The rapid expansion of electric mobility introduces significant challenges in optimizing vehicle routes, particularly under constraints related to energy consumption, limited battery capacity, and the availability of charging infrastructure. The Electric Vehicle Traveling Salesman Problem (EV-TSP) extends the classical Traveling Salesman Problem (TSP) by incorporating these constraints, thereby increasing its computational complexity. This chapter proposes a hybrid approach that combines Genetic Algorithms (GA) with Q-Learning, a reinforcement learning technique, to improve electric vehicle route optimization. Four GA variants are analyzed: three using fixed crossover operators (Order Crossover, Cycle Crossover, and Position Crossover), and an adaptive variant (GA-QL) that dynamically selects the most effective operator through Q-Learning. The methodology utilizes benchmark instances derived from the TSP-Lib dataset, which have been modified to reflect the constraints of electric vehicles. A comprehensive experimental evaluation is conducted, including statistical validation via the Bonferroni-Dunn post-hoc test. Results demonstrate that GA-QL achieves superior performance in terms of route efficiency, energy consumption, and solution stability. Its dynamic operator selection accelerates convergence and enhances adaptability to real-world factors such as traffic variations and stochastic energy usage. In addition to its theoretical contributions, this work outlines practical applications in urban logistics and smart mobility. The proposed framework supports energy-efficient routing and scalability, with future directions including multi-objective optimization, real-time adaptability, and deep reinforcement learning strategies.