Search academic texts

Information × Registration Number 0212U004406, 0110U003443 , R & D reports Title Development of the control methods for dynamic systems on the base of guaranteed estimation popup.stage_title Head Kuntsevich Vsevolod Mykhailovich, Registration Date 09-02-2012 Organization Space Research Institute popup.description2 The investigation object is methods and algorithms used for estimation of the state and parameters of dynamic systems and synthesis of their control under uncertainty of their properties, properties of noise and external disturbances. The aim of the work consists in the development of robust methods and algorithms for estimation of the state and parameters of dynamic systems and methods of synthesis of their robust control on the basis of the guaranteed approach to the mathematical interpretation of uncertainties without using assumptions about their probability nature. The investigation methods are methods of linear algebra, set theory, convex analysis, optimization and mathematical programming, methods of estimation and control under uncertainty conditions, computer modelling. For families of linear time-dependent discrete and nonlinear discrete in the general case DSs with limited disturbances, irreducible to a quasi-linear in E. Barbashina, we obtain upper bounds of invariant (limiting) sets of these systems. On this basis, a constructive method for synthesis of robustly stable control of nonlinear DSs from the minimization of the radius of invariant sets is proposed. For multidimensional domain structure, including DSs with distributed parameters, a method for synthesis of control based on approximate models obtained by identification is developed. The method provides an iterative refinement of the structure and parameters of the model DS by means of additional identification. We solve the problem of estimating the parameters of the linear controlled DS based on the results of discrete measurements of noise, which satisfy the proposed uncertainty model. This model generalizes the well-known set-theoretic uncertainty model, which uses only interval estimates of the values in each moment of discrete time and rate of change. We consider a nonlinear DS with a quadratic nonlinearity, which is a mathematical model of the angular motion of a spacecraft (SC) and many other classes of aerospace and marine facilities on the basis of quaternions, which are commonly used in modern strapdown inertial orientation. We propose a robust algorithm for estimating the parameters of the ellipsoidal spacecraft orientation according to the current measurements of modern on-board instrumentation. The algorithm represents an ellipsoidal analogue of the extended Kalman filter, but in most situations it has some advantages of set-theoretic interpretation of the uncertainty compared to the probability one. The efficiency of the developed methods is illustrated by the respective computer simulation and computational experiments. Product Description popup.authors Волосов Віктор Вікторович Губарев Вячеслав Федорович Компанієнко Лілія Григорівна Кравченко Аліна Вікторівна Кунцевич Всеволод Михайлович Личак Михайло Михайлович Максимюк Любов Володимирівна Царук Ніна Петрівна Шевченко Володимир Миколайович popup.nrat_date 2020-04-02 Close