I studied in in class and never really understood. I found teh best way : implement your own.
At the end this is a very "pragmatic" algorithm and it is more an execution routine than a full totalitarian method.
As for examples, you can see it in action on the classical (and very visual!) travelling salesman problem like here
I think you do not need much than understanding the components for starters : chromosomes (= solutions), mutation function (altering one solution), crossover (mixing two solutions, meaning selecting parents and actually mixing them), fintess function (quality of the solution), and generation function (candidate solutions generation) and (optionally) elitism policy (how much we protect best solutions from disappearance).
After having understood it, you will see the modular aspect of GA and you will be able to dig in more advanced content (evolutionnary strategy, coding, etc.) if needed.
Thing is that in GA you almost define everything, they are only classical execution routines. This is why I do not think books will add a lot to your understanding in this case.