Experimental parameters influencing the transesterification of rubber seed oil (RSO) to biodiesel using alumina (Al2O3) impregnated on calcined eggshells (Al2O3/calcined eggshells) were studied in the present work. Parameters were optimized using response surface methodology (RSM) and artificial neural network (ANN). A conversion of 98.9% was observed for RSO at optimum conditions of 12:1 methanol: oil molar ratio, 3 (wt%) catalyst concentration and 4 (h) of reaction time. A significant quadratic model with molar ratio as the most influencing process parameter and a coefficient of determination, R2, of value equal to 0.9379 is observed from RSM analysis. Best validation performance of 5.8595 at epoch-1 and R2 value equal to 0.9740 was observed from ANN modeling. On comparing RSM and ANN models, it is concluded that ANN is a better tool for predicting the conversion of RSO to biodiesel with minimum error.

中文翻译：

---

用 Al2O3 作为多相催化剂浸渍的煅烧蛋壳从橡胶籽油生产生物柴油：RSM 和 ANN 优化的比较研究

在目前的工作中，研究了使用浸渍在煅烧蛋壳 (Al2O3/煅烧蛋壳) 上的氧化铝 (Al2O3) 将橡胶籽油 (RSO) 酯交换为生物柴油的实验参数。使用响应面方法 (RSM) 和人工神经网络 (ANN) 优化参数。在甲醇：油摩尔比为 12:1、催化剂浓度为 3 (wt%) 和反应时间为 4 (h) 的最佳条件下，观察到 RSO 的转化率为 98.9%。从 RSM 分析中观察到一个重要的二次模型，其中摩尔比是影响最大的工艺参数，决定系数 R2 的值等于 0.9379。从 ANN 建模中观察到在 epoch-1 和 R2 值等于 0.9740 时的最佳验证性能为 5.8595。在比较 RSM 和 ANN 模型时，