Research On Automatic Fire Detection Technology With Aspirating Smoke Detection Based On Light Scattering Principle

Aspirating smoke detection is the fire detection technology developing rapidly in recent years which has characters such as high sensitivity and fast response, where has been used in hugue space site such as plants, warehouses, power plants emporiums, and in clean spaces such as sarchives, workshops of precise apparatus and products. the principle of this technology is detection to dispersion of somke particles produced in the forepart of fire. This technique has been one of important detecting dechnologies and become main mathod of fire detection and alarm in forepart or super forepart of fire, which has very expansive applicationThe paper has introduced the development of smoke detecting technology, studied space distribution character of the smoke particle scattering-polarization characters of some typical particles, then presented relative physical quantities about particles dispersion such as scattering coefficient, absorption coefficient, extinction coefficient, intensity function and calculation method of polarization.The paper has designed the system of automatic detection of aspirating smoke detetor. Significance of light path components protecting was presented. The detection and control circuit of inspiratory flow and partition of fire detecting were made.This paper proposes a smoke particle recognition method with BP neural network, describing the particle recognition method of BP neural network and the character extraction process of different kinds of smoke particles are illustrated in detail.Unable to detect smoke in early stage and false alarm are the main disadvantages using the traditional methods for smoke detection. Based on laser scattering by smoke particles, two sets of smoke detection and recognition systems are proposed in the thesis.system uses the semiconductor laser and PIN photodiode array as the transmitter and receiver. It has the advantage of high sensitivity and high recognition-rate. Smoke detection and recognition system can recognize different kinds of smoke particles by measuring the scattering intensity on different angles and polarization directions. So the system can be used for fire warning to achieve reliability and validity.