Inverse estimation of the particle size distribution using the Fruit Fly Optimization Algorithm

The Fruit Fly Optimization Algorithm (FOA) is applied to retrieve the particle size distribution (PSD) for the first time. The direct problems are solved by the modified Anomalous Diffraction Approximation (ADA) and the Lambert Beer Law. Firstly, three commonly used monomodal PSDs, i.e. the Rosin -Rammer (R-R) distribution, the normal (N-N) distribution and the logarithmic normal (L-N) distribution, and the bimodal Rosin-Rammer distribution function are estimated in the dependent model. All the results show that the FOA can be used as an effective technique to estimate the PSDs under the dependent model. Then, an optimal wavelength selection technique is proposed to improve the retrieval results of bimodal PSD. Finally, combined with two general functions, i.e. the Johnson's S-B (J-S-B)) function and the modified beta (M-beta) function, the FOA is employed to recover actual measurement aerosol PSDs over Beijing and Hangzhou obtained from the aerosol robotic network (AERONET). All the numerical simulations and experiment results demonstrate that the FOA can be used to retrieve actual measurement PSDs, and more reliable and accurate results can be obtained, if the J-S-B function is employed. (C) 2014 Elsevier Ltd. All rights reserved.