摘要
表面粗糙度是表面加工质量的重要指标之一,影响零件的使用寿命,因此在线预测表面粗糙度具有重要意义。由于BP神经网络的算法本身存在容易陷入局部极小值、收敛速度慢和全局搜索能力弱等缺陷,故采用遗传算法优化BP神经网络的结构和初始参数并设计基于进化神经网络的学习算法,建立BTA钻削在线预测的神经网络模型。仿真和实验结果表明,进化的BP神经网络能够很好的预测表面粗糙度,克服了BP神经网络容易陷入局部极小值的问题,为BTA钻削的研究提供了新的思路。
The surface roughness is one of the important indicators of machined surface quality and affect the service life of parts, So the on-line prediction surface roughness has great significance. Focusing on some disadvantages in BP neural network algorithm , such as low rate of convergence, easily falling into local minimum point and weak global search capability, so there is a genetic algorithm to optimize BP neural network configuration and initial parameters, the on-line prediction model of surface roughness in BTA drilling was proposed. The simulation and experimental results show that the evolution of the BP neural network can well predict the surface roughness and overcome the problems of easily falling into local minimum point, it also provides a new study and analysis method for the studying of BTA drilling.
出处
《组合机床与自动化加工技术》
北大核心
2014年第1期26-28,共3页
Modular Machine Tool & Automatic Manufacturing Technique
基金
国家自然科学基金(51175482)
山西省国际合作项目(2012081030)
关键词
BTA钻削
进化神经网络
表面粗糙度
遗传算法
BTA drilling
evolutionary neural network
surface roughness
genetic algorithm
