A Hybrid Evolutionary Approach Applied to the Economic Dispatch Problem with Prohibited Operating Zones and Uncertainties



Economic Dispatch Problem, evolutionary computation, robust optimization


In order to be able to implement a solution to a real world problem, it must be solved taking into account the uncertainties in the associated model. In fact, there is a possibility that the action of a small uncertainty on a nominal solution obtained for a real world problem may become completely meaningless in practice. This paper presents a hybrid algorithm, HEA-GA-SA, to solve the Economic Dispatch Problem with Prohibited Operating Zones and Uncertainties. HEA-GA-SA considers the worst case of uncertainties. The proposed algorithm proved to be effective to solve this problem. Despite having higher computational and dispatch costs than that of the nominal optimal solutions, the robust solutions found by HEA-GA-SA do not become unfeasible in the presence of uncertainties. Therefore, they can be implemented in the real world. The experiments showed that all nominal optimal solutions analyzed (including two in literature) became infeasible in the presence of uncertainties. They violated some of the constraints imposed on the problem, such as the Prohibited Operation Zones (POZs) or the power intervals in which each generator can operate. It was also observed that the presence of POZs makes the problem more difficult and vulnerable to uncertainty.

Author Biographies

Daniel Freitas Martins, Federal University of Viçosa - Campus Florestal

Daniel Freitas Martins is majoring in Computer Science at the Federal University of Viçosa - Campus Florestal. He is interested in topics related to Optimization, Symbolic Regression, Artificial Intelligence, Biomathematics and Applied Mathematics.

Marcus Henrique Soares Mendes, Federal University of Viçosa - Campus Florestal

Marcus Henrique Soares Mendes holds a PhD in Electrical Engineering from UFMG (2013). He is a Full Professor at UFV - Campus Florestal and is interested in the themes: Evolutionary Computing, Robust Optimization, Interval Analysis, Meta-heuristics, Genetic Programming, Symbolic Regression and Surrogate Models.


Bookmark and Share