الفهرس Only 14 pages are availabe for public view
 |  | from | 200 |  |  |
| | | from | 200 | | |
Abstract
This thesis proposes a modified version of the pal tide swarm optimizer for solving a class of non-linear Optimization problem A modification is applied to the originally developed particle swarm optimizer in terms of both the velocity arid the way each individual in the swarm updates its position inside the problem space Accordingly, members of the Swarm are prevented front exploring areas outside the feasible domain. Moreover, the chance of trapping. in local optima is reduced The proposed algorithm is used for evolving neural network to perform different tasks in power systems Also. the proposed algorithm is applied directly lo solve non-convex economic load dispatch problems Simulations and comparisons with conventional and/or modern techniques indicate that the proposed algorithm achieves very good accuracy in reasonable amount of time Furthermore, the algorithm utilizes very simple arithmetic and logic operations