Fitness decreasing but not converging


This is a more general question about evolutionary algorithms. I have an multi objective optimization problem where the fitness is decreasing as expected but instead of converging on one solution the standard deviation of every generation seems to increase.
Is this kind of behavior typical in evolutionary algorithms? What could be the reason for it?

Yes this kind of behavior is normal in the multi-objective optimisation problems. in simple words it means that as we move forward the generations comprise very fit and poor performing solutions but since the population of fit solutions are higher, they are pulling the average down and overal the generations are getting better and better.