Regresar

Mathematical Modeling and Optimization of Downdraft Gasifiers Using Artificial Neural Networks (ANN) and Stochastic Programming Techniques

Abstract:

The research study explores the modeling and optimization of multi-objective operation of biomass gasification facilities using of Artificial Neural Networks (ANN) and Stochastic non-linear Programming methods. This study underpins the modelling by starting from the classification of the information derived from the systemic analysis of the gasification facilities. The study is based on the multi-objective mathematical modeling of these facilities through the different optimization and Neural Networks techniques specified in the literature. A 3N experimental plan with 3 replicas is made to generate four models according to their performance indicators using Neural Networks, with satisfactory results and their evaluation based on regression of coefficients. The standard errors are calculated using biomasses with low, medium and high caloric power biomass. The experimental installation and the developed data acquisition systems are presented to validate the results. Numerical experimentation and the analysis show that such models could be used for developing operational system for existing design of downdraft installations.