Temperature declines are induced by cold waves, leading to reduced electric vehicle (EV) range and triggering failures in charging infrastructure. As a result, charging demand cannot be met, and the resilience of the EV charging networks (EVCN) is compromised. To address this issue, a resilience enhancement scheme based on intelligent navigation is proposed. The impacts of cold waves on the EVCN are comprehensively analyzed. The failure mechanisms and cascading characteristics of charging stations under cold wave conditions are investigated, and historical data are processed to establish a cascading failure model. To enhance supply-side resilience, mobile emergency generators are navigated to faulty stations for power compensation using a navigation model trained via graph reinforcement learning. In parallel, the same model is utilized to recommend suitable charging stations and optimize routing for EVs in need of charging, thereby improving resilience from the demand side. Through case studies, cascading load increases are identified as the primary cause of failures during cold waves. The proposed collaborative navigation approach ensures stable power delivery and rapid recovery under fault conditions, while reducing waiting times and fulfilling users' charging demand.