Abstract: Rockburst is one of the major engineering geological hazards in underground space development and mining.Rockburst propensity prediction is a major problem in rock engineering that must be solved.Aiming at the characteristics of multi-factor comprehensive influence in the process of rockburst prediction, the cloud atomization theory is used to test the rationality of weight fusion, and the comprehensive weight of the test is obtained.A multi-dimensional cloud model rockburst prediction method is used to generate a hierarchical integrated cloud that integrates multiple indicators.Finally, the reliability and practicability of the model are verified by several sets of typical rockburst examples at home and abroad, and compared with CRITIC-cloud model, entropy-cloud model and RS-TOPSIS model.The results show that the weighted fusion method based on cloud model theory can obtain more reasonable comprehensive weights.The multi-dimensional cloud model is effective for rockburst propensity prediction, and it can be used to determine the rockburst intensity classification intuitively, quickly and effectively.