Taking into account epistemic uncertainty in the preliminary aircraft design is relevant due to the lack of accurate and statistical information necessary for applying traditional deterministic and probabilistic mathematical models. When accurate initial data are not available, experts are usually involved in determining the parameters necessary to solve calculation and optimization problems. Experts form functions that reflect the distribution of values of non- deterministic parameters depending on the confidence level in the implementation of these values. To describe and use such functions in the preliminary aircraft design, it is proposed to use the theory of uncertainty, which enables one to calculate the numerical characteristics of the expressions that depend on uncertain parameters specified by uncertainty distributions. The paper focuses on solving the problem of taking into account the uncertainty of the optimized parameters, which arises when there is no confidence in the possibility of exact compliance with the selected design solutions. The developed models are studied within solving the task of the aircraft power plant parametric synthesis.