The primary goal of wireless sensor networks from a design point of view is to enhance the network's lifetime. Among the different options for reducing operational energy consumption, energy invested in routing and cluster head selection is considered to be very effective mechanisms. Both tasks have been considered as very challenging and difficult to obtain the efficient solution. Since it is difficult for traditional approaches to satisfy the requirements and difficulties, a heuristic solution focused on natural computation methods has provided a lot of naivety. The proposed work efforts to address these challenges using computational intelligence especially differential evolution and genetic algorithm. An energy efficient route discovery for dynamic network is designed with variations in DE, quick and adaptable routes were discovered for networks undergoing changes. A knowledge based DE has been designed depending on prior knowldge to redefine new routes for changing network. A hybrid mutation strategy under standard DE is designed for cluster head selection providing faster convergence characteristics. The proposed solutions were implemented under MATlab environment and the results have shown that the proposed solutions are performing better for different network configurations. Dynamic route discovery using KDE has achieved energy saving of 9.83 to 49.2 percentage as compared to RDE and 6.7 to 29.5 percentage as compared to PDE. Also energy saving attained in cluster head selection using proposed HMDE is 10 to 33 percentage better compared to dyPSO and 5 to 10 percentage better compared to SDE.