Πρόβλημα ένταξης μονάδων παραγωγής
Master Thesis
Author
Πατσιούδης, Δημήτριος Γ.
Date
2011-09-22View/ Open
Abstract
In this essay we study the unit commitment problem (UCP) of production units in a 24 hour time horizon. As in many research areas, the electrical power generation systems work following specific cycles. Studying those cycles raises the problem of efficient commitment at the lowest possible cost. The main methods of addressing the problem can be classified into three categories: traditional optimization methods, heuristic methods, and methods of artificial intelligence. Some mathematical models are inspired by biology which can be applied to artificial societies entities, such as colonization and evolving artificial ant colony system (Ant Colony System - ACS). An evolution of ant colony system, the Con-ANT (Continuous ANT) is studied and applied in the above mentioned problem, namely the commitment of electricity generation plants in an horizon of 24 hours. Also an application is developed using the programming language VB.NET of Microsoft, which simulates such a system and searches for a solution to the aforementioned problem. Moreover we conducted experiments using this application comparing results of the Con-ANT to other existing algorithms addressing the same problem, under the same plant configuration.