A Language for Platform Independent Communication and Storage in Multiobjective Optimisation

Abstract

In real-world multiobjective optimisation, where problem complexity is typically high, systems are increasingly reliant on distributed computation to reduce processing times. While a number of successful distributed techniques have been proposed, they are dependent on high degrees of commonality between communicating processes - requiring homogeneity with respect to platform, environment or implementation language. Such constraints seriously restrict the applicability of distributed optimisers. Consequently, it is important to define a standard independent language for the transmission and storage of agent data in disparate interacting systems. Based on XML, MOO_ML represents such a language.