2013/05/14 by Jie Sun, Adilson E. Motter
Computer Science · Mathematics · Physics and Astronomy · #Neural Networks Stability and Synchronization #Nonlinear Dynamics and Pattern Formation #Quantum optics and atomic interactions #cond-mat.dis-nn #math.OC #nlin.AO
paper · pdf · doi:10.1103/physrevlett.110.208701
published as Phys. Rev. Lett. 110, 208701 (2013)
openalex publication_date 2013/05/14 · arxiv created 2013/05/24 · arxiv updated 2013/05/28 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
A common goal in the control of a large network is to minimize the number of driver nodes or control inputs. Yet, the physical determination of control signals and the properties of the resulting control trajectories remain widely underexplored. Here we show that (i) numerical control fails in practice even for linear systems if the controllability Gramian is ill conditioned, which occurs frequently even when existing controllability criteria are satisfied unambiguously, (ii) the control trajectories are generally nonlocal in the phase space, and their lengths are strongly anti-correlated with the numerical success rate and number of control inputs, and (iii) numerical success rate increases abruptly from zero to nearly one as the number of control inputs is increased, a transformation we term numerical controllability transition. This reveals a trade-off between nonlocality of the control trajectory in the phase space and nonlocality of the control inputs in the network itself. The failure of numerical control cannot be overcome in general by merely increasing numerical precision--successful control requires instead increasing the number of control inputs beyond the numerical controllability transition.