Notice of RetractionImprovement Research of Evidence Theory in Mine Water Inrush Prediction

Abstract

Since the highly conflicting evidence could not be combined effectively through D-S evidence theory, a novel D-S data fusion method based on evidence quality is introduced in this paper. The concrete algorithm for the reliability of observer and the measure of quality function value based on observer reliability are proposed. In the data fusion, adopting the measure of evidence quality, the collected data from multi-sensor is assigned to the different weight according to the reliability of observer, and the probability assignment value is correspondingly adjusted. The improved D-S evidence theory, along with combining the highly conflicting evidence, is applied successfully to the prediction of mine water inrush. The experimental results show that the method is effective.