摘要
为了提高数控机床故障预测的能力,针对BP神经网络在数控机床故障预测中出现的收敛速度慢和训练容易陷入局部极值问题,提出了一种基于萤火虫算法优化BP神经网络的数据机床故障诊断算法。文章详细介绍了常见的数控机床故障类型和分类,在萤火虫优化算法和BP神经网络的基础上,建立了萤火虫算法优化BP神经网络的数控机床故障诊断模型,并提出了基于该模型的算法。该模型和算法采用萤火虫算法优化BP神经网络的初始权值和阈值,优化后的BP神经网络能对测试集进行更好的预测。实验结果表明,萤火虫算法优化BP神经网络的预测误差明显小于GRNN和PNN算法。该模型和算法具有很好的预测能力,可以快速、准确地完成数控机床故障诊断研究。
In order to improve the ability of numerical control machine tool fault prediction,this paper proposes a BP neural network fault diagnosis algorithm,which is based on firefly algorithm. The algorithm is against the problem of slow convergence speed and the training easy to fall into local extremum. This paper introduced the common fault types and classification of numerical control machine tool. On the basis of the firefly optimization algorithm and the BP neural network,the model of fault diagnosis is established. Meanwhile the algorithm of the above fault diagnosis is proposed. The model and algorithm adopts the firefly algorithm to optimize the BP neural network＇s initial weights and threshold. The experimental results show that prediction error of the BP neural network,optimized by the firefly algorithm,is less than the GRNN or PNN algorithm significantly. The prediction ability of the model and algorithm is good and the process of fault diagnosis can finish quickly and accurately.
出处
《机械设计与研究》
CSCD
北大核心
2014年第6期77-80,89,共5页
Machine Design And Research
基金
安徽省科技厅资助项目(1301zc02004)
安徽省教育厅资助项目(KJ2013Z197)
