Forwardstepping: A New Approach for Control of Dynamical Systems
Hanlei Wang, Wei Ren, Chien Chern Cheah · 2019
In this paper, we mainly focus on developing a new paradigm motivated by investigating the consensus problem of networked Lagrangian systems with unknown time-varying delay and switching topologies. We propose a class of dynamic feedback by adding differentiators to accommodate the discontinuity and uncertainty resulting from time-varying delay and switching topologies, yielding arbitrary times differentiable reference velocities. Using these reference velocities, we develop delay/topology-independent adaptive controllers with arbitrary times differentiable control torques for networked Lagrangian systems. This specific practice motivates the formulation of a new paradigm, referred to as forwardstepping, for handling discontinuity, time-varying uncertainty, and unavailable state measurement in control systems.