摘要:
In this talk we will focus on formability of multi-agent control approach based on Gaussian potential functions, each parametrized by an agent-related control parameter. This results in different characteristics of attractive and repulsive forces among agents and targets, as well as among agents themselves and dependence of the formation’s potential structure on a change of agent’s configurations. The undesired stable equilibria could be eliminated by mutual interaction of the agents and targets, i.e., by changing the characteristics of elementary potential functions. As a possible solution we propose an adaptive algorithm of the agent-related control parameter.
报告人简介:
Ivana Palunko is a postdoctoral researcher at the Department of Control and Computer Engineering, University of Zagreb, Faculty of Electrical Engineering and Computing (FER). She graduated from FER with a Masters degree (Dipl.-Ing.) in Electrical Engineering in 2007., majoring in Control Systems. In August 2008, she enrolled in the Ph.D program, the Control Systems major, at the Department of Electrical and Computer Engineering, University of New Mexico (UNM), Albuquerque, NM where she defended her Ph.D dissertation in August 2012. The tools utilized in her research are: modeling of nonlinear systems, nonlinear control, Lyapunov stability, adaptive control, bifurcation theory, optimal control, reinforcement learning, optimization theory, trust-based decentralized control.