The article provides a brief overview of the methods for forecasting technical and economic processes, reveals their features. A method for estimating the parameters of models with a varying value of the degree of time is proposed, which can be used to implement the method of growth models and the method based on representations of fuzzy sets. The features of methods based on growth models are considered. The simplest methods are compared, the implementation of which can be represented in the form of the following steps: 1) a curve is selected heuristically (expertly), smoothing the initial statistical data and extrapolating to their perspective, while the degrees of the terms of the model are chosen constant; 2) the parameters of the models are estimated, most often by the least squares method; 3) the adequacy of the model by the initial data and the value of the predicted parameter are evaluated. As analysis has shown, the first two stages are decisive in the process of increasing the accuracy of the forecast, and therefore it is advisable to improve them. To do this, it is proposed to use an iterative process, which allows you to simultaneously vary the coefficients at various degrees of the time parameter, as well as its degree. The iterative process is presented in detail by analytical calculations that are convenient in engineering applications.