登月舱软着陆的非线性神经元控制

A NONLINEAR NEUROCONTROL SCHEME FOR LUNAR SOFT LANDING

本文针对登月舱软着陆过程的控制问题，提出了一种非线性动态逆与状态反馈控制相结合的神经元控制系统设计方案。该方案包含两分：（１）借助前馈神经元网络通过学习逼近任意非线性映射的能力，建立被控系统的非线性动态逆神经元模型，用神经元网络实现被控非线性系统的线性化；（２）在线性化模型的基础上构造系统的神经元最优状态反馈控制器。本文给出的仿真结果显示出神经计算学在航天飞行器控制问题中所具有的潜在能力。

A neurocontrol scheme for lunar soft landing is proposed in this paper,which combines nonlinear dynamic inversion with optimal state feedback The scheme mainly consists of two parts First,the nonlinear dynamic inversion of the controlled object is modeled with an artificial neural network by means of its ability to learn to approximate any functions,and therefore,the controlled object is linearized by the neural inversion model Secondly,based on the linearized system another artificial neural network is used as a controller to realize certain optimal state feedback controllaw Finally,the effectiveness of the scheme described in this paper is investigated by computer simulation The simulation results are encouraging and show that neurocomputation could play important role in control of the future spaceships