The application of hybrid memetic algorithm of global constrained optimization in the search optimal control of nonlinear stochastic systems problem is researched in the paper. The suggested approach is based on control parametrization, which gives an opportunity to reduce the search optimal control problem to nonlinear programming problem. The global optimization memetic algorithm is suggested to solve the nonlinear programming problem. The term "memetic algorithms" is wildly used to denote method, which is based on evolutional, cultural-evolutional or another approach, that uses notions like population and individual learning process or other local improvement procedure to determine global extremum. In the developed algorithm the cultural evolution component is implemented by means of the solution of sub-task optimization using ant colony method or simulated annealing algorithm. Herewith mems (unit of cultural information) are used to generate perfect individual. Several alternative forms of parametrization are considered: in the form of expansion by Legendre polynomials system and in the form of cosine-waves. The efficiency of the proposed algorithms was analyzed by means of created software complex. The problem of damping rotational satellite motion by means of installed engine was solved as the example. Verification of obtained results using the solution detected with local variations method was done. The described results show the sufficient efficiency level and reflect the necessity of further research in this direction.


    Access

    Download


    Export, share and cite



    Title :

    APPLICATION OF HYBRID MEMETIC ALGORITHM IN OPTIMAL CONTROL NONLINEAR STOCHASTIC SYSTEMS WITH INCOMPLETE FEEDBACK PROBLEMS


    Contributors:


    Publication date :

    2018




    Type of media :

    Article (Journal)


    Type of material :

    Electronic Resource


    Language :

    Unknown