This research paper focuses on hybridization of two soft computing fields – chaos theory and evolutionary algorithms, specifically on the implementation of Chaotic map based Pseudo-Random Number Generator (CPRNG) into the process of parent selection in Success-History Based Adaptive Differential Evolution (SHADE) algorithm. The impact on performance of the algorithm is tested on CEC2015 benchmark set where five different chaotic maps are used for random integer generation. Performance comparison shows that there is a potential in replacing classic Pseudo-Random Number Generators (PRNGs) with chaotic ones. The results provided in this paper show that the choice of CPRNG for given problem is crucial in terms of affecting the performance of the algorithm, therefore the next research step will be focused on the development of the framework which will adapt to the solved problem and select the most suitable CPRNG or their combination.