We propose adaptive algorithms for constructing a spline-wavelet decomposition that provides an a priori given estimate for deviation of the main flow from the initial one. We introduce the notions of a pseudouniform grid and an adaptive type grid. We establish difference representations of the deviation of the main flows from the initial one. We estimate volumes of data used in the main flow under different characteristics of irregularity of the initial flow in the cases of a pseudouniform grid and an adaptive type grid under the same approximation. We also clarify the limit characteristics of these volumes in the case where the initial flow is generated by a differentiable function. Bibliography: 5 titles.