Development of an early warning fire detection system based on a laser spectroscopic carbon monoxide sensor using a 32-bit system-on-chip

Abstract

A distributed feedback laser-based carbon monoxide (CO) sensor has been developed for early warning detection of fire. The sensor, based on a system-on-chip with high-performance 32-bit microcontroller (MCU), features a digital lock-in amplifier (DLIA) for wavelength modulation spectroscopy (WMS) to improve detection sensitivity. The ramp scanning signal and a sinusoidally modulated signal were generated and superimposed by the MCU, and converted by a digital-to-analog converter to scan the absorption line of CO around 2.327 μm. The n-harmonic components obtained through multiplication of the n-th cosine wave data of the spectral signal, were then processed by integration and filter algorithms. The Allan-Werle deviation was used to evaluate the long-term performance of the sensor. The limit of detection (LoD) for CO was calculated to be 0.0875 ppm based on an optimal integration time of ∼24 s. In field tests, five small-scale standard fires were investigated in accord with EN54 and the results for the test substances – cotton rope (smoldering fire), polyurethane foam (smoldering fire), beech wood, a sheet of A4 paper and polyvinylchloride – verified the reliability and robustness of the developed sensor. Relatively fast response times (110 s for foam and A4 paper; 250 s for PVC, beech wood and cotton rope) were realized based on a threshold of 5 ppm CO. The CO sensor is considered suitable for integration into portable devices for industrial application, particularly for in-situ early warning fire detection.