基于支持向量回归机SVR的钻削力在线预测分析
|
摘要
钻削力预测是深孔加工质量监测和工艺参数优化等工作的重要手段。针对现有预测方法样本需求量大、网络结构复杂以及易陷入局部极值等问题,提出了一种新的基于支持向量回归机的钻削力预测方法。首先,对钻削轴向力和扭矩预测的主要影响因素进行分析,然后确定预测模型的输入输出参数,进一步建立了基于支持向量回归机的钻削力预测模型。仿真实例的预测结果表明:利用所构建的SVR预测模型对10组样本扭矩和轴向力预测的平均相对误差分别为1. 13%和1. 26%,远小于其他预测方法,说明所建立的模型预测精度高,具有较强的泛化能力。
Drilling force prediction is an important means of quality monitoring and parameter optimization in deep hole processing. Aiming at the problems of large sample demand,complex network structure and easy to fall into local extreme value,a new prediction method of drilling force based on support vector regression machine is proposed. Firstly,the main factors affecting the prediction of drilling axial force and torque are analyzed,then the input and output parameters of the prediction model are determined,and the prediction model of drilling force based on support vectorregression machine is established. The simulation results show that the average relative error of the SVR prediction model is 1. 13% and 1. 26% respectively,which is much smaller than other prediction methods. It shows that the SVR prediction model has high prediction accuracy and strong generalization ability.
Drilling force prediction is an important means of quality monitoring and parameter optimization in deep hole processing. Aiming at the problems of large sample demand,complex network structure and easy to fall into local extreme value,a new prediction method of drilling force based on support vector regression machine is proposed. Firstly,the main factors affecting the prediction of drilling axial force and torque are analyzed,then the input and output parameters of the prediction model are determined,and the prediction model of drilling force based on support vectorregression machine is established. The simulation results show that the average relative error of the SVR prediction model is 1. 13% and 1. 26% respectively,which is much smaller than other prediction methods. It shows that the SVR prediction model has high prediction accuracy and strong generalization ability.
引文
[1]唐宁,张军,李原,等.基于ANFIS的CFRP轴向钻削力预测[J].航空制造技术,2012,19:67-70.
[2]杨进,李鹏南,唐思文,等.金刚石涂层钻头钻削碳纤维增强复合材料钻削力和撕裂预测[J].机械科学与技术,2015,34(7):1123-1126.
[3]于海夫,薛惠峰. CFRP/铝合金叠层结构动态钻削力预测研究[J].机械科学与技术,2017,36(1):68-73.
[4]张柏寿,刘长毅.一种叠层结构钻削力预测模型的研究[J].机械制造与自动化,2017(4):105-109.
[5]刘洋,李鹏南,陈明,等.采用BP神经网络预测碳纤维增强树脂基复合材料的钻削力[J].机械科学与技术,2017,36(4):586-591.
[6]陈菁瑶,苗鸿宾,刘兴芳,等.基于混沌神经网络的钻削力在线预测[J].组合机床与自动化加工技术,2018(3):110-113.
[7]侍红岩,吴晓强,张春友.基于支持向量回归机的刀具寿命预测[J].工具技术,2015,49(11):47-50.
[8]方瑞明.支持向量机理论及其应用分析[M].北京:中国电力出版社,2007.
[9]白冰.基于支持向量回归机的切削力预测[J].工具技术,2016,50(10):32-35.
[10]宋亮,杨志霞,刘芳.应用双支持向量回归机的风速预测模型[J].重庆理工大学学报(自然科学),2013,27(8):85-89.
[2]杨进,李鹏南,唐思文,等.金刚石涂层钻头钻削碳纤维增强复合材料钻削力和撕裂预测[J].机械科学与技术,2015,34(7):1123-1126.
[3]于海夫,薛惠峰. CFRP/铝合金叠层结构动态钻削力预测研究[J].机械科学与技术,2017,36(1):68-73.
[4]张柏寿,刘长毅.一种叠层结构钻削力预测模型的研究[J].机械制造与自动化,2017(4):105-109.
[5]刘洋,李鹏南,陈明,等.采用BP神经网络预测碳纤维增强树脂基复合材料的钻削力[J].机械科学与技术,2017,36(4):586-591.
[6]陈菁瑶,苗鸿宾,刘兴芳,等.基于混沌神经网络的钻削力在线预测[J].组合机床与自动化加工技术,2018(3):110-113.
[7]侍红岩,吴晓强,张春友.基于支持向量回归机的刀具寿命预测[J].工具技术,2015,49(11):47-50.
[8]方瑞明.支持向量机理论及其应用分析[M].北京:中国电力出版社,2007.
[9]白冰.基于支持向量回归机的切削力预测[J].工具技术,2016,50(10):32-35.
[10]宋亮,杨志霞,刘芳.应用双支持向量回归机的风速预测模型[J].重庆理工大学学报(自然科学),2013,27(8):85-89.