On the Efficiency of the Evolutionary Approach to Dynamic Objects Identification in the Form of Differential Equations Systems

The study of dynamic systems often assumes building a model based on numerical data describing the system’s behavior, i.e., its identification. This problem is relevant, since it is impossible to study properties of a real object, make predictions and select correct control without an adequate mathematical model. The representation of the object’s model in a symbolic form as differential equations or their systems is the most common and convenient for further application. The paper examines the efficiency of the evolutionary approach to the structural-parametric identification on test problems, the problem of modeling the longitudinal perturbed motion of an aircraft and the problem of modeling the viscous fluid movement when heated it from below.