机构地区: 燕山大学机械工程学院
出 处: 《塑性工程学报》 2005年第3期67-70,共4页
摘 要: 在圆锥形零件拉深成形智能化控制的研究基础上,分析了盒形件拉深智能化控制需要解决的技术关键。采用LabVIEW虚拟仪器技术,可以令人满意地实现拉深过程中信号的监测与控制。采用神经网络技术,可实现材料性能参数的实时识别,利用LM(LevenbergMarquarat)优化算法达到了2‰的网络误差。最佳工艺参数的预测模型尚待深入研究,该文提出3种参考方案,即理论成形三极限预测模型、神经网络/模糊推论预测模型以及自适应反馈控制模型。 Based on the research achievements of intelligent deep drawing for an axis-symmetric part, the key technologies of intelligent deep drawing for a rectangular box are discussed in this paper. Signal monitoring and control can be realized nicely in the process of deep drawing by applying the technology of Lab VIEW VI (virtual instrument). Real-time identification of material properties can be achieved with artificial neural network, and the network error is stepped downward to 2 by the application of LM (Levenberg-Marquarat) optimization algorithm. The prediction model of optimal processing parameter remains to study in depth, and three reference schemes are put forward in this paper, i.e. three theoretical forming limits model, neural network/fuzzy inference model and adaptive feedback control model.
领 域: [金属学及工艺]