Sequential detection and concentration estimation of chemical vapors using range-resolved lidar with frequency-agile lasers

摘要

This paper extends our earlier work in developing statistically optimal algorithms for estimating the range-dependent concentration of multiple vapor materials using multiwavelength frequency-agile lidar with a fixed set of wavelength bursts to the case of a time series processor that recursively updates the estimates as new data become available. The concentration estimates are used to detect the presence of one or more vapor materials by a sequential approach that accoumulates likelihood in time for each range cell. A Bayesian emthodology is used to construct the concentration estimates with a prior concentration smoothness constraint chosen to produce numerically stabel results at alonger ranges having weak signal return. The approach is illustrated on synthetic and actual field test data collected by SBCCOM.