This article addresses the Dynamic Electric Vehicle Routing Problem with TimeWindows (DEVRPTW) using a hybrid approach blending genetic and Ant Colony Optimization (ACO) algorithms. It employs an Ant System algorithm (AS) with an integrated memory system that undergoes mutations for solution diversification. Testing on Schneider instances under static and dynamic conditions, with run time of 10 and 3 minutes respectively, reveals promising results. Compared to static solutions, deviations of 8.55% and 2.38% are observed in vehicle count and total distance. In a dynamic context, the algorithm maintains proximity to static results, with 10.99% and 4.41% deviations in vehicle count and distance. Instances R1 and R2 present challenges, suggesting potential improvements in memory and pheromone transfer during re-optimization.