Programação diária da operação de sistemas termelétricos utilizando algoritmo genético adaptativo e método de pontos interiores

Abstract

The growth of the electric energy consumption in the last years has generated the need of the increase in the amount of power sources, making the electricity sector undergo some large changes. This has provided the search for tools that promotes a better efficiency and security to the electrical power systems. A planning problem that is considered important in the daily operation of the power systems is the Unit Commitment, where the time schedule of the operation is defined, determining which machines will be online or offline, and which are the operating points. Those units must operate by load variation, respecting the operative and security constraints. This research proposes the resolution of the problem for the short-term planning, taking a set of constraints associated with the thermal generation and the power system. Among them, we can highlight the output power variation constraints of the machines and the security restrictions of the transmission system, avoided in most Unit Commitment studies. This problem is nonlinear, mixed-integer and has a large scale. The methodology used involves the utilization of an Adaptive Genetic Algorithm, for the Unit Commitment problem, and the Interior-Point Primal- Dual Predictor–Corrector Method, for DC power flow resolution in economic dispatch problem. Furthemore, this research proposes the implementation of cross-over and mutation operators of Genetic Algorithm based on a ring methodology applied in Unit Commitment matrix. The results were obtained through simulations in a mathematical simulation software, using the IEEE test systems with 30 bus and 9 generators, and another with 24 bus and 26 generators. The validation of the algorithm was done by comparing the results with other works in the literature.