Current aircraft cargo bay fire detection systems are generally based on smoke detection. Smoke detectors in modern aircraft are predominately photoelectric particle detectors that reliably detect smoke, but also detect dust, fog, and most other small particles. False alarms caused by these contaminants can be very costly to the airlines because they can cause flights to be diverted needlessly. To minimize these expenses, a new approach to cargo bay fire detection is needed.
This paper describes a novel fire detection system developed by the Goodrich Advanced Sensors Technical Center. The system uses multiple sensors of different technologies to provide a way of discriminating between real fire events and false triggers. The system uses infrared imaging along with multiple, distributed chemical sensors and smoke detectors, all feeding data to a digital signal processor. The processor merges data from the chemical sensors, smoke detectors, and processed images to determine if a fire (or potential fire) is present. Decision algorithms look at all this data in real-time and make the final decision about whether a fire is present.
In the paper, we present a short background of the problem we are solving, the reasons for choosing the technologies used, the design of the system, the signal processing methods and results from extensive system testing. We will also show that multiple sensing technologies are crucial to reducing false alarms in such systems.