In response to the problems of traditional building fire alarm systems relying on a single sensor

being susceptible to environmental interference leading to high false alarm rates and delayed responses

this paper designs an intelligent fire alarm system based on multi-sensor fusion. The system has built a multi-source data acquisition and communication architecture for smoke

temperature

CO

etc.

achieving software and hardware collaborative control; At the algorithmic level

a fusion judgment and dynamic threshold adjustment model is proposed to optimize alarm accuracy and response speed. Through experimental platform verification

the results show that the recognition accuracy of the system in typical fire scenarios has been improved by 21.6% compared to traditional methods

and the response time has been shortened by about 30%. It effectively reduces false alarms and missed alarms

and has good practical and promotional value.