Architecture Design and Application of GMOS Digital Foundation Building Automation System

With the increasing scale of Building Automation Systems (BAS) and the growing complexity of control requirements, conventional BAS architectures centered on physical controllers are increasingly constrained by limited computing capacity, rigid control logic, and insufficient support for cloud-based intelligent algorithms and elastic scheduling. To overcome these limitations, this paper proposes a cloud-based GMOS digital foundation architectural framework for building automation systems, which upgrades the traditional BAS architecture while maintaining compatibility with the existing system ecosystem. Specifically, the control capabilities of physical controllers are virtualized into schedulable software entities, and a cross-architecture compatibility mechanism is introduced to enable legacy control programs to execute seamlessly in cloud environments. In addition, a dynamic resource allocation method for virtual controllers is developed. Combined with data- and configuration-mapping mechanisms, this method supports parallel execution and unified management of multiple control programs. Furthermore, a script-based protocol parsing approach is designed to enable rapid integration and flexible extension of heterogeneous control protocols. Experimental results demonstrate that the proposed system operates stably over long-term deployments. The measured end-to-end average control latency is 105.1 ms, satisfying the real-time requirements of building automation control, while the system throughput remains stable across different load scenarios. In practical engineering applications, the proposed system can reduce the amount of on-site cabling by 47.37% and achieve an average daily energy saving rate of 11.17% while maintaining compatibility with the existing system ecosystem.The research results show that the proposed method can support the evolution of BAS towards cloudification, software-based systems and intelligence, and provides a feasible engineering and technical path for upgrading and transforming existing BAS and building a new generation of intelligent building control platforms.