Memory-based multi-population genetic learning for dynamic shortest path problems

View/ Open
Date
2019-06Abstract
This paper proposes a general algorithm framework for solving dynamic sequence optimization problems (DSOPs). The framework adapts a novel genetic learning (GL) algorithm to dynamic environments via a clustering-based multi-population strategy with a memory scheme, namely, multi-population GL (MPGL). The framework is instantiated for a 3D dynamic shortest path problem, which is developed in this paper. Experimental comparison studies show that MPGL is able to quickly adapt to new environments and it outperforms several ant colony optimization variants.
Description
The file attached to this record is the author's final peer reviewed version.
Citation : Diao, Y., Li, C., Zeng, S., Mavrovouniotis, M. and Yang, S. (2019) Memory-based multi-population genetic learning for dynamic shortest path problems. Proceedings of the 2019 IEEE Congress on Evolutionary Computation, Wellington, New Zealand, June 2019.
Research Institute : Institute of Artificial Intelligence (IAI)
Peer Reviewed : Yes