Research And Product Development Of MIR Flame Detector System

In recent years, the fire occurred in the oil and gas fields, gas stations, coal mines, power plants and other industrial sites, caused great loss of our lives and property. Traditional smoke, temperature sensitive detection methods have the disadvantages of short detection distance and response slowly can not meet the requirements, therefore, the development of a flame detection system which can alarm accurately and fastly becomes necessary.The paper surveyed the flame detector development situation at mainland and abroad, in-depth analysis of the fire characteristics and traditional flame detection method in terms of intrinsically safe, explosion-proof, detection distance, response time, false alarm and stability shortcomings and limitations. The prototype flame detector system is based on infrared pyroelectric principle and take cholic lithium sensor as sensitive unite to perceive infrared radiation intensity information that flame emitted, which will effectively improve the detection distance and sensitivity through of multiple infrared radiation detection technology(MIR), and have a good inhibition of complex background interference.The paper studied and designed a flame identify algorithm and control software. Three sensors have the different optical characteristics such as center wavelength and the pass bandwidth, make the difference of the radiated power and the radiated interference that pass through the sensors, based on the content ratio matrix, the three-channel infrared radiation power can be seperated into the flame power and the interference power. The algorithm contains Energy threshold analysis, bands correlation analysis, spectral analysis of flame flash frequency methods which can judge whether there is a real flame, and remove false fire signal which not meet the flame characteristics, the algorithm simulated in Matlab with several groups of real flame sample, the results show that it has a strong capability of anti-interference in sunlight, artificial light source, heat source, background.Finally, the prototype system verified the feasibility and effectiveness of the flame recognition algorithm, past product function, detection sensitivity, detect distance, anti-jamming capability, real-time response, anti-false alarm test. The test results show that the flame detection method is feasible and the system has fast alarm sensitivity, parameter configuration flexibility and variety of alarm output interface, meet the design requirements. The system has been working at heating molten salt tower in the the Zijingang solar thermal power generation site in July2012, it runs stable and reliable.