TakehikoNakama¹

CMC-Computers, Materials & Continua, Vol.14, No.1, pp. 35-60, 2009, DOI:10.3970/cmc.2009.014.035

Abstract Random noise perturbs objective functions in practical optimization problems, and genetic algorithms (GAs) have been proposed as an effective optimization tool for dealing with noisy objective functions. In this paper, we investigate GAs in a variety of noisy environments where fitness perturbation can occur in any form-for example, fitness evaluations can be concurrently disturbed by additive and multiplicative noise. We reveal the convergence properties of GAs by constructing and analyzing a Markov chain that explicitly models the evolution of the algorithms in noisy environments. We compute the one-step transition probabilities of the Markov chain and show that the chain has… More >

D. S. Liu¹, C. Y. Tsai¹, S. R. Lyu²

CMC-Computers, Materials & Continua, Vol.11, No.2, pp. 147-164, 2009, DOI:10.3970/cmc.2009.011.147

Abstract This study presents a novel numerical method for extracting the tempe -rature-dependent mechanical properties of the gold and aluminum thin-films. In the proposed approach, molecular dynamics (MD) simulations are performed to establish the load-displacement response of the thin substrate nanoindented at temperatures ranging from 300-900 K. A simple but effective procedure involving genetic algorithm (GA) and finite element method (FEM) is implemented to extract the material constants of the gold and aluminum substrates. The material constants are then used to construct the corresponding stress-strain curve, from which the elastic modulus, yield stress and the tangent modulus of the thin film… More >

CMC-Computers, Materials & Continua, Vol.10, No.1, pp. 65-96, 2009, DOI:10.3970/cmc.2009.010.065

Abstract This work proposed an alternative formulation for the prediction of martensite start temperature (Ms) of Fe-Mn-Si shape memory alloys (SMAs) depending on the various compositions and heat treatment techniques by using Neural Network (NN) and genetic programming (GP) soft computing techniques. The training and testing patterns of the proposed NN and GP formulations are based on well established experimental results from the literature. The NN and GP based formulation results are compared with experimental results and found to be quite reliable with a very high correlation (R2=0.955 for GEP and 0.999 for NN). More >