One possibility is to use each agent as an individual of GA's population. Then, you can apply crossover and mutation using any MAS negotiation method, and modify the internal state of agents. The hardest problem is to evaluate the fitness function; but it can be done locally forming coalitions and voting among them for the best local candidate. If you need more support, you can contact me again!

Thanks Hiram. That was very useful. Do you know of any papers or books that touch upon this topic. Thanks Again!

Read an IEEE conference article entitled "Genetic Algorithms in a Multi-Agent System". This paper will guide you more properly. Here is the DOI: dx.doi.org/10.1109/IJSIS.1998.685410

We have had encouraging results for multi-agent systems (cooperative mission planning of UAVs) using a hybrid genetic fuzzy approach. The system provided excellent performance, was robust, adaptive and scalable 

For reference: Ernest, N., Cohen K., and Schumacher, C., 2014, “Learning of Intelligent Controllers for Autonomous Unmanned Combat Aerial Vehicles by Genetic Cascading Fuzzy Methods”, SAE 2014 Aerospace Systems and Technology Conference, Cincinnati, OH.

http://papers.sae.org/2014-01-2174/

Thank you very much Kelly

Thanks Hosny