Parameters optimization of continuous casting process using teaching-learning-based optimization algorithm

Abstract

In the present work, continuous casting process is considered for its parameters optimization using a recently developed optimization algorithm known as teaching-learning-based optimization algorithm. The example considered is a multi-objective multi-constrained problem having three objectives and two constraints. The objectives of the process under consideration are maximization of lubrication index and minimization of peak friction and oscillation marks. Four process parameters are involved in the process namely casting speed, stroke, frequency and deviation from sinusoid. The same model was earlier attempted by researchers using genetic algorithm and differential evolution techniques. The results obtained in the present work outperformed the previous results in terms of fitness functions and number of generations. © 2012 Springer-Verlag.