摘要
利用模糊神经网络的学习功能和处理不确定情况的能力,合理划分煤粉在水平管道中的多种流动状态,解决煤粉流量的测量问题,并充分利用工作经验和专家知识,大大缩短网络训练时间.通过现场数据实验表明,此方法提高了测量的准确度,并克服了传统神经网络物理意义难以解释的不足.
Fuzzy neural network combines the capability of neural network in learning and the capability of fuzzy reasoning in handling uncertain information. Several states of flowing coal powder in horizontal pipe can be divided reasonably by using fuzzy neural network, and it can solve problem in the measurement of coal powder flowrate. In addition, work experience and expert knowledge are used sufficiently. The experiment results show that the accuracy is improved using this method, and the shortcoming is avoided that the physical meaning of neural network can not be interpreted.
出处
《东北大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
1998年第3期293-295,共3页
Journal of Northeastern University(Natural Science)
基金
国家"八五"科技攻关项目
关键词
模糊神经网络
隶属函数
煤粉
流动状态
高炉
fuzzy neural network, membership function, coal powder,flow state.