Construction tower cranes are prone to accidents such as instability, overturning, tower arm breakage, and collisions due to factors including the structure of the crane, workers' operational behavior, and meteorological conditions. Investigating safety dynamic monitoring technologies based on multi-sensor integration is crucial for identifying hazardous situations, issuing timely warnings, and preventing accidents. This study analyzes the causes of tower crane accidents by examining the tower crane system's composition and the factors triggering safety incidents. It proposes a safety control framework for tower cranes and establishes the mapping relationship between monitoring objects and sensors. A multi-sensor tower crane state perception system is designed to acquire real-time, multi-source monitoring data. Utilizing an established safety hazard identification model, a tower crane safety monitoring technology system is developed. The practical application of this system demonstrates that appropriately deploying multiple sensors enables effective real-time monitoring of tower crane operations. The analysis of monitoring data allows for the timely identification of hazards and the implementation of control measures, thereby significantly reducing the incidence of production accidents and providing essential technical support for safe building construction.