Trace gases analysis in pulsed photoacoustics based on swarm intelligence optimization

pplication of pulsed photoacoustic spectroscopy for in-situ measurements of trace gases with changeable spatial and temporal distribution requires high sensitivity, selectivity, and easy handling devices. In order to improve PAS characteristics in trace gases measurements, we have applied computational intelligence. Computational intelligence as a combination of learning, adaptation, and evolution may increase efficiency and precision of measurements and provides possibility of system–environment interactions. Two metaheuristic techniques are applied to improve accuracy in simultaneous determination of photoacoustic signal parameters: particle swarm optimization and artificial bee colony optimization. Swarm intelligences are applied to simultaneously determination of unknown parameters of photoacoustic signal: radius of the laser beam spatial profile r_L and vibrational-to-translational relaxation time τ_V-T . Experimental PA signals are generated in the SF6 + Ar mixture in multiphoton regime. Results produced by particle swarm optimization and artificial bee colony are discussed. Values of common parameters: population size and number of iterations in both algorithms are chosen to be the same. Advantages, such as easy implementation, high precision, and the possibility of finding solutions in a wide range of parameters make swarm intelligence algorithms efficient and perspective tool for in-situ measurements.