Stabilized Walking of Humanoid NAO Using Enhanced Spring-Loaded Inverted Pendulum Model on Uneven Terrain

Introduction

The structural resemblance between humanoid robots and humans enables them eligible in areas approved for human occupation for activities like providing services or assembly. In everyday life in factories, restaurants, hospitals, etc., robots are satisfying people and render them as a special companion of people. Humanoid robots are favored among robots because they can mimic human behavior in different environments to control the workforce. Various studies for humanoid robots are underway, but one of the issues among researchers is their stabilization on uneven surfaces and obstacle avoidance. Several papers are studied here concerning the stabilization and obstacle avoidance using different methods:

Navigation of robots is a trending topic, and various studies have been done on different types of robots for the environment with different complexity (Deepak et al., 2011; Kashyap & Pandey, 2018, 2020; Kashyap et al., 2018; Kumar et al., 2018; Lagaza et al., 2020; Pandey et al., 2019). Walking of humanoid robots on flat terrain is a complicated task, but it becomes more complicated while dealing with uneven terrain (Mandava & Vundavilli, 2018). In many connexions, the authors have used the error prediction to identify the best approach to outline the path. They have multiple degrees of freedom, and an underacted system makes it hard to control. Their dynamics (Kashyap, Parhi, & Kumar, 2020) should therefore be recognized, and their analysis should be simplified by different simplified models, i.e., 3D LIPM (linear inverted pendulum model) (Kashyap, Pandey, et al., 2020), LIPM plus flywheel (Kasaei et al., 2017) and SLIP (Wensing & Orin, 2014), etc. must be considered. Apart from improved mechanical devices, such as intuitive robotic interactions (Arumbakkam et al., 2010) with humans, stable locomotion over the uneven surface (Morisawa et al., 2011), and a connection between recovering from environmental disruptions, the operation of such devices still proceeds. In order to achieve a normal pattern (limiter cycle) of the center of the mass, similar to the movement and frictional forces for the existing technology, for example, in human walking is possible in view of the initial state and suitable model parameters. Rutschmann et al. (2012) have applied diffraction data control scheme in order to design SLIP routes for unstable field footholds. The 2D stepping controller is based on a SLIP bipedal has been created by authors (Garofalo et al., 2012).