面阵探测下的污染云团红外光谱仿真

研究污染云团的红外光谱仿真,对于利用仿真光谱进行光谱识别的算法研究十分重要.在单元探测器探测方式下污染云团的红外光谱仿真研究取得了一定成果,并且已经被应用于识别算法的研究工作中.针对基于成像光谱仪的污染云团识别算法研究缺乏实测数据的问题,利用具有高仿真精度的基于物理模型的污染云团扩散模型及其仿真结果网格化的特点,研究相应的云团红外光谱仿真多层模型,提出了面阵探测方式下污染云团红外光谱的仿真方法,生成了同时具有光谱维和空间维信息的数据立方体,为该研究领域提供了新的研究方法.提出的面阵探测方式下的污染云团红外光谱仿真直观地反映了污染云团的扩散,提供了完备且符合实际情况的污染云团红外光谱立方体数据,对于提高和完善红外光谱识别算法具有重要意义.

Simulation of pollutant-gas-cloud infrared spectra under plane-array detecting

The research on simulation of pollutant-gas-cloud infrared spectra is very important for studying the spectral identification algorithms by using simulated spectra. Some good results of the simulation of pollutant-gas-cloud infrared spectra under single-detector detecting are achieved, and have been used for studying the spectral identification. With the development of infrared detection technology, the infrared imaging spectrometer is used to detect pollutant gas cloud. The gas identification algorithms in the way of plane-array detecion based on the imaging spectrometer also need a number of measured gas-cloud infrared spectrum data cubes. Due to the lack of measured data in studying the spectral identification algorithm that is based on imaging spectrometer, the multiple-layer model of the cloud infrared spectrum is well studied by using the high-precision physics-based gas-cloud explosion model and its gridding simulation results, and the way of simulating pollutant-gas-cloud infrared spectra under plane-array-detector detecting is proposed to generate the infrared spectrum data cube with both spectral and spatial information, which obtains a new research method for the research field. Validations are made by comparing the measured data with the simulated data, and the comparison contains three parts: i) the comparison of measured gas-cloud explosion with simulated gas-cloud explosion, ii) the comparison of spectral identification imaging results between the measured and the simulated gas-cloud infrared spectrum data cubes, and iii) the comparison between the measured and the simulated gas-cloud infrared spectra. The comparison results have two aspects: the first aspect is that the simulated gas-cloud explosion is consistent with the measured explosion and has little difference in separate parts, and the second aspect is that the simulated gas-cloud spectra have relative errors of less than 10% compared with the measured gas-cloud spectra. The conclusion is that the simulation model of pollutant-gas-cloud infrared spectra under plane-array detecting is correct, which is obtained from the validation results that simulated gas-cloud infrared spectrum data cubes are highly precise, whether in the comparison with spectral identification imaging results or in the comparison with gas-cloud spectra. The simulation of pollutant-gas-cloud infrared spectra under plane-array detecting which directly reflects the explosion of pollutant gas cloud and provides complete and realistic infrared spectrum data cube of pollutant gas cloud, is significant for improving and perfecting the spectral identification algorithms.