Adaptive and robust evidence theory with applications in prediction of floor water inrush in coal mine. (April 2017)

• Record Type: Journal Article
• Authors: Li, Fenglian; Zhang, Xueying; Chen, Xiaolei; Tian, Yu-Chu
• Other Names: Wu Zhipeng guest-editor.
• Is Part Of: Transactions of the Institute of Measurement and Control. Volume 39:Number 4(2017)
• Page Start: 483
• Page End: 493
• Publication Date: 2017-04
• Subjects: Data fusion -- evidence theory -- joint focal element -- water inrush prediction -- coal mine; Automatic control -- Periodicals; Measuring instruments -- Periodicals; Commande automatique -- Périodiques; Mesure -- Instruments -- Périodiques; 681.2
• DOI: 10.1177/0142331216687816
• Languages: English
• ISSNs: 0142-3312

Abstract

The Internet of Things generates rich information either from different sources or the same source via different measurement methods. This demands data fusion for decision making. Despite the progress in data fusion, existing data fusion techniques, such as the classic Dempster–Shafer evidence Theory, face challenges when dealing with highly conflicting sources of evidence. To address this problem, an Adaptive and Robust evidence Theory (ART) is presented in this paper through a robust combination of conjunctive and disjunctive rules. It is capable of handling both conflicting and reliable sources of evidence. When the sources of evidence are reliable, the conjunctive rule plays a predominant role, whereas if the sources of evidence are in high conflict the disjunctive rule is critical. Our ART approach was compared with existing representative evidence theory methods through two examples, and was further applied in the prediction of floor water inrush in coal mines. The ART approach presented in this paper was demonstrated to behave better than the existing methods.