Output details
11 - Computer Science and Informatics
University of Portsmouth
Global optimization based on novel heuristics, low-discrepancy sequences and genetic algorithms
<12>This hybrid heuristic for constrained global optimization (GO) performs initial uniform search of the parameter space and after locating regions of attraction, incorporates adaptive, evolutionary heuristics at the later stages, balancing the method’s exploration/exploitation effort. Tests on a number of benchmark problems and comparison with other optimization techniques demonstrate superior performance of the method. It facilitates the advancements in heuristic GO by providing the research community with a powerful tool for solving complex optimization problems of higher dimensionality. The method was later used in the development of an EPSRC KTN Industrial Mathematics project “Refining hydroprocess modelling used in BP refineries”.