As a representative infrastructure for consumer Autonomous Intelligent Systems (AISs), the Internet of Electric Vehicles (IoEV) integrates real-time sensing, communication, and edge computing to support distributed, resource-constrained electric vehicles (EVs) in delivering intelligent, low-latency mobility services. Digital Twin (DT) technology plays a pivotal role in IoEVs by creating virtual replicas of EVs, enabling predictive analytics and intelligent decision-making. However, frequent DT migrations driven by high EV mobility, coupled with the dynamic availability of sensing, communication, and computation resources at the edge, present significant challenges in ensuring timely and accurate DT updates. To address these issues, we propose an adaptive DT migration framework that holistically integrates sensing, communication, and computation into a unified optimization model tailored for dynamic IoEV environments. Specifically, we formulate a joint problem of DT migration and multi-modal resource allocation, aiming to minimize DT updating error while satisfying system capacity constraints. To handle the dynamic and heterogeneous nature of resource availability in IoEVs, we design a Dynamic Contextual Distributed Auction (DC-DA) algorithm that enables EVs and edge nodes to make decentralized, context-aware decisions for coordinated DT migration and resource provisioning. Extensive simulations demonstrate that our approach significantly improves DT update accuracy, reduces service latency, and enhances resource utilization compared to baseline methods, offering a scalable and robust solution for next-generation intelligent electric vehicle systems.