The problem of reconstructing the characteristics of disperse particles from measurements of scattered radiation is considered. To solve this problem, the neural network method, based on the approximation of the parameters of particles by a linear combination of the results of measurements, is used. The capabilities of the method are studied on the examples of the reconstruction of the radius and the refractive index of spherical particles from measurements (for example, in flow-type cytometers) of the luminance of radiation scattered by individual particles, as well as the reconstruction of the mean radius, the coefficient of variation, and the refractive index from measurements of the luminance of radiation scattered by an ensemble of particles. Errors in the reconstruction of the characteristics of disperse particles depending on the structure of the neural network and the parameters of particles are studied. (C) 2004 MAIK "Nauka/Interperiodica".
