Neural networks for aerosol particles characterization

- Retrieval of parameters of spherical homogeneous nonabsorbing aerosol particle is investigated.
- Multilayer perceptron neural networks with one, two, and three inputs are built.
- Optimization of amount and values of angles at nephelometric measurement is considered.
- Dependence of accuracy of particle parameters retrieval on the input data is simulated.
- Retrieval error does not exceed 5% at multiplicative error 15% in size interval 1.778-56.234 μm.

Multilayer perceptron neural networks with one, two and three inputs are built to retrieve parameters of spherical homogeneous nonabsorbing particle. The refractive index ranges from 1.3 to 1.7; particle radius ranges from 0.251 μm to 56.234 μm. The logarithms of the scattered radiation intensity are used as input signals. The problem of the most informative scattering angles selection is elucidated. It is shown that polychromatic illumination helps one to increase significantly the retrieval accuracy. In the absence of measurement errors relative error of radius retrieval by the neural network with three inputs is 0.54%, relative error of the refractive index retrieval is 0.84%. The effect of measurement errors on the result of retrieval is simulated.