Dynamically transitioning between surfaces of varying inclinations to achieve uneven-terrain walking

Abstract

This paper focuses on how to generate dynamic transitions in order to make our robot COMAN (COmpliant huMANoid) dynamically traverse inclined terrains. The novel approach addresses dynamic walking on inclined surfaces by dividing the walking motion into two phases: transition and incline walking. During the transition phase, the humanoid robot performs a 3-dimensional movement in order to transfer its body between surfaces of different inclinations, which is then followed by the incline-walking phase. The transition phase is less trivial to execute than the incline walking itself. In this paper, we first formulate the equations of a 3D (non linear) Inverted Pendulum, and then we derive an equivalent model. Subsequently, we introduce a trajectory generator based on this model and validate it experimentally by performing, with COMAN, dynamic transitions from the horizontal ground to a 10 • slope.

Improving anthropomorphic robot stability using advanced intelligent control interfaces

“…In fact, there are considered movements on inclined terrain [4], [7], [12], [13], [14], [16], [24], [25] which by means of an explicit trajectory planning [8], [10], [13] with a strategy based on the ZMP criterion [4], [8], [10], [12], [13], [16], [23], [25] or CPG [24], where a careful analysis of the robot's step [4], [8], [10], [13], [14], [16], [23], [24], [25] and in which the inertia [4], [7], [8], [14], [16], [25] of the various components of the robot have not been avoided [15]. The control methods are applied to the CoM center of mass [4], [7], [8], [10], [12], [13], [16], [24], [25] or to the CoG center of gravity [10], [14], the difference between the positions of t...…”

“…The control methods are applied to the CoM center of mass [4], [7], [8], [10], [12], [13], [16], [24], [25] or to the CoG center of gravity [10], [14], the difference between the positions of the two points are not significant in an uniform gravitational field. The control methods are based on the model of the linear inverted pendulum LIP [4], [7], [8], [13], [23], [25] and use neural networks [6], [16], [24], fuzzy logic [7], [12] predictive [8], balance [4], [7], [8], [10], [12], [13], [16], [24] or energy efficiency calculation [14], [16], [26]. The proposed theoretical results were simulated in the virtual environment [12], [16], [24], [25] with the Webots simulation software [13], [16] or with the application Choregraphe [23] and tested in the real environment on the Nao robot [8], [13], [16], [23],…”

“…The control methods are based on the model of the linear inverted pendulum LIP [4], [7], [8], [13], [23], [25] and use neural networks [6], [16], [24], fuzzy logic [7], [12] predictive [8], balance [4], [7], [8], [10], [12], [13], [16], [24] or energy efficiency calculation [14], [16], [26]. The proposed theoretical results were simulated in the virtual environment [12], [16], [24], [25] with the Webots simulation software [13], [16] or with the application Choregraphe [23] and tested in the real environment on the Nao robot [8], [13], [16], [23], [25] or other models of walking robots [4], [7], [10], [12], [24]. Stability control was carried out either online [10], [12] or offline [1], [25].…”

“…The Ishikawa diagram of AWR stability (Figure 1) integrates the components resulting from the analysis of AWR movement on uneven, unstructured and obstructed terrain, to which two first-rank components are added, i.e. the study of the static balance by CoM trajectory [4], [7], [8], [10], [12], [13], [16], [24], [25] and the investigation of the dynamic balance by the trajectory of ZMP [4], [8], [10], [12], [13], [16], [23], [25], as our synthesis above reveals. It is also demonstrated that high static stability ensures a slow shift, but cannot meet a high-speed requirement.…”

“…The dynamic stability of the anthropomorphic robotic platform balance is ensured in the conditions of the projection of the ZMP in the support range defined by the AWR legs [11], [13] and is more difficult to maintain than in the case of multiple legs robots. The errors generated by the AWR modeling, by theoretical calculation reported to the effects and working environment in the control of walking on a horizontal or inclined plane, were included in the Ishikawa diagram based on strategies, methods and algorithms, either by study in plan horizontally and then extrapolate on the slope, or through the reverse linear inverted method pendulum (LIMP) [4], [7], [8], [13], [23], [25], dual-length linear inverted method pendulum (DLLIMP), virtual support point (VSP), energy efficiency [14], [16], [26] dynamic model adopted, sensory analysis of motion robot on unstructured and uneven terrain [11], [17], [18], [19], [21], [22]. The proposed method offers a stable walk with high speed for the NAO robot only in the conditions of a relatively straight route, with no sharp changes in direction.…”