Multi-dimensional Normal Cloud Model For Rockburst Classification Prediction

Rockburst is a kind of complex geological hazard under high geostress environment.With the continuous increase of the burial depth of underground engineering,the problem of rockburst is becoming more and more serious,which poses a threat to the safety of construction personnel and production equipment.In order to improve the safety and reliability of underground engineering,forecasting and controlling rockburst hazards has become more important today.The establishment of rockburst classification prediction model accuratly is the basis and prerequisite for preventing rockburst disasters.On the other hand,the excavation environment of deep projects is complex,and there are many uncertain factors in the prediction process of rockburst.Therefore,comprehensive consideration of the interactions between multiple influencing factors,reasonable allocation of forecast index weights are of great significance for establishing a reasonable and effective rockburst grading prediction model and obtaining accurate rockburst grading judgment resultsAccording to the existed research results of rockburst classification predictions,we can get the main factors affecting rockburst occurrence are lithology,surrounding rock stress,and energy.Based on the rockburst occurrence mechanism,the rock uniaxial compressive strength ?c,rock uniaxial tensile strength ?t,the maximum tangential stress of surrounding rock ??,elastic strain energy index of rock wet,rock stress coefficient??/?c and rock brittleness coefficient ?c/?t are used to establish a rockburst prediction index system.At the same time,different index grading standards are determinted according to the characteristics of prediction indicators,which provides a reference standard for the classification prediction model of rockburstA weight calculation model of rockburst was established based on the improved FC-RS weight calculation method.In view of the differences of attribute characteristics of rockburst prediction indicators,the improved FC-RS weight calculation model is used to evaluate the contribution rate of different prediction indicators.The contructed model can make full use of the initial sample data,avoid the loss of effective information during data processing,and improve the lacking of zero importance of non-nuclear attributes,which make the weight distribution of rockburst prediction index is more reasonableA simple multidimensional normal cloud model was established according to the rockburst prediction index system.The digital characteristics determined by the rockburst classification standard and the rockburst sample data are inputed to the multidimensional forward cloud generator,then the determination of rockburst samples were obtained,and the rockburst level of samples were judged according to the principle of maximum determination.By comparing with the actual situation,the multidimensional normal cloud model used for the study of rockburst grading prediction is reasonable and feasibleA weighted classification prediction model of rockburst was constructed by combining the improved FC-RS weight calculation model and the multidimensional normal cloud model.By compared with the one-dimensional normal cloud model,finite interval cloud model,and fuzzy mathematical model,it is found that the contructed model has the advantage of considering the interaction between multiple prediction indicators and effectively deal with the uncertainty in the rockburst prediction process,which can ensure the prediction results are more reliable.The established model is used to the Sujiapo Phosphate Mine,the calculated results of rockburst grades were completely consistent with the actual project,which further verifying the accuracy and reliability of the modelThe rationality of weight distribution of rockburst prediction index is improved,the correlation between several influencing factors of rockburst is considered,and the applicability of established model in practical engineering is verificated in this study,which provided a new research path for the study of rockburst grade prediction.