Uma técnica para modelagem de cenários consistentes na previsão da demanda por energia elétrica

• Ano de defesa: 1986
• Autor(a) principal: Ribas, José Roberto
• Orientador(a): Mariotto, Fábio L.
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Link de acesso: http://hdl.handle.net/10438/11126

The available qualitative techniques for modeling scenarios have been recognized by their extreme limitation, evidenced in the beginning of the activities of the process, as the initial conceiving step itself. The main restrictions have been: • unexistence of a tool wich tests the internaI structural consistency of the model, or through the utilization of economic relationships with theoric foundation, but with inperfect interface with the environment, or through the use of binary variations for validation tests; • 'a priori' fixation of possible scenarios, are generally classified under three adjectives - optimist, more plausible and pessimist - biased exactly by the attributes of people who provide this information. The study deals with the employment of a tool for iteration among a technique that helps on the generation of models, based on relational logic with about the real world decisions. It aims forecasting system, four values variations, and expectations based on the knowledge of those who make on building an exploratory qualitative where scenarios are achieved through essentialy intuitive and descriptive approach, for electricity· regional demand. This kind of approach, introduced by J. Gershuny, mainly aims with the supply of a methodological aid for consistency of scenarios qualitatively generated. The development and structure of the model is done by steps, beginning with a single relationship and following with introductions of variables and effects wich improve the explanation ofthe model. The study introduces a set of relationships for the regional demand of electricity in the main expediture sectors residencial, comercial and industrial - and the resulting scenarios for most plausible variations of its exogenous variations. At the end it is concluded. that such technique is useful in models that:. • include social variables that are relevants and aren't easy to measure; • believe in the importance of \ externaI consistency among results generated by the model and those expected for decision making; • let the responsability for those who make decisions to understand model's conceptual structure foundation. Once this procedure has been adopted, the author recommends that the model would be validated through an iterative adjustment process with the participation of who make decisions. The quantitative techniques can be used after, with the model as a consistency element.