利用人工神经网络建立水滑石晶化过程的关联方程

Neural Networks for Study of Mg-Al Hydrotalcite of Continuous Aging Process

采用人工神经网络方法对镁铝水滑石的连续晶化过程进行了研究，以便找出最佳的晶化条件。利用神经网络建立的模型，分析了成核料液的浓度、流量以及晶化温度、晶化时间对形成的水滑石晶体的影响，并由此建立了关联方程。研究表明在浓度保持在0．4～0．6mol／mL的范围内比较好，流量要控制在18mL／min左右为好，晶化时间在400min左右比较好，晶化温度选择100℃较好。

By means of artificial neural network（ANN）, continuous aging proess of Mg-Al hydrotalcite is under review to seek optimal operating condition. The effects of continue aging process such as concentration, flow, temperature and aging time for hydrotalcite lattice on operating condition is analyzed. On the basis of experimental data, the model of relation equations is built. The optimal condition requires a concentration of 0.4 - 0.6 mol/mL, a flow at 18 mL/min and aging time at 100℃.