UDC 681.515
SETTING WEIGHT MATRIX OF SDRE CONTROLLER WITH BIOINSPIRED OPTIMIZATION ALGORITHMS
A. I. Bobikov, associate professor, RGRTU; This email address is being protected from spambots. You need JavaScript enabled to view it.
A. V. Shabirina, student of group 430M RGRTU; This email address is being protected from spambots. You need JavaScript enabled to view it.
We consider the problem of setting weight matrices of State Dependent Riccati Equation (SDRE) controller using bioinspired algorithms. Particular attention is paid to the particle swarm optimization algorithm (PSO) and bacterial foraging algorithm (BFO). The control object is a spherical tank with liquid which is a complex nonlinear object. So it is impossible to apply the conventional methods of nonlinear control to this object. Method of state-dependent Riccati equation (SDRE) is applicable to wide variety of nonlinear systems, and assures great flexibility of design by the choice of weighting matrices Q(x) and R(x). The aim of this article is to find the weight matrices of SDRE controller specified nonlinear control object, by bacterial foraging algorithm, particle swarm optimization algorithm and by the method based on hybridization of two considered algorithms. In conclusion, the proposed algorithms have been successfully applied for given non-linear object, and we present comparison of simulation results.
Key words: non-linear object, sub-optimal control, particle swarm optimization, bacterial foraging optimization, weight matrix, State-dependent Riccati equation (SDRE) method, SDRE controller, minimization of objective function.