Multi-stage evolution of single- and multi-objective MCLP

摘要: Maximal covering location problems have efficiently been solved using evolutionary computation. The multi-stage placement of charging stations for electric cars is an instance this problem which addressed in study. It particularly challenging, because a final solution constructed multiple steps, cannot be relocated easily and intermediate solutions should optimal with respect to certain objectives. This paper extended version work published Spieker et al. (Innovations intelligent systems applications (INISTA), 2015 international symposium on. IEEE, pp 1–7, 2015). In work, it was shown that through decomposition, incremental genetic algorithm benefits from having stages. On the other hand, decremental strategy does not profit reduced computational complexity. We extend our previous by including multi-objective optimization station placement, allowing us only optimize toward (weighted) demand coverage, but also include second objective, taking into account traffic density. reachable part full Pareto front at each stage bound chosen respective front. By careful choice selection strategy, particular focus can set. exploited comply concrete implementation goals adjust evolved both static dynamic changes requirements.