Transformation of independent variables of problem for increasing the POD method efficiency

Problems of the increase of the proper orthogonal decomposition method (POD) approximation efficiency by transformation of independent variables of investigated problem are considered. Transformation carried out taking into account the physical process character, problem symmetries and investigated fields topological peculiarities permits to reflect dominant spatio-temporal structures of the investigated problem using approximative representation with small number of basis functions. Examples demonstrating the efficiency of the described approach are presented.