This study endeavors to optimize the logistical distribution of electric vehicles (EVs), considering the characteristic that customer demand can be split into several discrete orders. With the objective of minimizing the fixed, routing, charging, and time window penalty costs for the EVs, a multi-type EV charging strategies and routing optimization model is formulated that considers discrete split demands. Given the characteristics of this model, an improved genetic-simulated annealing algorithm is designed. The effectiveness of the algorithm is validated through empirical analysis. The findings indicate that the algorithm can efficiently optimize EV charging strategies and distribution routes under discrete split demands. Notably, the partial charging strategy not only reduces total costs but also shortens charging time compared to the full charging strategy. Furthermore, sensitivity analysis reveals that as charging waiting time increases, the time window penalty costs rise for both strategies. However, the cost growth rate for the partial charging strategy is notably lower than that of the full charging strategy, suggesting its superior suitability in scenarios involving prolonged charging waiting times. This research offers valuable guidance for logistics companies seeking to optimize EV distribution operations.