Research of FDI for Artificial Neural Network Failure Detection Method Based on Model Residual

摘要：

FDI （Failure detection and isolation ）is widely used in field of flight control, because the failure detection method base on mathematic model has the shortness that the precision of the failure detection method based on linear model depends on accuracy of the mathematic model in some degree, the FDI method based on neural network overcomes this shortness. In this paper hierarchy neural network is used to simply the structure of network and improve the performance of network according to the failure tree analysis and the research based on neural network FDI method to one kind UAV is done. Simulation in MATLAB is done to testify to this method.

关键词： Fault Detection and Isolation (FDI) artificial neural network residual hierarchical simulation

作者: SHI Yan QI Xiaohui

作者单位: Department of Optical and Electrical Engineering, Ordnance Engineering College, Shijiazhuang 050003, China

会议类型: 国际会议

会议名称: 第七届国际测试技术研讨会

会议地点: 北京

会议语种:英文

在线出版日期: 2007-08-05（万方平台首次上网日期，不代表论文的发表时间）