IR Neural Network Flame Detector Stops False Alarms - Multi-spectrum IR sensing and neural network processing together in one device

摘要

There are five primary optical flame sensing technologies in use today, all of them based on either UV spectrum or IR spectrum sensing. Depending on the application, the designer selects a specific technology. While these devices meet basic flame detection needs, they are prone to false alarms caused by reflected sunlight, arc welding, hot piping, machinery, and other false-alarm sources, "False alarms are more than a nuisance. They are both a productivity and cost issue," says Ardem Antabian, product manager at General Monitors (www.generalmonitors.com). "From the perspective of productivity, a false alarm incident in most process plants requires a system shutdown. Turning off and then restarting a process line may take anywhere from an hour to a whole shift."