基于遗传算法和神经网络的攻丝扭矩预测  

Torque Forecasting for Tapping Based on Genetic Algorithm and Neural Network

作　　者：吴金妹[1] 

机构地区：[1]泉州师范学院物理与信息工程学院

出　　处：《北华大学学报（自然科学版）》2012年第2期241-244,共4页Journal of Beihua University（Natural Science）

基　　金：泉州师范学院校自选项目(2008KJ03)

摘　　要：针对振动攻丝工艺参数与攻丝扭矩之间的高度非线性关系问题，利用神经网络的基本原理，结合遗传算法理论建立了工艺参数和攻丝扭矩之间的关系模型．将网络模型的预测结果与实验结果进行了比较，显示出了GA—BP预测模型的可靠性．Aiming at complicacy relationship between vibration tapping technological parameter and tapping torque, the theory of neural network was introduced and joining genetic algorithm, the relation function of technological parameter and tapping torque were founded. The comparison has been carried out between the results predicted by network and the experiment results, which shows that the GA-BP network is stable.

关 键 词：神经网络 遗传算法 振动攻丝 攻丝扭矩 

分 类 号：TG506.5[金属学及工艺—金属切削加工及机床]