APPLICATION OF MULTI-EVIDENCE REASONING FOR IDENTIFICATION VECTORS OF FLOW FIELD
In the course of the measurement velocity field, because of measurement error, there are inevitably miss-vectors in the original flow field which is measured. The neural network has been an important means for its characteristic to identify miss-vector. In this paper, based on simulating the identification process of cerebra to miss-vector; a new Hopfield neural network model of multi-evidence reasoning is founded. And the identification performance of this model is tested by numerical simulation experiment. The experiment result indicates that, compared with other network model of single-evidence reasoning, the new network model of multi-evidence reasoning can simulate cerebra ideation better, it has better identification function.
multi-evidence reasoning Hopfield neural networks identification miss-vector flow field
Longhua Wu
Center for Eeo -Environmental Modeling, Hohai University, Nanjing 210098, China
国际会议
第16届亚太地区国际水利学大会暨第3届水工水力学国际研讨会(16th IAHR-APD Congress and 3rd Symoposium of IAHR-ISHS)
南京
英文
1984-1988
2008-10-20（万方平台首次上网日期，不代表论文的发表时间）