【摘要】 空间轮轨接触是铁道车辆系统动力学仿真对象模型中的重要柔性环节,基于材料线弹性假设的Hertz理论,可以给出两弹性体一般接触的复杂的隐式非线性关系。支持向量回归是统计学习理论中提出的新一代学习算法,轮轨接触计算可以使用它离线学习到一个显式的非线性关系,其逼近精度和表示复杂度可得到很好的折中,有利于铁道车辆系统动力学仿真的在线计算。更多还原

【Abstract】 Wheel rail contact is an important flexible element in the object model for the system dynamics simulation of rail vehicles. Hertz theory deduces an implicit and complex nonlinear relation of two contacting objects based on material linear elastic assumption.SVR (Support Vector Regression)is a learning algorithm of new generation proposed by statistical learning theory,with which an explicit nonlinear relation of wheel rail contact compntation can be learned off-line.The approximation accuracy and expression complexity is well compromised, which is useful in the on-line computation of system dynamics simulation for the rail vehicles.更多还原