模糊神经网络控制器及其在航天器姿态控制系统中的应用研究

A Fuzzy Neural Network Controller and Its Application Study for the Spacecraft Attitude Control System

在航天器姿态控制领域，模糊控制技术作为一种辅助控制方法得到越来越多的关注。本文提出了一种基于模糊聚类算法和分布式模糊神经网络的模糊神经网络控制器，并将其用于《东方红三号》卫星应急星地大回路姿态控制系统之中。仿真研究表明，这种模糊神经网络控制器能够有效地从样本数据提取信息，实现分区域控制，并且具有较强的鲁棒性。该方法具有较为满意的控制效果。

In the field of spacecraft attitude control, the controllers based on fuzzy logic have received increasing attentions as auxiliary control strategies. A fuzzy neural network controller based on fuzzy clustering algorithm and distributed neural network is presented in this paper. This paper makes a study of fuzzy neural network controllers for the emergent satellite-ground long loop attitude control system of DFH-3 satellite, and some good results are achieved by simulations. The study demonstrates the ability to construct the heterogeneous controller from pattern data collected and the good robustness of the fuzzy neural network controller discussed above.