Rotar, Corina; Dumitrescu, Dan; Lung, Rodica Optimization using an evolutionary hyperplane guided approach. (English) Zbl 1183.90369 Acta Univ. Apulensis, Math. Inform. 12, 49-62 (2006). A new search strategy for multiobjective evolutionary algorithms is suggested that is based on the definition of a so-called “guiding hyperplane” that is used to guide the search towards the Pareto front. The guiding hyperplane is defined as passing through the ideal point with respect to the individuals in the current generation, and as being orthonormal to the largest eigenvalue of the scatter matrix of the individuals in the current generation. Two performance measures are considered based on the guiding hyperplane: The individuals fitness is evaluated based on their Euclidean distance from the guiding hyperplane, and their diversity is measured using projections onto the guiding hyperplane. The resulting Target Hyperplane Evolutionary Algorithm THEA is tested on example problems taken from the test suite of Zitzler, Deb and Thiele (1999). Numerical results on selected example problems show the superiority of the THEA over SPEA2 and NSGA2. Reviewer: Kathrin Klamroth (Erlangen) MSC: 90C29 Multi-objective and goal programming 90B50 Management decision making, including multiple objectives Keywords:evolutionary computation; multicriterial optimization PDFBibTeX XMLCite \textit{C. Rotar} et al., Acta Univ. Apulensis, Math. Inform. 12, 49--62 (2006; Zbl 1183.90369) Full Text: EuDML