Operating Parameter Optimization for Abnormal Operating Modes Correction in Sintering Process

The sintering process in iron ore smelting plays a critical role in the production of suitable raw materials for blast furnaces. However, the complexity of the process, which involves multiple variables, can lead to abnormal operating modes such as over-burning and under-burning, negatively affecting production stability and product quality. Traditional operating parameter optimization methods based on empirical rules are often inadequate due to the high non-linearity and interdependency of process parameters. This study proposes an operating parameter optimization strategy for correction of abnormal operating modes in the sintering process, in which a multilayer feedforward neural network approach is used to predict the burn-through point, which can identify abnormal modes. In addition, a strategy for optimizing the operating parameters is developed using the nondominated sorting genetic algorithm II, which is used to adjust key operating parameters to prevent over-burning and underburning. The result of the experiment using industrial realworld data shows the effectiveness of the proposed method in reducing abnormal operation occurrences by 37.75 %, improving the stability of the sintering process.