Abstract: A new method based on rough sets and genetic algorithm was presented to predict the rockburst tendency in this paper. Taking the rockburst tendency of working face of the stope for example, the influencing factors of rockburst tendency were taken as the condition attributes, and the result of rockburst tendency of working face was taken as the decision attribute. The engineering data were chosen to make up the sample set. The sample set was randomly divided into two parts. One was the training sample set, and the other was the testing sample set. The attribute values of the two sets were quantified respectively, and then the decision tables of the training and testing samples were set up. Genetic algorithm was used to reduce the decision table of the training samples. Rough sets were used to extract the decision rules of rockburst tendency from the reduction results. Then the rockburst tendency of the testing samples were predicted by these decision rules. The results show that the predicted results agree well with the actual ones, which proves that the method is effective and available and can be applied to predict the rockburst tendency in actual engineering.

Key words: rockburst tendency, prediction, rough sets, genetic algorithm, decision rules