Innovative Production Machines and Systems Conference - Intelligent Optimisation Techniques

Welcome

Afshin — Fri, 06 Jul 2007 10:51:28 +0100

Dear Participants,

Welcome to the Intelligent Optimisation Techniques session of IPROMS 2007 virtual conference. My name is Afshin Ghanbarzadeh, the co-chairman for for this session.

I want to encourage all conference participants and visitors to feel free to download any of the conference papers and to actively take part in the discussions. The authors are eager to get your useful comments. Do not forget to also visit the conference’s blog especially if you want to share you humours side with others.

Authors, may I also use this medium to remind you all to subscribe to Group Notification on the

An EA framework for uncertain optimization problem

mbhattac — Fri, 22 Jun 2007 14:17:48 +0100

Many engineering design projects involve solving complex optimization problems. Evolutionary algorithms (EA) have been widely accepted as efficient optimizers for complex real life problems. However, many real life optimization problems involve time-variant noisy environment, which pose major challenges to EA-based optimization. Presence of noise interferes with the evaluation and the selection process of EA and adversely affects the performance of the algorithm [7]. Also presence of noise means fitness function can not be evaluated and it has to be estimated instead. Several approaches have been tried to overcome this problem, such as introduction of diversity (hyper mutation, random immigrants, special operators) or incorporation of memory of the past (diploidy, case based memory) [8]. In this paper we propose a method, DPGA (distributed population genetic algorithm) that uses a distributed population based architecture to simulate a distributed, self-adaptive memory of the solution space. Local regression is used in each sub-population to estimate the fitness. Specific problem category considered is that of optimization of functions with time variant noisy fitness. Successful applications to benchmark test problems ascertain the proposed methodâ€™s superior performance in terms of both adaptability and accuracy.

Qualitative and quantitative airfoil design optimisation using interactive genetic algorithms

Alexandra Brintrup — Fri, 22 Jun 2007 14:17:48 +0100

This paper introduces the necessity of qualitative views on the evolutionary optimisation of airfoil shape design and applies multi-objective and parallel interactive genetic algorithms to combine qualitative evaluation with quantitative shape optimisation. The interactive evaluation enables user to embed domain specific knowledge which is frequently hard to describe. A comparison of the multi-objective and parallel IGA results is made using the fitness convergence, diversity and user preference performance metrics. Although the multi-objective IGA provides more diverse results the parallel IGA obtains significantly better fitness convergence and user preference.

Distributed and adaptive clustering architecture for dynamic sensor networks

Medhat Awadalla — Fri, 22 Jun 2007 14:17:48 +0100

Clustering is an effective topology control approach in sensor networks. This paper proposes a distributed and adaptive clustering architecture for dynamic sensor networks. The proposed architecture comprises an approach for energy-efficient clustering with adaptive node activity for achieving a good performance in terms of system lifetime and network coverage quality. This architecture demonstrates a uniform cluster head distribution across the network in addition to a desirable network coverage. Furthermore, this paper presents an analytical approach to disclose the relationship between network density and coverage quality.

PLANNING IN MULTI AGENT SYSTEMS BASED ON REINFORCEMENT LEARNING

bastanfard — Fri, 22 Jun 2007 14:17:47 +0100

In this paper, a method for feature selection by reinforcement learning is considered. Reinforcement learning is the problem faced by an agent that learns behavior through trial and error interactions with a dynamic environment. The work described here is a multi agent system in which the agents which try to coordinate to reach common goals, would be learned by considering the ability of the features to separate the sample data in order to discriminate them as well as possible. This discrimination is under Linear Discriminate Analysis; in addition the correlation coefficient does as the weight

Toward the Application of Genetic Algorithms to Real World Resource Constrained Project Scheduling Problems.

John Lancaster — Fri, 22 Jun 2007 14:17:46 +0100

Much research has been invested in the optimisation of the Resource Constrained Project Scheduling Problem (RCPSP) using Genetic Algorithms. Reviews of this work can be found in Lancaster and Ozbayrak [1] and Kolisch and Hartmann [2]. This research however doesnâ€™t extend to the solution of real world RCPSP. As part of ongoing research the authors describe a practical implementation of Genetic Algorithm optimisation within a commercial scheduling package applied to simple real world problems. The paper will show the effectiveness of genetic algorithms when applied to real world RCPSP.