A Method Integrating Self-Organizing Maps to Predict the Probability of Barrier Removal Zhicheng Zhang and Fre´de´ric Vanderhaegen University of Valenciennes,

CONTENTS 30.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498 30.2 A Method Integrating Self-Organizing Maps Algorithm . . . . . . . . . . . . . . . . . . . 498 30.3 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 503 30.4 Discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507 30.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 510

Barrier Removal (BR) is a particular violation in the field of human reliability analysis. The analysis of BR may be undertaken in terms of benefit, cost and possible deficit. Moreover, during the BR analysis, the data of all three indicators for each barrier class are usually provided in terms of some performance criteria, e.g., productivity, quality, safety and workload. This paper addresses the problem of how to balance the multi-variable BR data on different performance criteria and capture the complex nonlinear relationships that exist between these sub-criteria. The application of artificial intelligent techniques, which can analyze the multi-dimensional BR indicator data with the sophisticated visualization technique, is vital for the sustainable study of BR.