Exploiting molecular dynamics for multi-objective optimization

Gas molecules within an enclosure will always tend to a homogenous and uniform equilibrium with maximum entropy, even without any prior knowledge on the geometry and state of the enclosure. Furthermore, if an uneven potential field was present, more molecules will tend to reside in the lower potential region as dictated by the Maxwell–Boltzmann distribution. The inherent diverse behavior of molecular system and their converging drift pressure in potential fields seems to be applicable for the contrary goals of proximity and diversity in multi-objective optimization. Inspired by this association, this paper will explore the notion of exploiting molecular motion to solve multi-objective problems. By adapting the algorithmic structure of molecular dynamics, which essentially represents a technique for the computer simulation of molecular motion, a molecular system that is relevant for multi-objective optimization is proposed, known as molecular dynamics optimizer (MDO). The performance of MDO was compared with other conventional multi-objective optimizers, specifically EA and PSO, in several multi-objective benchmark problems and the experimental results demonstrated that MDO is indeed a viable and practical approach for multi-objective optimization.