4. Vulnerability Index Method Based on Partition Variable Weight Theory

A new theory and approach for evaluating coal seam floor water bursting is described in this chapter. To quantitatively analyze the effects of each of these factors on water inrush and their relative importance to water inrush, and then to make a comprehensive evaluation, this study has utilized information fusion technology, geographic information technology, and modern mathematical method to propose a vulnerable index method based on zoning variable weight theory. An information fusion architecture from multi-source to evaluate water bursting has been built according to information pretreatment, target assessment, situation assessment, risk assessment, and process assessment of the five-step process with application of information fusion technology, and information fusion algorithm has been determined by combining the linear (AHP) and nonlinear (ANN) mathematical methods. With the information fusion architecture and information fusion algorithm, an information fusion model--JDL floor water inrush information fusion model based on GIS is built after quantitatively determining the impact of every factor on water inrush and their relative importance. Based on coal seam floor water inrush model evaluation method with constant weight, we use variable weights theory and K-means clustering algorithm of dynamic clustering to divide variable weight range threshold of each index of main factors, and construct a conclusive theoretical method of state variable weight vector and adjustable weight parameters. Then, we propose a new theory and method to evaluate the risk of water bursting, which is vulnerable index method based on zoning variable weight theory. Zoning variable weight model not only can reveal the influences of the sudden changes of index values of these main factors on water inrush in the evaluation area, but can also reflect the influence due to the change of one factor under the influence of several main factors combined with each other, without being neutralized by other indexes. A key technical problem with vulnerability assessment of coal seam floor water bursting evaluating is solved through changing weights of evaluation area units with the changes of index values.