A hybrid technique for path planning of humanoid robot NAO in static and dynamic terrains

The humanoid robot is widely used because of its ability to imitate human actions. The selection of navigational techniques is of prime importance because the quality of the opted technique directly affects the success of output. In this paper, the hybridization of the Dynamic Window Approach (DWA) and the Teaching–Learning-Based Optimization (TLBO) technique and its implementation on the NAO humanoid robot for navigation have been presented. The input is based on the location of obstacles and the target. The parameters are provided to the DWA technique, which decides the optimum velocity. The intermediate result is feed to the TLBO technique, which operates based on the teacher phase and the learner phase. This hybridization provides an optimum angle to take a turn and avoids the obstacles while moving towards the target. The current article concentrates on implementing hybridized techniques in static and dynamic terrains. Single NAO and some random obstacles are chosen for static navigation. For dynamic terrains, multiple NAOs and some static obstacles are considered. In this case, one humanoid robot acts as a dynamic obstacle to another. In the dynamic terrain, there is a possibility of inter-collision amongst NAOs. To avoid inter-collision, a Petri-Net controller has been designed and implemented in all NAOs. Simulation and experimental results on humanoid NAOs demonstrate target attainment with collision-free optimal paths. Experimental and simulated results of the proposed technique present an acceptable relation under 5 % and 6 % for a single robot and multiple robots, respectively. The proposed technique has been compared with previously developed techniques in complex, danger and dynamic terrains. In comparison with previously developed techniques, it is evident that the proposed technique is robust and efficient for the path planning of humanoid robots.