机构地区: 北京航空航天大学
出 处: 《玻璃钢/复合材料》 2007年第1期6-8,共3页
摘 要: 研究目的是通过人工神经网络方法重构玻璃钢天线罩曲面。径向基函数(RBF)神经网络具有很强的非线性逼近能力,根据给定的天线罩外表面数据及采集的天线罩内表面数据,采用RBF神经网络对玻璃钢天线罩进行逆向工程曲面重构。典型工程实例的计算结果证明,该方法的拟合、重构精度高,并且训练速度快,具有很高的实用推广价值。 The objective of the research is to reconstruct FRP radome surface through the artificial neural network. Radial basis function (RBF) has strong capability of nonlinear approximation. External data of FRP radome are given and inner data of FRP radome collected by reverse engineering technique. RBF neural network is applied to reconstruct FRP radome surface. A typical example is employed to test the approach. The results prove the precision of fitting and reconstructing precise. The train rate of RBF neural network is fast. The method is worth practice and generalizing.
关 键 词: 径向基函数 逆向工程 曲面重构 玻璃钢 天线罩
领 域: [化学工程]