![]() | Only 14 pages are availabe for public view |
Abstract Unit commitment is a rising problem whose importance increases day after day due to economical aspects regarding problems in supplying fossil fuel resources, other aspects of reliability of supplied energy are also more stressing now than it ever is. This thesis presents a simulated annealing algorithm to solve the unit commitment problem. Ulili/.ing the powerful aspects of simulated annealing technique of adapting to any cost function forms asking for no special aspects for the production cost pattern such as convexity or linearity. Unit commitment is a very important aspect in modern power system management as the influence of competition and individual producers of energy stress more and more lor cost reduction of energy production schedules. The probabilistic production simulation (PI’S) index of the expected energy not served (EENS) is utilized as a measure for the energy production scheduling reliability. A new flexible reserve setting (I:RS) technique is also presented to obtain a less cost as well as a less EENS solution, i.e. a more reliable solution. The approximate method of cumulants of a random variable is used to get the reliability probabilistic indices, this ensures faster computation and more ability to handle different cases of production schedules to compare between them. By the introduction of such modification to the classical unit commitment problem is hopefully a beginning to open an approach to the unit commitment problem as a multi-objective optimization problem regarding many other factors of network and units constraints, limits and penalty values. |