基于神经网络的汽车用铝合金热处理工艺优化

Optimization of Heat Treatment Process of Aluminum Alloy for Automobile Based on Neural Network

采用7×35×2三层拓扑结构,以锻造铝合金牌号、退火温度、退火时间、固溶温度、固溶时间、时效温度、时效时间作为输入层参数,以耐磨损性能和冲击性能作为输出层参数,构建了汽车用锻造铝合金热处理工艺优化神经网络模型,并进行了模型训练、预测验证和生产线应用。结果表明,汽车用锻造铝合金用神经网络优化模型的优势较明显,预测性较好,且精度性较高。和生产线传统工艺相比,通过神经网络优化模型热处理的试样磨损体积减小22%、冲击吸收功增大了79%。

Thaking the forging aluminum alloy grades, annealing temperature, annealing time, solution temperature, solution time, aging temperature and aging time as input layer parameters, and taking wear resistance and impact properties as output layer parameters, the neural network model of heat treatment process optimization of wrought aluminum alloy for automobile was built. The model was trained, and the forecast verification and the production line application were carried out. The results show that the advantage of the neural network optimization model of forging aluminum alloy for the automobile is obvious, the prediction is good, and the precision is high. Compared with traditional production line, the wear volume of the samples heat-treated by the neural network optimization model is reduced by 22% and the impact absorption energy is increased by 79%.