【摘要】 涂布工序是烟草薄片制造的关键工序之一。涂布率作为涂布效果的重要指标，对烟草薄片的质量有较大影响。针对涂布工序的涂布率预测与工艺参数优化问题，采用灰色关联度分析，提取关键影响因子，将遗传-粒子群算法与BP神经网络模型相结合，搭建了GA-PSO-BP涂布率预测模型。结果表明，GA-PSO-BP模型预测精度显著高于BP模型，实现了对涂布率的有效预测。在此基础上，利用遗传算法求解了涂布工序的最佳工艺参数组合，为烟草薄片生产中涂布工序的参数控制提供了参考。更多还原

【Abstract】 Coating process is one of the key processes in reconstituted tobacco production. As an important index of coating process, coating rate has great influence on the quality of tobacco sheets. Aiming at the parameter optimization problem of coating process, 12 possible influencing factors are considered, and the key influencing factors are extracted by grey correlation analysis. The coating rate prediction model is established by BP neural network, which is combined with GA-PSO algorithm. The results show that the prediction accuracy of GA-PSO-BP model is significantly higher than that of BP model, and the coating rate can be predicted effectively. On this basis, genetic algorithm is used to solve the optimal process parameters of coating process, which provided reference for parameter control of coating process in tobacco sheet production.更多还原