机构地区: 燕山大学机械工程学院
出 处: 《机械工程学报》 2005年第4期199-202,共4页
摘 要: 在板材成形智能化控制的四个基本要素中,材料性能参数的实时识别及最优工艺参数的预测是最复杂的两个技术关键。识别和预测精度的高低,将直接影响到智能化控制成功与否。以宽板V形自由弯曲智能化控制为研究对象,采用基于LM算法的前馈神经网络模型,通过实时监测量来实时识别所需的材料性能参数,并预测最优的工艺参数,取得了令人满意的收敛精度。在样本数据范围内,当模型的收敛精度为0.1%时,识别和预测的泛化精度均在5%以内。 In the four basic factors on the intellectuali zation of sheet metal forming, the real-time identifica tion of the material performance parameter and the prediction of the optimum technological parameter are the two most complicated technical keys. The accuracy of identification and prediction will have direct effect on the success of the intelligence control.Taking the intelligence control of V-shape free bending of wide sheet metal as an object of study ,a feed front neural network model is used based on LM algorithm. By means of real-time monitoring and measure to identify the material performance parameter and predict the optimum technological parameter, and satisfied accuracy of convergence is achieved . Within the sample database range, generalization precisions of identification and prediction are within 5% when convergence precision is within 0.1%.
领 域: [自动化与计算机技术] [自动化与计算机技术]