Search

In this paper we consider a controlled dynamical system consisting of two objects in the presence of external perturbations. The dynamics of the first object (the main object of the system) and the second object (the auxiliary object of the system) are described respectively by linear and nonlinear discrete-time recurrent vector equations. In the system under consideration there are two levels of control: basic level (the level I) that is dominant level and auxiliary level (the level II) that is subordinate level. The quality of the control process on every level of the control system is estimated by corresponding convex terminal functional. Both the control levels have different informational and control connections defined in advance. In this paper we study the problem of optimization of guaranteed result for program control the final state of phase vectors of the objects in the presence of external perturbations. For this problem we propose a mathematical formalization of two-level hierarchical minimax program control the final state of the dynamical system in the presence of external perturbations and incomplete information. In this paper for solving of the investigated problem is proposed the algorithm that has a form of a recurrent procedure of solving a linear and a convex mathematical programming problems, and a finite optimization problems. [ABSTRACT FROM AUTHOR]

Nowadays, with the popularity of the computers, various industries and fields are widely applying computer information technology, which brings about huge demand for a variety of application software. In order to develop software meeting various needs with most economical cost and best quality, programmers must design efficient algorithms. A superior algorithm can not only soul up one thing, but also maximize the benefits and generate the smallest overhead. As one of the common algorithms, dynamic programming algorithms are used to solving problems with some sort of optimal properties. When solving problems with a large amount of sub-problems that needs repetitive calculations, the ordinary sub-recursive method requires to consume exponential time, and dynamic programming algorithm can reduce the time complexity of the algorithm to the polynomial level, according to which we can conclude that dynamic programming algorithm is a very efficient compared to other algorithms reducing the computational complexity and enriching the computational results. In this paper, we expound the concept, basic elements, properties, core, solving steps and difficulties of the dynamic programming algorithm besides, establish the dynamic programming model of the production planning problem [ABSTRACT FROM AUTHOR]

In this paper, a new nonlinear interval-based programming (NIP) method based on Radial basis function (RBF) approximation models and RBF global search technique method is proposed. In NIP, searching for the extreme responses of objective and constraints are integrated with the main optimization, which leads to extremely low efficiency. Approximation models are commonly used to promote the computational efficiency. Consequently, two inevitable problems are encountered. The first one is how to obtain the global minimum and maximum in the sub-optimizations. The second one is how to diminish the approximation errors on the response bounds of system. The present method combined with RBF global search technique shows a good feature to overcome these problems. High accuracy and low computational cost can be achieved simultaneously. Two numerical examples are used to test the effectiveness of the present method. [ABSTRACT FROM AUTHOR]