This chapter presents a new parallel multiobjective evolutionary algorithm, based on the island model, where the objective space is exploited to distribute the individuals among the processors. The algorithm, which generalizes the well-known cone separation method, mitigates most of its drawbacks. The new algorithm has been employed to speed-up the optimization of fuzzy rule-based classifiers. The fuzzy classifiers are used to build an emulator of the Ship Classification Unit (SCU) contained in modern influence mines. Having an accurate emulator of a mine’s SCU is helpful when needing: (i) to accurately evaluate the risk of traversal of a mined region by vessels/AUVs, (ii) to assess the improvements of ship signature balancing processes, and (iii) to support in-vehicle decision making in autonomous unmanned mine disposal.
CITATION STYLE
Cococcioni, M. (2015). Mspmea: The microcones separation parallel multiobjective evolutionary algorithm and its application to fuzzy rule-based ship classification. Studies in Computational Intelligence, 621, 445–465. https://doi.org/10.1007/978-3-319-26450-9_17
Mendeley helps you to discover research relevant for your work.