The corresponding equations of unsteady boundary layer, by introducing the appropriate variable transformations, momentum and energy equations and one similarity parameter set, are transformed into generalized partial nonlinear differentia] equation. These parameters express the influence of the outer flow velocity, and the flow history in the boundary layer on the boundary layer characteristics. Since the equation contains the sums of terms equal to the number of parameters, it is necessary to limit the number of parameters for numerical integration. So it is very important that, the chosen set of parameters possesses the following two properties: 1. the first parameter is to be "strong" enough, so that the solutions lies close to the exact solution, and 2. the following parameters introduce in the solution small corrections only, and provide a sufficiently fast convergence. For this purpose, the modem parameter method has been developed to calculate boundary layers, known as generalized similarity method. The numerical integration of the generalized equation with boundary conditions has been performed by difference schemes and using Tridiagonal Algorithm Method with iterations in full two parametric approximation, where the first, unsteady and dynamic parameters will remain, while all others will be let. to be equal to zero, and in two-once localized parametric approximation, where also the first unsteady and dynamic parameters remain, while all others will be equal to zero and where the derivatives with respect to the first, unsteady parameter will be considered equal to zero, while the derivatives with respect to the first porous parameter will be considered equal to zero. The obtained results show that for both the confuser and this diffuser regions as well as for both the accelerating and decelerating flows, there are differences between their values, especially close to the separation point of the boundary layer, and is very important for particular problem with laminar-turbulent, transition region on the contour.