基于神经网络的汽车齿圈锻压工艺优化  

Optimization of Forging Process of Automobile Gear Ring Based on Neural Network

作　　者：高洪 徐田恬 GAO Hong;XU Tiantian(College of Machanical and Vehicle Engineering,Changchun University,Changchun 130000,China)

机构地区：[1]长春大学机械与车辆工程学院,吉林长春130000

出　　处：《热加工工艺》2020年第23期100-103,共4页Hot Working Technology

摘　　要：以始锻温度、终锻温度、锻压比和模具预热温度为输入参数,以耐磨损性能(磨损体积)为输出参数,以tansig函数为隐含层传递函数,以purelin函数为输出层传递函数,采用4×20×1三层拓扑结构,构建了汽车齿圈盘体的神经网络优化模型。结果表明,模型的平均相对训练误差5.37%,平均相对预测误差5.98%,模型预测能力较好,预测精度较高。与企业现用工艺相比,采用神经网络优化工艺锻压的20CrMnTi汽车齿圈的磨损体积减小10%,耐磨损性能得到明显提高。Taking the initial forging temperature, final forging temperature, forging ratio and die preheating temperature as the input parameters, the wear resistance(wear volume) as the output parameters, Tansig function as the hidden layer transfer function, and purelin function as the output layer transfer function, the neural network optimization model of the automobile ring gear was constructed by using 4 ×20 ×1 three-layer topology structure. The results show that the average relative training error of the model is 5.37%, and the average relative prediction error is 5.98%. The model has good prediction ability and high prediction accuracy. Compared with that of the current process of the enterprise, the wear volume of 20CrMnTi steel automobile ring gear forged by the neural network optimization process is reduced by 10%, and the wear resistance is significantly improved.

关 键 词：汽车齿圈 神经网络优化 锻压工艺 耐磨损性能 相对训练误差 相对预测误差 

分 类 号：TG319[金属学及工艺—金属压力加工] TG162