Abstract: Seeing that proper weights are difficult to be attached to the indexes in the existing rockburst disaster assessment system, the paper was aimed at developing a combination weighting algorithm based on coupling AHP method with entropy weight method. To ensure the consistency between subjective-and-objective weights and preference coefficients, an optimization decision model was built by introducing Lagrange function, by which the proper CW of each index can be obtained and the factors influencing predicting rockburst can be ranked in order of influence . On the basis of the basic theory of TOPSIS, the CW-TOPSIS model for comprehensive evaluating rockburst was built for analyzing closeness degree of samples and predicting rockburst ranking level. The model was applied to the panel 83003 in Laohutai coal mine .The panel was predicted to have a medium risk of rockburst and the tectonic stress would be the leading factor causing the rockburst. The prediction conforms with the reality.

Key words: rockburst prediction, combined weights(CW), preference coefficient, CW-TOPSIS evaluation model, primary and secondary factors