(Universidade Federal do Pará, 2013) LEITE, Jandecy Cabral; TOSTES, Maria Emília de Lima
Due to the industrial growth in Northern Brazil, especially at the Manaus Industrial Pole (PIM), mathematical tools are needed to assist experts in selecting and sizing harmonic filters that neutralize the harmful effects of harmonics generated by nonlinear loads and ensure compliance with power quality standards. Additionally, passive harmonic filters (PHF) can act as reactive power compensators, achieving significant energy billing savings. This thesis develops a mathematical method and a computational tool for selecting the configuration and design parameters of PHFs in industrial systems, formulating the problem as a multi-objective optimization that considers both harmonic distortion reduction and the project's economic effectiveness, complying with Brazilian and international standards. The computational solution is based on the NSGA-II genetic algorithm, generating a Pareto front of optimal solutions that allow designers to select the most suitable option. The tool includes innovative features, such as automatic selection of filter configuration and number of branches, and considers various operational scenarios. Validation was performed through practical examples that demonstrated the advantages of the proposed method.
