摘要
针对环境温度变化及数控机床本身发热导致机床加工精度降低的问题,提出一种结合萤火虫算法与BP神经网络的数控机床热误差补偿方法。该方法两次使用萤火虫优化算法,第一次用于筛选传感器数据,第二次用于优化BP神经网络,然后通过优质训练库和GSO-BP神经网络进行建模,实现高精度的误差补偿。实验结果表明,本文方法预测误差与实际误差更加接近,补偿后残余误差更小,且补偿稳定性好。
Aiming at the problem that ambient temperature and CNC machine tool heat generation led tothe low machining precision of machine tool, the author put forward a CNC machine tool thermal errorcompensation method in combination of firefly algorithm with BP neural network. This method utilizedfirefly optimization algorithm twice, once for screening sensor data and another time for optimizing BPneural network. Through quality training library and GSO-BP neural network, the model was establishedto achieve high precision error compensation. Test results showed the error predicted by the method saidin this article was close to actual error with less residual error after compensation and good compensationstability.
出处
《天津冶金》
CAS
2017年第5期35-39,共5页
Tianjin Metallurgy
关键词
热误差补偿
萤火虫算法
BP神经网络
数控机床
thermal error compensation
firefly algorithm
BP neural network
CNC machine tool
