Mobile devices are resource-limited, and task migration has become an important and attractive feature of mobile clouds. To validate task migration, we propose a novel approach to the simulation of task migration in a pervasive cloud environment. Our approach is based on Colored Petri Net (CPN). In this research, we expanded the semantics of a CPN and created two task migration models with different task migration policies: one that took account of context information and one that did not. We evaluated the two models using CPN-based simulation and analyzed their task migration accessibility, integrity during the migration process, reliability, and the stability of the pervasive cloud system after task migration. The energy consumption and costs of the two models were also investigated. Our results suggest that CPN with context sensing task migration can minimize energy consumption while preserving good overall performance.