摘要
不确定性信息的处理是目前制造业的关键和难点,“信息熵”的应用为解决此类问题提供一种新途径,但目前国内用于实际并不多。文章以变切削深度加工过程为例,用神经网络控制,发现系统响应速度慢;改用基于信息熵的目标函数,其中选用均匀分布函数作为不确定系统的输出分布概率函数,发现响应速度提高,但震荡次数增多;后用最大熵原理来取得概率函数,取得很好的控制效果。这为熵的优化理论提供实践证明。
How to deal with uncertain information is the key and difficult problem in the modem manufacturing, and information entropy is a good idea to the problem. In this paper, varied cutting depth process is trialed to study the application effect of entropy The output responses slow when controlled by BP nettral network. It becomes fast, but shocks more times while minimum entropy theory is applied to the controller, where the probability of output is set to uniform distributed function. Finally the controller is reversed to get the probability according to maximum entropy theory, better output is get this time.
出处
《组合机床与自动化加工技术》
2005年第10期59-61,共3页
Modular Machine Tool & Automatic Manufacturing Technique
关键词
信息熵
神经网络
切削加工
智能控制
information entropy
neural networks
cutting process
intelligent control
作者简介
常少莉（1979-）,女,陕西宝鸡人,华南理工大学在读研究生,主要研究方向：现代制造与计算机控制,（E-mail）bengnv@163.com.
