一种新的非线性离散动态大系统递阶控制的神经网络方法

A New Method for Hierarchical Optimization of Nonlinear Large-Scale Discrete-Time Dynamic Systems on the Basis of Neural Networks

本文给出了一种新的具有全集成化特征的、求解非线性动态大规模系统递阶控制问题（LOCP）的神经网络模型（LHCNN），该神经网络具有结构简洁、紧凑、高效的特点，适合于动态大规模系统的实时优化控制．

A novel neural network (LHCNN) for hierarchical control of nonlinear large-scale discretetime dynamic systems is put forward in the paper. The neural network is constructed compactly and concisely,and can be used to solve the problems effectively. It is more suitable for practical applications to real-time optimal control of large-scale dynamic systems.