Invariance and Robustness in interconnected systems: Coordination from multi-agents to living microorganism.

• Alejandro Vignoni ^[1] ; Jesús Picó ^[1]
  1. [1] Universidad Politécnica de Valencia

     Universidad Politécnica de Valencia

     Valencia, España

     
• Localización: XI SIMPOSIO CEA DE INGENIERÍA DE CONTROL: Automática y Biología celular: una combinación emergente, 2013, ISBN 978 84 695 7298 6, pág. 109
• Idioma: inglés
• Enlaces
  • Texto Completo Libro (pdf)
• Resumen

  • This work is focused in the study of coordination of interconnected systems using invariance and robustness properties. Coordinationis understood as driving the systems involved in order to obtain some desired collective behavior. This can be achievedeither in a centralized way or in a distributed one along the systems, always taking into account the systems constraints and capabilities.Starting from set invariance principles, we develop several methodologies to deal with the coordination problem. Thishas been done from di erent perspectives depending on the class of systems we are working with. On the one hand, working withmulti-agent systems, we develop a methodology based in reference conditioning with sliding modes. The approach addresses theproblem of coordinating systems with possibly di erent dynamics (e.g. linear and nonlinear, di erent orders, constraints, etc.). Theidea behind our approach to the coordination problem is that in order to coordinate the systems, we can shape their local referencesas function of the achievable performance of each system and information each system has about their neighbors in the group.On the other hand, but starting also from invariance principles, we develop a strategy to make a set invariance in a group of generegulatory (GR) circuits in synthetic microorganism. This kind of systems is of a di erent nature than the previous one, and isnot very often treated in control theory, thus new methodologies are required. The strategy involves invariance ideas, and a sort ofback-stepping technique to make a group of interconnected non-linear GR circuit to work in a linear regime in a coordinated way.This methodology is used then to design GR circuits and to obtain coordinated behaviors at the population level, and all in all, toimprove the performance of the system and its robustness with respect to intrinsic and extrinsic noise.