On the convergence of puck clustering systems

Kazadi, Sanza; Abdul-Khaliq, A.; Goodman, R.M.

doi:10.1016/s0921-8890(01)00171-3

Cited by 39 publications

(30 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Therefore, macroscopic models offer a direct description of the collective group behavior and are computationally more efficient than their microscopic counterparts since, even if they must be solved numerically, their computation time is independent of the number of agents in the system. Although for a long period early in collective robotics research there was relatively little work in modeling of multi-robot systems, recently physicists and engineers have dedicated more attention to this problem (see, for example, related work performed by Kazadi et al [17], Lerman and Galstyan [23], and Sugawara and Sano [36,37]). Moreover, modeling methodologies for swarm robotics systems must take into account mobility, individual intelligence, intrinsic stochastic properties of the collective coordination based on SI-principles, and, potentially, several different modalities of interaction among individuals and between an individual and the environment (e.g., mechanical, electromagnetic, chemical).…”

Section: Introductionmentioning

confidence: 99%

Macroscopic Modeling of Aggregation Experiments using Embodied Agents in Teams of Constant and Time-Varying Sizes

Agassounon¹,

Martinoli

Easton³

2004

Autonomous Robots

View full text Add to dashboard Cite

In this paper, we present discrete-time, nonspatial, macroscopic models able to capture the dynamics of collective aggregation experiments using groups of embodied agents endowed with reactive controllers. The strength of the proposed models is that they have been built up incrementally, with matching between models and embodied simulations verified at each step as new complexity was added. Precise heuristic criteria based on geometrical considerations and systematic tests with one or two embodied agents prevent the introduction of free parameters into the models. The collective aggregation experiments presented in this paper are concerned with the gathering and clustering of small objects initially scattered in an enclosed arena. Experiments were carried out with teams consisting of one to ten individuals, using groups of both constant and time-varying sizes. In the latter case, the number of active workers was controlled by a simple, fully distributed, thresholdbased algorithm whose aim was to allocate an appropriate number of individuals to a time-evolving aggregation demand. To this purpose, agents exclusively used their local perception to estimate the availability of work. Results show that models can deliver both qualitatively and quantitatively correct predictions and they represent a useful tool for generalizing the dynamics of these highly stochastic, asynchronous, nonlinear systems, often outperforming intuitive reasoning. Finally, in addition to discussions of small prediction discrepancies and difficulties in generating quantitatively correct macroscopic models, we conclude the paper by reviewing the intrinsic limitations of the current modeling methodology and by proposing a few suggestions for future work.

show abstract

Section: Introductionmentioning

confidence: 99%

Macroscopic Modeling of Aggregation Experiments using Embodied Agents in Teams of Constant and Time-Varying Sizes

Agassounon¹,

Martinoli

Easton³

2004

Autonomous Robots

View full text Add to dashboard Cite

show abstract

“…The global probabilistic behavior emerges based on the geometrical shapes of the robots and pucks, and the individual behaviors provided by neural-based controllers. More detailed theoretical derivations related to this subject can be found in [24].…”

Section: Methodsmentioning

confidence: 99%

Supervised Learning in Robotic Swarms: From Training Samples to Emergent Behavior

Vorobyev

Vardy

Banzhaf

2014

Springer Tracts in Advanced Robotics

View full text Add to dashboard Cite

Emergent behavior in swarm robotic systems is key to obtaining complex behavior by a group of relatively simple agents. The question is how to design the individual behaviors of agents in such a way that the desired global behavior emerges. Different approaches have been proposed to solve this problem: from biologically inspired probabilistic behavioral models to evolutionary techniques. In some situations, however, creating a complex probabilistic model of the behavior or developing a proper setup for an evolutionary process can be challenging. In this paper we propose a new method, based on supervised learning on a relatively small number of training samples. We apply our method to the well-known clustering problem and show that this approach yields the desired global clustering behavior.

