BP神经网络补偿快速成型系统闭环误差的研究

RESEARCH INTO THE COMPENSATION OF CLOSED LOOP ERROR OF RAPID PROTOTYPING SYSTEM BY BP NERVE NETWORK

提出快速成型系统闭环误差反馈的必要性,将BP神经网络引入快速成型技术,建立了快速成型误差补偿模型,探求出一种快速成型系统闭环误差的反馈方法,并通过实例进行了验证。结果表明,基于BP神经网络控制的最大绝对误差和相对误差均比传统误差补偿方法大为降低。

The necessity of closed loop error feedback of rapid prototyping（RP） system is proposed and an error compensation model of RP is created. Based on this model, a feedback method of RP is established and proved. The results show that both the maximum absolute error and the relative error based on BP nerve network are much lower than those of the traditional method.