基于BP神经网络的磨削加工误差补偿研究

Study on Error Compensation of Grinding Processes Based on BP Neural Networks

通过分析磨削加工过程中的误差产生的原因,使用学习训练后的BP神经网络建立了双面磨床磨削加工误差补偿模型,详细阐述了网络模型的组织结构与核心算法,提出了实现磨削加工实时误差补偿技术的硬件组成方法,并通过实验验证了模型的使用效果,结果证明,基于BP神经网络的磨削加工误差补偿模型可以有效减小磨削误差,提高磨削加工精度。

Through synthetic analysis the causes of grinding error, an error compensation method of double-sided grinding machine＇s grinding processes based on training BP neural networks was builded. The structure and algorithem of the networks model were expatiated details in this paper. Finally, implementation method of hardware system for real-time error compensation was given, and the use effect of model was tested through the experiment. The results show that the error compensation networks model can reduce grinding error efficiently and improve the accuracy of grinding processes.