Implementation of intelligent navigational techniques for inter-collision avoidance of multiple humanoid robots in complex environment

Over the past few decades, research in humanoid robots has been amplified rapidly. This paper displays the attainment of an optimum steering angle to avoid hurdles and reach the target with minimum effort. To achieve this objective, three steps optimization procedure is considered. A hybridization of regression analysis (RA), cell decomposition (CD), and whale optimization algorithm (WOA) are designed and implemented in humanoid NAO for prime trajectory with the least computational cost. RA supplies the first set of steering angles to CD based on the training in the given workspace. CD provides a second set of steering angles to WOA, which results in an optimum steering angle based on its characteristics of hunting prey. The proposed RA-CD-WOA algorithm is evaluated in simulated and experimental workspaces for a single NAO. The proposed algorithm and its standalone algorithms are compared for several iterations, demonstrating the requirement for hybridization. It is also examined for multiple NAOs on a common platform that may lead to a deadlock condition during navigation. To elude this condition, the dining philosopher controller is integrated with the base algorithm, which results in prioritizing a NAO and solving the problem. Further, the RA-CD-WOA algorithm is compared with an existing technique that displays its robustness and effectiveness for robot navigation.