Collaborative Persistent Excitation in RKHS Embedded Adaptive Estimation with Consensus
Jia Guo, Fumin Zhang, Andrew J. Kurdila · 2022 American Control Conference (ACC) · 2022
In this paper, we extend the adaptive consensus estimation scheme proposed in [1], [2] and study the functional parameter convergence in reproducing kernel Hilbert spaces (RKHS). Inspired by the collaborative persistence of excitation condition in [3], we propose a collaborative PE (C-PE) condition which which can be fulfilled by the multiagent team and guarantees functional parameter convergence of adaptive consensus estimation in RKHS. We first derive a necessary condition of the C-PE which relates the collective trajectories of agents and the PE subspace. Then we prove the pointwise function convergence given the C-PE condition. A numerical example is presented to illustrate the results.