Khomenko V. Multiobjective synthesis of controls systems with anisotropic regulators based on stochastic multiagent optimization

The object of the study are management processes multimass electromechanical systems with parametric uncertainty keruvannya.Metoyu object of the thesis is to improve control accuracy multimass electromechanical systems with uncertain parameters on the basis of multi-criteria approach to the synthesis of robust stochastic keruvannya.Metody research tasks is a solution based on the methods of automatic control theory formation of stochastic problems of robust control multibody systems; Methods for describing the state space of the control systems; stochastic methods of synthesis of robust control systems on the basis of minimizing anizotropiynoi system standards for the correct definition of the vector control objectives; methods of analysis and synthesis of stochastic robust control systems for the synthesis of anisotropic regulators; Modern methods of computer mathematics, analytical and numerical optimization methods for the implementation of the proposed methods and stochastic robust control algorithms; methods of multi-criteria optimization for constructing a set of Pareto - optimal solutions, numerical methods bagatoroevoi stochastic multi-agent optimization to solve the problem of multi-criteria synthesis stochastic robust control, numerical simulation methods and MATLAB software package for the synthesis of stochastic robust control systems, simulation of dynamic characteristics and evaluation indicators of the quality of the synthesized systems. results of the thesis are used in the synthesis, development and implementation of automated control systems of mobile units, technological processes and complexes of scientific and production enterprise "Hartron Arcos" (Mr .. Kharkov). Scientific novelty: for the first time developed a method of multicriteria synthesis stochastic robust control of multibody systems with uncertain parameters through the use of anisotropic controllers, allowing for a variety of requirements for management systems in various modes of operation and improve control accuracy compared to existing systems with typical controllers; further developed the method of selecting a target vector stochastic robust control by forming appropriate multi-criteria optimization problem - the vector of nonlinear mathematical programming. The possibility of such an approach is shown based on the concept of functional multi-tenant set of vectors able to synthesize a stochastic robust control system to ensure compliance with the requirements for the control system; developed method for solving multiobjective programming problem, to which reduced multicriteria synthesis of stochastic robust control by constructing a set of Pareto - optimal solutions based on bagatoroevoi stochastic multi-agent optimization, enabling significantly narrow the field of possible optimal solutions of the problem of multicriteria optimization and thereby significantly simplify the procedure a single selection of optimal solutions. Scope of use: the design of computer networks, the educational process.7452