In this paper a modification to the adaptive mechanism in Success-History based Adaptive Differential Evolution (SHADE) with Linear decrease in population size (L-SHADE) is proposed in order to overcome the premature convergence, while optimizing higher dimensional problems. This modification can be also useful in constrained optimization, where the improved exploration could help in overcoming constrained areas. The proposed modification of the algorithm is tested on the CEC2015 benchmark set and compared to its basis – L-SHADE.