KEYWORDSAbstract. With the ability to mimic human behaviour, humanoid robots have become a topic of major interest among research fellows dealing with robotic investigation. The current work is focused on the design of a novel navigation controller based on the logic of the regression analysis to be used in the path planning and navigation of humanoid robots. The current investigation focuses on static and dynamic path planning of humanoid NAOs. The static path planning represents a single NAO navigating through random static obstacles. The dynamic path planning represents multiple humanoid NAOs navigating through random static obstacles and acting as dynamic obstacles for each other. A Petri-net controller is designed to avoid the collision among multiple NAOs in dynamic path planning. To reduce path length and time travel and provide the shortest possible path, an advanced regression controller is implemented in the NAOs in both simulation and experimental environments. Finally, a comparison has been performed between the simulation and experimental results, and good agreement is observed between both of the results with a minimal percentage of error. The proposed navigation controller is also tested against other existing navigational technologies to validate better e ciency.
SummaryThe present paper discusses on development and implementation of back-propagation neural network integrated modified DAYANI method for path control of a two-wheeled self-balancing robot in an obstacle cluttered environment. A five-layered back-propagation neural network has been instigated to find out the intensity of various weight factors considering seven navigational parameters as obtained from the modified DAYANI method. The intensity of weight factors is found out using the neural technique with input parameters such as number of visible intersecting obstacles along the goal direction, minimum visible front obstacle distances as obtained from the sensors, minimum left side obstacle distance within the visible left side range of the robot, average of left side obstacle distances, minimum right side obstacle distance within the visible right side range of the robot, average of right side obstacle distances and goal distance from the robot’s probable next position. Comparison between simulation and experimental exercises is carried out for verifying the robustness of the proposed controller. Also, the authenticity of the proposed controller is verified through a comparative analysis between the results obtained by other existing techniques with the current technique in an exactly similar test scenario and an enhancement of the results is witnessed.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.