show abstract

“…Nestes trabalhos foram realizados estudos com grupos mistos de robôs divididos em robôs que possuíam algum tipo de comunicação centralizada e robôs que baseavam suas decisões apenas em observações locais. Em Kazadi et al (2002) foram estudadas as propriedades gerais de agregação em Sistemas Multi-Robô usando modelos fenomenológicos macroscópicos.…”

Section: Trabalhos Relacionadosunclassified

Algoritmo distribuído para atribuição dinâmica de tarefas em um enxame de robôs

Mendonça¹,

Nedjah

Mourelle

2013

Proceeding Series of the Brazilian Society of Computational and Applied Mathematics

View full text Add to dashboard Cite

Abstract-This paper proposes a distributed control algorithm to implement dynamic task allocation in a swarm robotics environment. In this context, each robot that integrates the swarm must run the algorithm periodically in order to control the underlying actions and decisions. The algorithm was implemented and extensively tested. The corresponding performance and effectiveness are promising.Keywords-Dynamic task allocation, Swarm robotics, Swarm intelligence, Multi-Robot Systems.Resumo-Este artigo propõe um algoritmo de controle distribuído para a alocação dinâmica de tarefas em um ambiente de robótica de enxame. Neste contexto, cada robô que integra o enxame deve executar este algoritmo periodicamente para controlar suas ações e decisões. O algoritmo foi implementado e amplamente testado. Assim seu desempenho e eficácia são promissores.Palavras-chave-Alocação dinâmica de tarefas, Robótica de enxame, Inteligência de enxame, Sistemas multirobôs 1 Introdução A Inteligência de Enxame foi proposta no final de 1980 através da observação do comportamento social de espécies de insetos e pássaros mostrada em Bonabeau et al. (1999). Estas espécies realizam tarefas simples coletivamente com o intuito de alcançar objetivos complexos e impossíveis de serem realizados por apenas um indiví-duo do enxame. A ideia central deste comportamento coletivoé a de executar uma tarefa complexa dividindo-a em tarefas simples que são facilmente executadas pelos membros do enxame. A abordagem da robótica de enxame surgiu a partir desta observação estendida a grupos de robôs mó-veis com capacidade de processamento limitada.A alocação dinâmica de tarefasé um processo necessário para o bom gerenciamento do enxame. Este permite distribuir as tarefas, a serem realizadas, entre os robôs do enxame de tal maneira que uma proporção pré-definida de execução de tarefas seja respeitada. A proporção deveria ser determinada no sentido de permitir o alcance do objetivo global do enxame. A alocação de tarefasé um processo dinâmico pois precisa ser continuamente ajustada em respostaàs alterações no ambiente e/ou no desempenho do enxame. Uma solução imediata para resolver este problema baseia-se na abordagem centralizada. No entanto, uma alocacão distribuída representa uma melhor aproximacão do comportamento dos enxames de espécies sociais, onde não existe um mecanismo de controle centralizado. Portanto, a alocação de tarefas em enxames de robôs deve surgir como resultado de um processo distribuído. Esta descentralização aumenta a complexidade do problema, pois o robô não tem uma visão completa do ambiente. Cada robô deverá tomar decisões de controle locais sem o conhecimento completo do que outros robôs fizeram no passado, estão fazendo agora ou irão fazer no futuro.Diversas aplicações que são automatizadas usando a robótica de enxame, requerem uma alocação dinâmica de tarefas. Por exemplo, em situações que representem um risco ou inviabilizem a presença humana para a execução de uma tarefa, o enxame de robôs seria capaz de se auto-organizar para formar g...

show abstract

On the convergence of puck clustering systems

Cited by 39 publications

References 10 publications

Macroscopic Modeling of Aggregation Experiments using Embodied Agents in Teams of Constant and Time-Varying Sizes

Macroscopic Modeling of Aggregation Experiments using Embodied Agents in Teams of Constant and Time-Varying Sizes

Supervised Learning in Robotic Swarms: From Training Samples to Emergent Behavior

Algoritmo distribuído para atribuição dinâmica de tarefas em um enxame de robôs

Contact Info

Product

Resources

About