Title :
Numerical method of network operator for multiobjective synthesis of optimal control system
Author :
Diveev, A.I. ; Sofronova, E.A.
Author_Institution :
Sci. Dorodnicyn Comput. Centre, RAS, Moscow, Russia
Abstract :
The problem of multiobjective synthesis of optimal control is considered. It is necessary to find the control as a function of problem space coordinates. Initial values of the control object are in closed bounded domain. Functionals of control quality include multiple integrals over domain of initial values. To solve the problem of multiobjective synthesis the numerical method was developed. This method is based on genetic algorithm and special data structure, network operator. The numerical method allows automatic search of mathematical equations that describe dependence of control from the problem space coordinates. The solution of the synthesis problem for space vehicle descent with indeterminate initial conditions at entry in the dense layers of atmosphere is given.
Keywords :
control system synthesis; genetic algorithms; numerical analysis; optimal control; genetic algorithm; mathematical equations automatic search; multiobjective synthesis; network operator; numerical method; optimal control system; problem space function; special data structure; Automatic control; Automation; Control system synthesis; Control systems; Data structures; Equations; Genetic programming; Mathematical model; Network synthesis; Optimal control;
Conference_Titel :
Control and Automation, 2009. ICCA 2009. IEEE International Conference on
Conference_Location :
Christchurch
Print_ISBN :
978-1-4244-4706-0
Electronic_ISBN :
978-1-4244-4707-7
DOI :
10.1109/ICCA.2009.5410619