Paper Conference

Proceedings of BSO Conference 2012: 1st Conference of IBPSA-England


Implementation of Pareto-Archive NSGA-II Algorithms to a Nearly-Zero-Energy Building Optimisation Problem

Mohamed Hamdy, Matti Palonen, Ala Hasan

Abstract: Finding cost-optimal solutions for nearly-zero energy buildings (nZEB) according to the new recast of the European Energy Performance of Buildings Directive EPBD- recast 2010 is a challenging task that requires exploring numerous possible combinations of energy-saving and energy-supply measures. Multiobjective simulation-based optimisation can be effectively used to find solutions close-to the Paretooptimal front. The current study tests the performance of three optimisation algorithms based on NSGA-II (elitist non-dominated sorting genetic algorithm, by Deb et al. 2002) in finding the costoptimal and nZEB solutions for a problem that has a large discrete solution-space. Optimisation runs with low number of evaluations show that a presented active-archive strategy (unconstrained archive for possible non-dominated adult solutions) can significantly improve the repeatability of the original NSGA-II in finding a diverse set of solutions near-to the true Pareto-optimal front.
Pages: 181 - 188