A typical safety monitoring system architecture with basic functions including data collection, parameter setting, data processing and backup, warning judgment and processing is proposed based on analysis of potential safety hazards. With the theories such as T-S model, domain processing and fuzzy control, a fuzzy control model based on multi-sensor data fusion is established, and a safety monitoring system software is developed. Test result shows that the software supports both analog and real-time data, and can perform real-time monitoring of safety parameters like temperature and smoke with satisfactory online hazard warning performance, thus able to be applied in environmental safety monitoring.