Quantitative Analysis, Based On State Space Theory Of The Gas Concentration

Optical technique is becoming one of the most important way in gas monitoring field. It is better than traditional way not only due to its precision or quickness, but also due to its safety and practicability. This thesis mainly focuses on the improvement of the optical technique algorithm in gas monitoring.As a monitoring technique using spectroscopy, the differential optical spectroscopy technique(DOAS) relies on the narrow molecular absorption bands and absorption strength to gain concentrations of the trace gases. DOAS technique has been widely used in the practical atmosphere gas detection. However, due to the least squares method and the separation of broad bands and narrow bands, DOAS can't not work well in the situation where exists big noise. Especially when the atmosphere conditions have changed, DOAS can't accurately recover the concentration.A new gas concentration analyse algorithm based on state space theory has been developed. The algorithm uses state space equation which consists of state equation and measure equation to recover the concentration. In the situation when the noise property is unknown, adaptive kalman filtering theory is used to estimate the unknown noise . The algorithm recovers gas concentration accurately, even if the SNR is low. Compared to DOAS, adaptive kalman filter based on state space theory can also works well when there exists state-space model error. A gas monitoring system which is used to verify our algorithm has been designed and builded up. Finally, It is proved that state space theory algorithm is better than DOAS in the pratical situation.At last, a state space theory algorithm which can recover the gas concentration and its temperature parameters simultaneously is developed detailedly.