion planning strategies with enhancement in navigational parameters like trajectory distance and trajectory time. Both simulation and experimental analysis has been performed by the humanoid using the developed controllers and the results obtained from both the analysis are validated. An analysis of the results from simulation and experiment demonstrates strong agreement with each other and a reasonable tolerance for error. The generated controllers are compared to existing methodologies to validate the findings and determine their efficacy, and significant advancements in respect to navigational parameters are seen. Finally, application areas with future extension of current research are highlighted.