基于神经网络的汽车用AlSi7Mg铝合金压铸工艺优化

Optimization of Die-casting Process of AlSi7Mg Aluminum Alloy for Automobile Based on Artificial Neural Network

建立了高精度、高通用性的AlSi7Mg铝合金压铸工艺的人工神经网络模型,优化了AlSi7Mg合金的铸件性能。通过相关度分析,确定了输入变量对AlSi7Mg合金力学性能的相关性。结果表明,铸态AlSi7Mg合金的力学性能对Mg元素含量最为敏感。不同压铸参数对合金力学性能的相关度排序为Mg元素含量>>浇铸温度>模具预热温度>快充速度。最后预测了Mg元素含量对铸态AlSi7Mg合金力学性能影响,从结果中得出最佳的Mg元素含量应在0.35wt%~0.45wt%之间。

An artificial neural network model of die-casting process with high precision and universal AlSi7Mg aluminum alloy was established, and the casting properties of the AlSi7Mg alloy were optimized. The correlation of input variables on mechanical properties of AlSi7Mg alloy was determined by correlation analysis. The results show that the mechanical properties of as-cast AlSi7Mg alloy are most sensitive to Mg element content. The order of correlation of different die-casting parameters on the mechanical properties of the alloy is Mg element content >> casting temperature > die preheating temperature > fast filling speed. Finally, the influence of Mg element content on the mechanical properties of as-casting AlSi7Mg alloy is predicted, and it is concluded from the results that the optimal Mg element content should be 0.35wt%-0.45wt%.