Variance modulated task prioritization in Whole-Body Control

Abstract

Abstract-Whole-Body Control methods offer the potential to execute several tasks on highly redundant robots, such as humanoids. Unfortunately, task combinations often result in incompatibilities which generate undesirable behaviors. Prioritization techniques can prevent tasks from perturbing one another but often to the detriment of the lower precedence tasks. For many tasks, static prioritization is not necessary or even appropriate because tasks can often be achieved in variable ways, as in reaching. In this paper, we show that such task variability can be used to modulate task priorities during execution, to temporarily deviate certain tasks as needed, in the presence of incompatibilities. We first present a method for mapping from task variance to task priority and then provide an approach for computing task variance. Through three common conflict scenarios, we demonstrate that mapping from task variance to priorities reactively solves a number of task incompatibilities.

Learning Robust Task Priorities and Gains for Control of Redundant Robots

“…A derivative-free stochastic optimization is used to learn the optimal parameters. Lober et al [8] propose a framework that modifies a set of Fig. 1: Despite being tuned to perform several motions in simulation, a hand-tuned controller can easily fail when transferred on the real humanoid and when performing different tasks that challenge its balance.…”

Probabilistic Learning of Torque Controllers from Kinematic and Force Constraints

“…configuration and operational spaces) into a common space, that of torque commands. With the definition of Γ (p) in (13), torque commands can be combined using (7). The problem of learning control commands and their respective importance is thus framed as the learning of reference trajectories as Gaussian distributions N (µ (p) , Σ (p) ), and generating Gaussian-distributed torque commands N (τ (p) , Σ (p) τ ), which encapsulate the control reference and its importance with respect to other controllers.…”

Optimization-Based Control Approaches to Humanoid Balancing

“…Recently, the work in [41,42] has questioned the notion of priorities: while some priorities are needed for robustness and safety reasons, trying to achieve incompatible tasks at the reactive level does not make sense. Thus, on-line task optimization is needed in order to feed reactive whole-body controllers with references which are compatible one with another and more importantly with constraints, among which balance is a very important one.…”