The aim of the study was to study the possibility of using the Dirichlet distribution as a statistical model of the process of dynamic interaction of large-span structures with aerodynamic load. As an object of research, a model of a hangar building was chosen for the maintenance of two AirbusA-380 aircraft at a scale of 1: 500 blown in a wind tunnel at five directions of wind flow at angles of 0 °, 30 °, 45 °, 60 °, 90 ° to the front of the structure ... It is shown that the statistical model of the Dirichlet distribution has sufficient flexibility and realism and compares favorably with the models built on the basis of the normal distribution, beta distribution, gamma distribution and Poisson distribution when describing the processes of intense dynamic interaction of wind load with structural elements. The problem of finding an integral assessment of the state of the processes of interaction of the wind load with the hangar building for the maintenance of two aircraft was reduced to calculating the self-organization parameter of the distribution over the surface of the building model of the aerodynamic coefficients, which was taken as the ratio of the total weighted amount of the Dirichlet models detected in the analyzed numerical series of the registered signal i-th dimension with negative external entropy to the total weighted number of Dirichlet models with positive external entropy.
 The study showed that the dimensionless informational and statistical indicator of self-organization makes it possible from a unified standpoint to assess the states that differ from each other in terms of external signs, the processes of interaction and their dynamics. The integral indicator can be used to rank the efficiency of operating systems and can be useful in assessing the state of the processes of dynamic interaction of the operating pressure and objects and structures of various shapes and purposes.