Genetic Algorithms Application for the Photoacoustic Signal Temporal Shape Analysis and Energy Density Spatial Distribution Calculation

摘要

Recently, a few numerical methods based on the photoacoustic (PA) signal temporal shape analysis and energy density spatial distribution calculation, which is directly related to the laser beam spatial profile, have been presented. It has been shown that these methods allow a precise reproduction of the spatial profile and the radius of the laser beam, determining the vibrational-to-translational (V-T) relaxation time with good accuracy. Their applicability has been shown and confirmed for the analysis of an arbitrary symmetric laser beam spatial profile in cylindrical geometry. Here, the application of genetic optimization for solving the problem of a simultaneous laser beam spatial profile and V-T relaxation time determination by pulsed PAs is presented. Real-coded genetic algorithms are used to calculate the mentioned relaxation time by fitting the experimental signal δp(r, t) with the theoretical one. The aim is to find combinations of PA signal parameters, namely, the radius of the laser beam and the V-T relaxation time that provide the best match with the given signal. A calculated PA signal with a known profile is used to simulate an experimental signal, and the sum of the square deviations representing deviations of the given and fitted signals is minimized by means of genetic optimization. In that way, the genetic algorithms are used to simultaneously estimate the radius of the laser beam and the V-T relaxation time efficiently and with high accuracy. Compared to previous methods, the presented method is much simpler and requires less time to compute.