【摘要】 研究目的是建立复杂非线性动态响应时间序列预测模型。摩擦和柔性的影响使柔性机构动态响应为高度非线性,实时控制的难度很大。应用改进的Elman动态神经网络技术,建立了柔性机构运动参数动态响应预测模型,对柔性机构的运动参数进行实时预测。通过机构动态响应的预测实例,证明该方法计算速度较快,精度较高,为实现复杂大系统的实时控制提供了一个有效可行的方法。更多还原

【Abstract】 The purpose of the research is to establish complicated nonlinear parameters time series prediction model.Considered influ-ence of friction and flexibleness,the characters of flexible mechanism are highly nonlinear.The dynamical responses of mechanism become more uncertain because of those influence.It is more difficult to control the mechanism at the real time.Via improved El-man dynamical artificial neural network,the prediction model was established to forecast kinetic parameters.A flexible mechanism example was applied to test this method.The results proved that the calculate speed of the model was fast and the precision was high.The method provided an available way on controlling at the real time for complicated large systems.更多还原