Machining condition optimization by genetic algorithms and simulated annealing

Optimal machining conditions are the key to economical machining operations. In this work, some benchmark machining models are evaluated for optimal machining conditions. These machining models are complex because of non-linearities and non-convexity. In this research, we have used Genetic Algorithms and Simulated Annealing as optimization methods for solving the benchmark models. An extension of the Simulated Annealing algorithm, Continuous Simulated Annealing, is also used. The results are evaluated and compared with each other as well as with previously published results which used gradient based methods, such as SUMT (Sequential Unconstrained Minimization Technique), Box's Complex Search, Hill Algorithm (Sequential Search Technique), GRG (Generalized Reduced Gradient), etc. We conclude that Genetic Algorithms, Simulated Annealing and the Continuous Simulated Annealing which are non-gradient based optimization techniques are reliable and accurate for solving machining optimization problems and offer certain advantages over gradient based methods.