摘要
管道流流型是两相流体输送系统的一个具有重要工程意义的参数.目前对该参数的检测方法主要采用软测量方法.本文提出了一种采用光学层析技术与CHNN神经网络相结合的新方法来解决气-固管道流流型识别问题.采用了光学层析技术获取流动固相的空间分布信息(即投影数据),而采用神经网络对投影数据进行了聚类分析.实验结果表明,该方法的处理速度快及流型识别率高,具有工程实际意义.
The pipe flow regime is an important parameter of pneumatic conveyor system and it is mainly measured by means of soft measurement at present. This paper puts forward a new method combining optical tomography with CHNN (Continuous Hopfield Neural Networks) to solve the problem of regime identification of gas-solid pipe flow. The optical tomography is adopted to obtain the spatial distributing information (i.e. projection data) of flowing solid phase, while the neural network is used to cluster the projection data. The experimental results show that this method performs with high processing rate and good identification effect, as well as engineering practicability.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2004年第4期523-526,共4页
Pattern Recognition and Artificial Intelligence
基金
国家自然科学基金(No.59975019)
关键词
神经网络
气-固管道流
流型识别
光学测量
Neural Networks
Gas-Solid Pipe Flow
Regime Identification
Optical Measurement
