A fire protection system takes on a critical significance to building operation. This paper describes the use of digital twins (DTs) and semantic web technologies for the intelligent control of building fire protection (BFP) systems in fire accidents. Specifically, a data fusion stage and several information-based control mechanisms are involved in the use of the two technologies. A designed BFP ontology is considered as the semantic model and basis of data fusion between the static building geometric information and the dynamic sensing data. The above information is incorporated into a DT data model, which is considered as the mapping of physical space. Moreover, rule models and process models are developed to achieve intelligent control mechanisms and keep the DT data model synced with the physical space. A case study based on a fire accident simulation was conducted to verify the feasibility of the use of DTs and semantic web technologies.