In this study, a hybrid of support vector regression and a modified tunicate swarm algorithm (SVR-MTSA) strategy is developed to optimize the process parameters for recovery of Pb from the residual of the liberator cells. The lead recovery efficiency in the selected process was strongly nonlinear and depended on several process parameters including temperature, processing time, the content of coke, Na₂CO₃, and Fe in the precursor. The results confirmed a good agreement between the efficiencies obtained experimentally and those predicted by the model. It is also shown that using the optimal process parameters suggested by the model, achieving a Pb recovery of more than 99% was possible. Sensitivity analysis using the proposed SVR-MTSA model revealed that temperature, coke content, processing time, Na₂CO₃ amount, and Fe content of the raw material had the most significant effect on the efficiency of the lead recovery, respectively.

在这项研究中，开发了支持向量回归和改进的被囊类群算法 (SVR-MTSA) 策略的混合体，以优化从解放细胞残余物中回收 Pb 的工艺参数。所选工艺中的铅回收效率具有很强的非线性，取决于几个工艺参数，包括温度、处理时间、焦炭含量、Na ₂ CO ₃，以及前体中的 Fe。结果证实了实验获得的效率与模型预测的效率之间的良好一致性。还表明，使用模型建议的最佳工艺参数，实现超过 99% 的 Pb 回收率是可能的。使用建议的 SVR-MTSA 模型的敏感性分析表明，温度、焦炭含量、处理时间、Na ₂ CO ₃量和原料的 Fe 含量分别对铅回收效率的影响最显着。