Dear Gbenga Dada

Please check your solutions (variables) to be in the allowable range.

Thanks

You may take the help of these following papers....

https://arxiv.org/pdf/1004.4170.pdf

https://www.researchgate.net/profile/Xin-She_Yang/publication/309365898_New_directional_bat_algorithm_for_continuous_optimization_problems/links/5817403808aeffbed6c30ba7.pdf

Article New directional bat algorithm for continuous optimization problems

Indeed Rosenbrock's test example function is non-negative everywhere, is non-convex, and has a unique optimum.

By the way - to call it a benchmark is to stretch the term a little - it's too small and easy for that, as even the very bad algorithm known as "steepest descent" solves it (albeit after many, many small steps). :-)

If you check the formulation of this well-known function, you will detect that mathematically it should be positive. You can see the range and solutions for this function in several papers. 

you refer xin she yang papers on bat algorithm

Thanks for your responses.