This paper presents a parallel control system designed for managing emergent events in an urban rail transit system (URTS). The key features of the system include cloud computing, information fusion and expert systems. A multiagent approach is used for modeling diversities and interactions in the URTS. Partially due to its ability to interact with the real system via parallel execution, such a parallel control system would provide a cost-effective solution for emergency personnel training, verification of emergency measures, and validation of emergency plans and management policies.