<p>This article considers the problem of optimizing the control of forecasting and estimating the state of a&#xa0;dynamical object. The dynamics of the controlled object are described by a&#xa0;quasilinear discrete-time vector-matrix recurrent equation containing a&#xa0;phase vector, control action vector, and disturbance vector. The realizations of the phase vector of the control object and the disturbance vector in each time period are by a&#xa0;convex compact polyhedrons, and the realizations of the control vector are bounded by a&#xa0;finite set, in the corresponding finite-dimensional vector spaces. A deterministic minimax approach is used to formalization the problem under study, allowing for optimization of the guaranteed solution result. To assess the quality of the estimation result, an objective function is used whose values are determined by the Chebyshev radius of the reachability set (forecasting set) in a&#xa0;given time period, formed relative to the phase vector of the controlled object. Within the framework of the developed mathematical model, problems of <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\varepsilon\)</EquationSource> </InlineEquation>-optimal open-loop control and adaptive control for minimax estimation of the phase state of the controlled object in the final time period are formulated. To solve this problem, a&#xa0;technique is proposed that is implemented as a&#xa0;finite set of one-step operations on vectors in finite-dimensional vector spaces, solving problems of linear and convex mathematical programming and discrete optimization. The results presented in this article can be used in the development and creation of navigation systems, as well as automated control systems for complex technical and economic facilities.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Optimization of Adaptive Control of Forecasting of the Phase Vector State of a Quasilinear Discrete-Time Controlled Dynamical System

  • Andrey F. Shorikov

摘要

This article considers the problem of optimizing the control of forecasting and estimating the state of a dynamical object. The dynamics of the controlled object are described by a quasilinear discrete-time vector-matrix recurrent equation containing a phase vector, control action vector, and disturbance vector. The realizations of the phase vector of the control object and the disturbance vector in each time period are by a convex compact polyhedrons, and the realizations of the control vector are bounded by a finite set, in the corresponding finite-dimensional vector spaces. A deterministic minimax approach is used to formalization the problem under study, allowing for optimization of the guaranteed solution result. To assess the quality of the estimation result, an objective function is used whose values are determined by the Chebyshev radius of the reachability set (forecasting set) in a given time period, formed relative to the phase vector of the controlled object. Within the framework of the developed mathematical model, problems of \(\varepsilon\) -optimal open-loop control and adaptive control for minimax estimation of the phase state of the controlled object in the final time period are formulated. To solve this problem, a technique is proposed that is implemented as a finite set of one-step operations on vectors in finite-dimensional vector spaces, solving problems of linear and convex mathematical programming and discrete optimization. The results presented in this article can be used in the development and creation of navigation systems, as well as automated control systems for complex technical and economic facilities.