LAUSR

Abstract Sinter cooler, as sinter cooling process equipment, is used to cool the hot sinter to be lower than specified temperature. While the hot gas produced in sinter cooling process is usually recovered and utilized to reduce the energy consumption in recent years. In present study, a global sensitivity-based optimization design method synthesizing both the performances of sinter cooling and energy conservation of sinter cooler was proposed. Firstly, based on the Latin hypercube sampling technique and radial basis neural network approach, an accurate metamodel of the sinter cooler was established. Secondly, to select the significant operating parameters, a global sensitivity analysis was performed to evaluate exactly the sensibility of the indicator parameters with respect to operating parameters. In the analysis, a function decomposition method was firstly introduced into sinter cooler analysis to obtain the global sensitivity indices of operating parameters. Finally, based on the results of global sensitivity analysis, a multi-objective optimization was performed to achieve the optimal indicator parameters and obtain corresponding operating conditions using nondominated sorting genetic algorithm II (NSGA-II). At the end, a Pareto frontier as a group of non-dominated optimal solutions was suggested for engineers to be selected according to actual situation.