Nature-Inspired Computing and Optimization by Srikanta Patnaik Xin-She Yang & Kazumi Nakamatsu

Nature-Inspired Computing and Optimization by Srikanta Patnaik Xin-She Yang & Kazumi Nakamatsu

Author:Srikanta Patnaik, Xin-She Yang & Kazumi Nakamatsu
Language: eng
Format: epub
Publisher: Springer International Publishing, Cham


3.7 Multi-objective Invasive Weed Optimization

The multi-objective invasive weeds optimization proposed by Kundu et al. [25] is based upon the soil occupying and new colony generation behaviour of the weeds. The seeds of the weeds are first scattered in suitable farmlands and grow into weeds. The newly generated colonies usually grow around the existing weeds and those grown in arable areas have higher chance of survival. This forms the base of the invasive weed-based multi-objective optimization algorithm. Firstly, in MOIWO, a set of weeds are generated randomly as the initial population and the objective function of each potential weed are evaluated. The entire population is then sorted using the fuzzy ranking and each member weed of the population is allowed to generate seeds with higher potential members. Generation of new weed colonies improves the search space exploration. The newly generated seeds are then scattered over the search space by using normal distribution and standard deviation functions while breeding is based upon competency between the members. As soon as the weed population crosses a threshold limit, fuzzy ranking is applied again and the best members are kept for further evolution whereas the weaker ones will be eliminated to maintain competitiveness between the members of the population. Consequently, new weeds come closer to the parent weeds, thus reducing the distance between them and transforming from diversification at the beginning to intensification by the end. This process is repeated until the stopping criterion is met.



Download



Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.