论文部分内容阅读
在讨论各种非线性动态系统辨识模型的基础上,给出相应模型的神经网络实现方案,并首次提出实现非线性动态系统的回归状态模型的新型神经网络结构—神经网络状态空间辨识模型.从理论上证明了使用神经网络实现这些模型的可行性.
Based on the discussion of various nonlinear dynamic system identification models, a neural network realization scheme of the corresponding model is given. A new neural network structure - neural network state space identification model is proposed for the first time to realize the regression state model of nonlinear dynamic system. The feasibility of using neural network to realize these models is proved theoretically.