الفهرس | Only 14 pages are availabe for public view |
Abstract This thesis addresses the problem of thermal generation scheduling. This problem is considered one of the most important power system planning problems. Thermal units are specially addressed since they form the major part of the generation plants in the Egyptian Electrical Power System. In the scheduling problem, a schedule of turning ON and OFF the generating units is prepared and the share of every online unit is determined over a certain scheduling horizon. The objective of the scheduling problem is to minimize the overall operating cost without violating any of the constraints. The scheduling problem consists of two main parts; the economic dispatch problem (EDP) and the unit commitment problem (UCP). The former belongs to nonlinear programming problems while the later is a combinatorial optimization problem. Dynamic economic dispatch problem (DEDP) is a special type of the generation scheduling problem in which all the units are considered online during the entire scheduling horizon. In the recent years, Artificial Intelligence (AI) techniques proved themselves as very efficient tools in tackling various electrical power system applications. Artificial Neural Networks (ANN) is one of the most important AI techniques in which the human brain is mimicked. In literature, ANN was successfully applied to many electrical power system applications. Fault diagnosis, power system control, machine control, and load forecasting are among the famous examples of these applications. All these applications can be classified as pattern recognition and classification problems. As an optimization tool, special type of ANN called Hopfield Neural Network (HNN) was designed to minimize an energy function. |