Probability Collectives: A multi-agent approach for solving combinatorial optimization problems

Kulkarni, Anand J.; Tai, Kang

doi:10.1016/j.asoc.2009.09.006

Cited by 65 publications

(32 citation statements)

References 49 publications

(88 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…An aerial robot model is a vector which records a probability distribution of the actual character of the modeled aerial robot [38]. Each aerial robot has two basic models about each other aerial robot a.…”

Section: Bayesian Strategymentioning

confidence: 99%

Multi Aerial Robot Planning

Sebbane

2014

Intelligent Systems, Control and Automation: Science and Engineering

View full text Add to dashboard Cite

Multi-robot systems are a major research topic in robotics. Designing, testing and deploying in the real world a large number of aerial robots is a concrete possibility due to the recent technological advances. The first section of this chapter treats the different aspects of cooperation in a multi-agent systems. A cooperative control should be designed in terms of the available feedback information. A cascade-type guidance law is proposed, followed by consensus approach and flocking behavior. Since information flow over the network changes over time, cooperative control must react accordingly but ensure group cooperative behavior which is the major issue in analysis and synthesis. Connectivity and convergence of formations are also studied. Team approach is followed by deterministic decision making. Plans may be required for a team of aerial robots to plan for sensing, plan for action or plan for communication. Distributed receding horizon control as well as conflict resolution, artificial potentials and symbolic planning are thus analyzed. Then, association with limited communications is studied, followed by genetic algorithms and game theory reasoning. Next, multi-agent decision making under uncertainty is considered, formulating the Bayesian decentralized team decision problem, with and without explicit communication. Algorithms for optimal planning are then introduced as well as for task allocation and distributed chance constrained task allocation. Finally, some case studies are presented such as reconnaissance mission that can be defined as the road search problem or the general vehicle routing problem. Then, an approach is considered to coordinate a group of aerial robots without a central supervision, by using only local interactions between the robots. The third case is the optimization of perimeter patrol operation. If an aerial robot must be close from a location to monitor it correctly and the number of aerial robots does not allow covering each site simultaneously, a path planning problem arises. Finally stochastic strategies for surveillance are presented.Y.

show abstract

Section: Bayesian Strategymentioning

confidence: 99%

Multi Aerial Robot Planning

Sebbane

2014

Intelligent Systems, Control and Automation: Science and Engineering

View full text Add to dashboard Cite

show abstract

“…may significantly get affected when applied for solving constrained problems [1][2][3][4][5][6]. The most adopted approach for handling constraints is the penalty function approach which can be referred to as a generalized constraint handling method because of its simplicity and ability to handle nonlinear constraints and it can be used with most of the unconstrained optimization methods [4,7,8].…”

Section: Introductionmentioning

confidence: 99%

Constraint handling in Firefly Algorithm

Deshpande¹,

Phatnani²,

Kulkami³

2013

2013 IEEE International Conference on Cybernetics (CYBCO)

Self Cite

View full text Add to dashboard Cite

Most of the contemporary nature-/bio-inspired techniques are unconstrained algorithms. Their performance may get affected when dealing with the constrained problems. There are number of constraint handling techniques developed for these algorithms. This paper intends to compare the performance of the emerging metaheuristic swarm optimization technique of Firefly Algorithm when incorporated with the generalized constrained handling techniques such as penalty function method, feasibility-based rule and the combination of both, i.e. combined approach. Seven well known test problems have been solved. The results obtained using the three constraint handling techniques are compared and discussed with regard to the robustness, computational cost, rate of convergence, etc. The associated strengths, weaknesses and future research directions are also discussed.

show abstract

“…In comparison, MIMIC directly uses truncated samples and PBIL adopts the idea of GA which simulates the evolution of a population to find the final solution. Besides, PC pays more attention on how individual can influence system's performance with the presence of the other agents than the DRL does and thus is particularly suitable for distributed problems (Kulkarni and Tai 2010a;Kulkarni et al 2010b;Autry and Brian 2008;Busoniu et al 2008).…”

mentioning

confidence: 97%

“…PC is also used to improve the Metropolis-Hastings sampling and mechanism design work (Wolpert and Tumer 1999, 2001, 2002. More widely, it has also been used in some combinatorial problems such as multi-depot multiple traveling salesmen problem (MTSP) (Kulkarni and Tai 2010a;Kulkarni et al 2010b), the singlE−depot MTSP (Kulkarni and Tai 2010a;Kulkarni et al 2010b), the fleet assignment problem (Antoine et al 2004), school table scheduling problems (Autry and Brian 2008) and vehicle routing problems (Kulkarni and Tai 2010a;Kulkarni et al 2010b). …”

mentioning

confidence: 98%

A modified probability collectives optimization algorithm based on trust region method and a new temperature annealing schedule

Yang

2015

Soft Comput

View full text Add to dashboard Cite

This article presents a distributed random search optimization method, the trust region probability collectives (TRPC) method, for unconstrained optimization problems without closed forms. Through analyzing the framework of the original probability collectives (PC) algorithm, three potential requirements on solving the original PC model are first identified. Then an interior point trust region method for bound constrained minimization is adopted to satisfy these requirements. Besides, the temperature annealing schedule is also redesigned to improve the algorithmic performance. Since the new annealing schedule is linked to the gradient, it is much more flexible and efficient than the original one. Ten benchmark functions are used to test the modified algorithm. Numerical results show that TRPC is superior to the PC algorithm in iteration times, accuracy, and robustness.

show abstract

Probability Collectives: A multi-agent approach for solving combinatorial optimization problems

Cited by 65 publications

References 49 publications

Multi Aerial Robot Planning

Multi Aerial Robot Planning

Constraint handling in Firefly Algorithm

A modified probability collectives optimization algorithm based on trust region method and a new temperature annealing schedule

Contact Info

Product

Resources

About