An ecology-based evolutionary algorithm to evolve solutions to complex problems

Evolutionary algorithms have shown great promise in evolving novel solutions to real-world problems, but the complexity of those solutions is limited, unlike the apparently open-ended evolution that occurs in the natural world. In part, nature surmounts these complexity barriers with ecological dynamics that generate a diverse array of raw materials for evolution to build upon. The authors previously introduced Eco-EA, an evolutionary algorithm that integrates these natural ecological dynamics to promote and maintain diversity in the evolving population. Here, we apply the Eco-EA to the real-world software engineering problem of evolving behavioral models for deployed nodes in a remote sensor network for flood monitoring. We show that the Eco-EA evolves good behavioral models faster than a traditional EA, generates a more diverse suite of models than a traditional EA, and creates models that are themselves more evolvable than those created by a traditional EA.