Abstract: To quantify risk of rock burst,a GRA-SPA-EW decision model was built,using the theories of GRA and SPA.A set of pairs for indicators and risk grades of rock burst was established,absolute relevance degrees and comprehensive contact degrees between them were calculated,and the hazard level of rock burst was determined.Then,the indexes' weights of rock burst were determined by method of entropy weight,and an identical degree matrix of the most highly vulnerable area and lowest vulnerable area of rock burst was constructed,also its risk susceptibility was evaluated.Finally,a typical coal mine was taken as an example,five working faces were selected as research objects,and the risk of rock burst on each working face was evaluated.Results show that the model comprehensively considers the identity,difference,and opposition between index factors of the evaluation object,quantitatively analyzes the contact coefficients,and highlights the differences between evaluation objects through multi-attribute decision making of EW.

Key words: rock burst, grey relational analysis(GRA), set pair analysis (SPA), entropy weight(EW), multi-attribute decision, risk evaluation