Aplicação do Filtro com Reamostragem por Importância Sequencial (SIR) na Simulação de Dados de Processo de Precipitação: Desenvolvimento de um Aplicativo Computacional para Auxílio na Tomada de Decisão em Refinaria de Alumina

Abstract

This study describes the development of a computational application based on mathematical models of mass balance and particle filters to simulate the precipitation process in the Bayer circuit. The goal is to provide a tool to assist process engineers in daily decision-making, enhancing the control of productivity and alumina quality. Simulated results were compared with real plant data, showing average relative errors of 4.7%, validating the model's effectiveness.