he investigations are in progress associated with estimating the efficiency of the use of promising wide-band lidars for sensing of microphysical parameters of cloud aerosol in the near- infrared spectral region. The iteration method proposed before has been generalized for the case of the multiwave laser sensing. In the framework of a closed numerical simulation the validation has been made of a new optimization algorithm, based on the combination of the ideology of the method of artificial neuron nets with a genetic search. The reconstruction of altitude profiles of the cloud particle spectrum was made possible up to considerable optical depths under conditions of the multiple scattering noise.
lidar; transfer equation; genetic algorithms; neuron nets