摘要
为提高对钛合金叶片精锻过程中模具磨损量分析的效率和准确率,以TC11钛合金叶片精锻成形过程为研究对象,通过有限元分析软件Deform-3D进行数值模拟,结合修正的Archard磨损模型情况建立了叶片精锻过程的模具磨损样本数据,应用遗传算法-极限学习机(GA-ELM)模型预测模具磨损量。以模具磨损量作为输出参数,以相关的叶片精锻工艺参数作为输入参数,对模具磨损量进行预测;并结合遗传算法优化的GA-BP神经网络模型、原始ELM模型的预测结果进行对比。最后,通过Deform有限元软件分析的模具磨损量验证了GA-ELM模型预测结果的精度和可靠性。结果表明,利用GA-ELM模型预测的模具磨损量具有较高的精度,与其他算法相比具有优越性。
In order to improve the efficiency and accuracy of analysis on mold wear amount in the fine forging process of titanium alloy blade,taking the fine forging process of TC11 titanium alloy blade as the research object,the numerical simulation was carried out by the finite element analysis software Deform-3 D.Combined with the modified Archard wear model situation,the data of mold wear sample for blade in the fine forging process was established,and the mold wear amount was predicted by the genetic algorithm-extreme learning machine(GA-ELM)model.Taking the mold wear amount as output parameters,taking the related blade fine forging process parameters as input parameters,the mold wear amount was predicted.Combined with GA-BP neural network model optimized by genetic algorithm and original ELM model,the prediction results were compared.Finally,the mold wear amount analyzed by finite element software Deform verifies the accuracy and reliability of GA-ELM model prediction result.The results show that the mold wear amount predicted by the GAELM model has a higher accuracy,and it is superior to other algorithms.
作者
梅益
刘洪波
罗宁康
李亚勇
龙孟伟
Mei Yi;Liu Hongbo;Luo Ningkang;Li Yayong;Long Mengwei(College of Mechanical Engineering,Guizhou University,Guiyang 550025,China)
出处
《锻压技术》
CAS
CSCD
北大核心
2020年第10期130-136,共7页
Forging & Stamping Technology
基金
贵州省科技支撑计划(黔科合支撑[2019]2019)。
