摘要
采用平衡的倒摆小车所记录下来的数据，经处理后用有师学习方法来训练前馈神经网络。训练后的神经网络可以替换专家，并且表示了一组专家尚未意识到或者表达不出来的规则。用神经网络表示的规则与使用Ｑｕｉｎｌａｎ的ＩＤ３推理算法推导出的规则进行了比较。实验结果表明，神经网络算法学习出来的规则较ＩＤ３算法推导出的规则更为有效，且更有应用价值。本文最后尝试将该规则应用于火箭的姿态控制。
Presents a method of training a feedforward neural network using supervised learning scheme to balance an inverted pendulum and cart system. The data used to train the neural network was obtained from a human expert doing the same task. The trained neural network can replace the human expert and uncovers a set of rules which the human expert can not express. Comparisons was made between the neural network learned rule and a decision tree rule deducted by Quilan's ID3 induction algorithm using the same set of data. Experiment results show that the neural network learned rule is more robust. At the same time, this finds that the neural network learned rule can be modified to do a similar and more important task——the attitude control of a rocket.
出处
《数据采集与处理》
CSCD
1998年第2期131-135,共5页
Journal of Data Acquisition and Processing
关键词
专家系统
神经网络
控制器
intelligent control
expert system
neural network
machine learning
