The use of electric vehicles (EVs) in the United States is growing at a rapid pace, with some projections estimating that the majority of cars will be electric by the year 2050. This transition aims to mitigate greenhouse gas emissions and provide a more sustainable means of transportation. However, as EV adoption rates rise, it becomes more challenging to support EV drivers with sufficient charging infrastructure that can be counted on even during extreme scenarios, such as natural disasters or road shutdowns. The inacessibility of adequate charging stations during these extreme scenarios can have catastrophic fallout. Therefore, building resilience into EV infrastructure is essential not only for daily operations but also for disaster preparedness and recovery. This requires the development of methodologies to forecast charging demand, identify weakness in current charging infrastructure, and optimize capacity expansion decisions. This thesis presents proof-of-concept computational tools which can be used for modeling, analyzing, and optimizing the resilience of EV charging infrastructure under both normal and disrupted conditions. Specifically, the study first quantifies the impact of natural disasters on traffic flow utilizing a long-short-term-memory neural network to analyze time series traffic flow data and build a counterfactual prediction for traffic flow at the time of a real-world natural disaster. The same traffic data is used to estimate trip volumes on the road network by building a least squares model to generate a realistic route schedule. The trip generation model provides input to an agent-based simulation developed to evaluate the performance of the existing EV charging network under disruption conditions, including charging station outage, road closure, and cold weather, while capturing the impact on EV users. Key metrics are collected from the simulation and provide a foundation for an objective function to a bi-level optimization model, which identifies optimal placement of additional charging stalls. By integrating data-driven disruption analysis, behavioral simulation, and network optimization, this research offers a computational framework to support resilient EV infrastructure planning.