Using previous result during a Otimization

[Norton Daniel Barth]

Fist of all thanks for any help

I would like to know if is possible using result of previous pass on genetic otimization for the next generation, 

For example, if I try to normalize the max Profit, I need to know the current max e min price, so f(x) = Profit/(Max(Profit)-Min(Profit))

How can I store the previous values instead to run a optimization, set the max e min like a inputs and restart the optimization?