Comparação de técnicas metaheurísticas de otimização no processo de hidrólise do amido de mandioca para obtenção de álcool

Abstract

Synonymous with renewable fuel that pollutes less in comparison with those derived from petroleum, alcohol returned to occupy a prominent place in the energy scenario of the country and has begun to be desired by many other countries. In addition to sugar cane, ethanol can also be obtained through other sources of raw materials such as: corn, sugar beet, barley, cassava, among others. What prevents the production of alcohol for commercial purposes using these other sources, is the high cost. The application of computational optimization techniques can provide a solution regarding the minimization of these costs. In this work metaheuristics techniques were applied in the optimization of the hydrolysis of cassava starch for production of fuel alcohol and an analysis of its economic viability as an alternative to alcohol from cane sugar. The objective is to optimize the hydrolysis process of cassava starch using the techniques Genetic Algorithm (GA), Simulated Annealing (SA), GASA, which is a hybrid algorithm which implements the AG to generate an optimized initial solution and then applies the SA to improve this solution, SAG, which is another hybrid algorithmimplementing the SA to generate an initial population of results and then apply the processes of the GA population to improve this result and extract the best solution, and tabu search. The results obtained with both techniques were compared and it was found that the genetic algorithm, the GASA and SAG have found the optimal solution for the process with determination. It is worth mentioning that the SAG has found the optimal solution in a larger number of tests, demonstrating a best performance. Finally, simulated the impacts, environmental and social production and consumption of alcohol from cassava and also estimated the benefits from the adoption of this product as an alternative energy source for the alcohol from sugar cane.