抄録
In this paper, we report a new multi-objective evolutionary algorithm: Archive-based Micro Genetic Algorithm (AMGA), which has developed by Engineous Software. Motivation of the development was mainly driven by two aspects: (1) Recent progress of MOEA studies, (2) Accumulated practice in engineering optimization. In the research field of MOEA(Multi Objective Evolutionary Algorithm), a number of excellent algorithms have been arising year by year since the beginning of 2000. And significant amount of application examples have been accumulated so far. This has been enabled positive cycle between algorithm development and filed application, and AMGA is one of such results. In this paper, beginning with descriptions of algorithmic features which characterize AMGA form other MOEAs, we show performance of proposed algorithm through benchmark problems both in mathematical test functions and in engineering problems, and discuss effectiveness of the algorithm.
