Evolving a Multiagent Controller for Micro Aerial Vehicles

Abstract

Abstract-Micro Aerial Vehicles (MAVs) are notoriously difficult to control as they are light, susceptible to minor fluctuations in the environment, and obey highly non-linear dynamics. Indeed, traditional control methods, particularly those relying on difficult to obtain models of the interaction between an MAV and its environment have been unable to provide adequate control beyond simple maneuvers. In this paper, we address the problem of controlling an MAV (which has segmented control surfaces) by evolving a neurocontroller and fine-tuning it using multiagent coordination techniques. This approach is based on a control strategy that learns to map MAV states (position, velocity) to MAV actions (e.g., actuator position) to achieve good performance (e.g., flight time) by maximizing an objective function. The main difficulty with this approach is defining the objective functions at the MAV level that allow good performance. In addition, to provide added robustness, we investigate a multiagent approach to control where each control surface aims to optimize a local objective. Our results show that this approach not only provides good MAV control, but provides robustness to (i) wind gusts by a factor of six; (ii) turbulence by a factor of four; and (iii) hardware failures by a factor of eight over a traditional control method.

Observer-based leader-following formation control of uncertain multiple agents by integral sliding mode

“…The technology of multi-agent systems has been praised as a novel paradigm for designing and implementing many real systems, such as distributed sensor networks [4], robots [5][6][7][8], satellite flying [9] and aerial vehicles [10]. In reality, a multi-agent system has some advantages over a single agent, including but not limited to efficiency, reliability, extensibility, robustness and flexibility [11,12].…”

“…A cooperative task can be achieved by the interactions between these followers and the leader. The scheme in [4,6,9,10] has been successfully applied to the anal- Uncertainties exist everywhere. Each individual agent may suffer from uncertainties, i.e., external disturbances, unmodelled dynamics and parameter perturbations.…”

“…However, the classic leader-following scheme is centralized, meaning that the control systems in [4,6,9,10] heavily depend on the leader and suffer from the "single point of failure'' problem. With the development of communication technology, it is desired to consider the communication topology in the scheme because the adaptability and practicability of multiple agents can be strengthened [15,16].…”

On The Control Of Opinion In Social Networks

“…Instead individuals seeking to influence the network must interact with no special privileges (beyond collusion within a centralized control system) over the natural diffusion model. Going further, multiple independent distributed controllers (using, for example, multi-agent neurocontrollers [22,106]) more closely model the problem of social network control via grass roots mobilization. These may operate in isolation or with communication depending on the objectives to be considered.…”

Modelo De Neuro Co-Evolução Com Inspiração Quântica Aplicado a Problemas De Coordenação

“…Dá-se o nome de Sistemas Multi-Agente (SMA) à área que estuda a interação entre os agentes em um ambiente, podendo ser encontrados na literatura estudos nas mais diversas aplicações como diagnósticos (Davidson et al, 2006), gerenciamento de energias renováveis (Abdessalem & Karaa, 2014), monitoramento de sistemas (Mcarthur, Strachan & Jahn, 2004), restauração de sistemas elétricos (Nagata & Sasaki, 2002), automação de sub-estações (Buse et al, 2003), controle de veículos aéreos (Salichon & Tumer, 2012) , controle de tráfego (Purohit & Mantri, 2013) , leilão de energia (Correa, Vellasco & Figueiredo, 2013) e futebol de robôs (Mendoza, Vellasco & Figueiredo, 2014).…”