基于伪阻抗学习的神经网络状态估计方法

A Method of State Estimation of Neural Network Based on Pseudo-impedance Learning Algorithm

针对反向传播算法收敛速度慢，且常收敛于局部极小值的缺陷，讨论了伪阻抗学习算法；并利用神经网络的学习能力和非线性特性，讨论了非线性动态系统的状态估计方法。仿真表明，该学习算法收敛速度快、稳定性好。将该算法用于状态估计，具有良好的状态跟踪能力。

A pseudo-impedance learning algorithm is discussed to overcome the drawbacks of the slow convergence speed and the local minimum of the back-propagation algorithm.In the meantime,by utilizing the learning capacity and nonlinear characteristic of neural networks,a method of state estimation is discussed for nonlinear dynamic systems. Simulation shows that the learning algorithm has a fast convergence speed and good stability.The ability of following the tracks of states can be obtained when the algorithm is applied to state estimation.