![]() | Only 14 pages are availabe for public view |
Abstract This thesis is concerned with Genetic Algorithms (GAs), and their applications in control systems. The two types of GAs, classical GA. (CGA.), and real-coded GA (RGA), are considered. Two problems face both types of GAs: the first problem is the convergence to suboptimal solution, and the other problem is the slow speed of convergence. These problems highly reduce the potential of Gain real time control applications. The main contribution of this thesis is the modification of GAs to ensure fast convergence to global• solution. The proposed algorithms are: The Modified Immune Genetic Algorithm (MIGA), Hybridized Immunity Mechanism based on Evolutionary Algorithm (HIMEA), and a new using for genetic algorithms on-line. The second contribution is the practical application of a GA to an industrial process. Two models are obtained for multizone electrical heat furnace using GA. To show the effectiveness of the proposed methods, several applications are carried out. The applications are concentrated in three directions: non-linear . |