As IoT services continue to evolve, they provide similar functionalities while exhibiting significant variations in Quality of Service (QoS) levels. Despite this increase in services, meeting user satisfaction becomes increasingly complex. This situation underscores the need for the composition of IoT services capable of providing innovative functionalities. A critical challenge in the IoT service composition process lies in selecting the most appropriate atomic services to incorporate into the composition. In this research, we propose a method for modeling the IoT service composition process using the meta-heuristic optimization algorithm MGWO, while considering global QoS constraints. Our model enhances the balance between exploration and exploitation. This balance mitigates the risk of converging on local optima during the optimization process. To effectively address constraint violations associated with user-specified requirements, we introduce a concept termed Modified Violation Constraint Handling (MVCH). This methodology culminates in the development of a novel hybrid model, MGWO-MVCH, specifically designed for IoT service composition with an emphasis on optimizing quality of service.